Thanks for the great article! Intel surprised me a bit since this Ice Lake wonder looks like Zen2 or even better. Now, the only question is when we can get Ice Lake cores in Xeon E/W form?
We all knew what the CPU result was going to be, and really this is a great foot in the door for AMD for when they have a competitive mobile product. Given the relentless execution after execution, I'm hopeful for something good here. I suspect they've left laptops until last as that's where a lot of Intel's dirty tactics are in play and will be the hardest market to crack. Working with MS will undoubtedly give AMD some ideas of where they can optimise for Windows on their end and also get MS to optimise on theirs. Given that AMD are actually competitive now, I suspect Microsoft are more interested in dealing with them.
The surprise for me was the GPU leap from Intel. I was expecting AMD's GPU option to be far better and Intel's to just be the usual perfunctory stuff, good only for multimedia and maybe Minecraft. That.... could be a sign of things to come, methinks.
I really hope you're right. It's great to finally see competition. I am actually considering upgrading my main PC for the first time in years. What I don't want is Intel abusing its market position to cane AMD in the GPU market with an inferior product.
I did not expect such a big improvement in IPC from Intel either. Looks like they were holding back. A desktop equivalent is going to snatch the CPU crown back from AMD. I just hope the platform price is not insane.
Really interesting. It just proves that AMD needs to upper their game in laptops, still it is jsut telling me that I should wait for Zen 2 mobile at this point.
7nm AMD laptop chips are expected to show up in January at CES, potentially even shipping at that point. With the 15% IPC boost, that will help a lot. If Zen3 brings another 15% IPC boost(something that has been rumored), then in January 2021 we might see AMD beating Intel in the laptop space(since Intel doesn't improve IPC all that often).
I think it will probably be higher once Zen3 ships as node will either have had a process improvement or shrunk to 5nm. Current AMD mobile is pretty bad, hope they get competitive and bring laptop prices down.
Ehhh, Sunny Cove's successor will bring an equal IPC increase, from what Intel has stated.
AMD has themselves to blame. Why tf would you delay the architecture of both GPU and CPU and process node as well like that? I mean, the money lost from doing that, as OEMs will have little reason to move away from Intel's superior products, outweighs the money "saved" from this delay, no? Not like a Zen 2 APU won't come anyway, so why not do it before rather than after?
If AMD are smart they'll jump one architecture ahead. Starting with Zen 4 in 1.5 year's time, they should be smart enough to jump straight from Zen 2 APU to Zen 4.
I don't think it is just the IPC boost. The two chips in the comparison here had memory bandwidth differences of 80%!
Simply looking at a few of the SPEC INT tests, that was very clear that if AMD even on Zen had similar memory bandwidth to the Ice Lake chip, it likely would have been spitting distance. GPU workloads are also very heavily memory bandwidth constrained and AMD's Vega here was slightly ahead of Intel. If it had 80% higher memory bandwidth it probably would have been 20-30% faster in many of those games/benchmarks.
Even some of the not heavily memory constrained workloads like Handbrake, faster memory does improve performance. That 80% memory bandwidth difference may well have been a 5-8% performance hit to AMD.
BUT AMD shipped it with only DDR4-2400 support. It is the chip they brought to the fight.
If their Zen 2 manages to both have lower platform power (doesn't need to be parity with Intel or better, but 10% better would go a huge long way towards making it less of a decision for a lot of people), brings its 15% better IPC and if it keeps its clock speeds AND manages to bring DDR4L-3000/DDR4x-3760 or whatever compatibility and suspect Intel is done for in the laptop space.
That should give AMD several more wins in CPU performance, bring it to parity or near parity in most of the others leaving Intel with only a few wins in that. For GPU performance, AMD wouldn't even really need to update the iGPU. Just give it that extra bandwidth and Zen 2 CPU behind it and it likely would be kicking Intel's butt by 20-40%. Upgraded on top of having that, yes please.
I don't think it is about IPC or CPU power anymore. Yes, AMD is slower, but not in a way that would bother me. It is more about power management that is and always was sore spot of AMD in notebooks. And no - 7nm is not the solution unfortunately. It may help under load, but under standby it can actually hurt battery life (leakage is higher). Most improvements in battery life are done by architectural changes (more specialized units, powering down unused parts etc).
Well, this driver is still under development and in the first test, they were running cooler than Intel, so likely there is tuning to be done, still. They just haven't done much laptop work in recent years. They're going to be behind on more than the hardware, which is why Microsoft's involvement is really promising, really. Of course they still have some minor hardware matters to sort out, as you point out.
Do you think a lot of help came from the sizeable memory speed differences? Do you think it benefitted cpu or gpu more? Is there a way to underclock the Intel memory so you can see the differences that the memory brings? Thank you.
It's perplexing: AMD created an entirely new SKU for Microsoft (the AMD RYZEN™ 7 3780U Microsoft Surface® Edition Processor), but AMD still left this CU 11-equipped iGPU at the barebones 2400 MHz speed.
Why not rate this MS-only SKU at 2933MHz or 3200MHz? My only thought: high-speed DDR4 was just going to exacerbate AMD's already-too-large power consumption.
It isnt just speed, the intel chip uses LPDDR4X. That's an entirely different beat from LPDDR4, let alone normal DDR4.
AMD would need to redesign their memory controller, and they have just...not done it. The writing was on the wall, and I have no idea why AMD didnt put LPDDR4X compatibility in their chips, hell I dont know why intel waited so long. The sheer voltage difference makes a huge impact in the mobile space.
You are correct, pushing those speeds at normal DDR4 voltage levels would have tanked battery life.
Sigh, it is just speed. DDR4-2400 to DDR4-3200 is simply speed: there is no "entirely new controller" needed. The Zen+ desktop counterpart is rated between DDR4-2666 to 2933.
LPDDR4X is almost identical to LPDDR4: "LPDDR4X is identical to LPDDR4 except additional power is saved by reducing the I/O voltage (Vddq) to 0.6 V from 1.1 V." Whoever confused you that LPDDR4X is "an entirely different beat" from LPDDR4 is talking out of their ass and I caution you to believe anything else they ever say.
And, no: DDR4-3200 vs DDR4-2400 would've tanked battery life, but simply made it somewhat worse. DDR4-3200 can still run on the stock 1.2V that SO-DIMM DDR4 relies on, but it's pricier and you'd still pay the MHz power penalty.
I don't think RAM speed/voltage has ever "tanked" a laptop's battery life: shaking my head here...
I'm quite sure you're wrong here. The problem isn't the memory itself (as long as you get default 1.2V modules, which exist up to ddr4-3200 itself), but the cpu. Zen(+) cpus require higher SoC voltage for higher memory speeds (memory frequency is tied to the on-die interconnect frequency). And as far as I know, this makes quite a sizeable difference - not enough to really matter on the desktop, but enough to matter on mobile. (Although I thought Zen+ could use default SoC voltage up to ddr4-2666, but I could be wrong on that.)
Ryzen had huge problems with memory speed and even compatibility at launch. No doubt they had to play it safe on laptops. They should have it mostly sorted out with Zen 2 laptop, it is why the notebooks are a gen behind where as intel notebook are usually a gen ahead.
We both agree it would be bad for battery life and a clear AMD failure. But, the details...more errors:
1. Zen+ is rated up to DDR4-2933. 3200 is a short jump. Even then, AMD couldn't even rate this custom SKU to 2666 (the bare minimum of Zen+). AMD put zero work into this custom SKU (whose only saving grace is graphics and even that was neutered). It's obviously a low-volume part (relative to what AMD sells otherwise) or such a high-profile design win.
2. If AMD can't rate (= bin) *any* of its mobile SoC batches to support even 2666MHz at normal voltages, I'd be shocked.
For any random Zen+ silicon, sure, it'd need more voltage. The whole impetus for my comments are that AMD created an entire SKU for Microsoft and seemed to take it out of oven half-baked.
Or, perhaps they had binned the GPU side too much that very few of those CU 11 units could've survived a second binning on the memory controller.
So all that being said, yes it had a huge impact. GPU based workloads are heavily memory speed dependent. Going from 2400 to 3200MHz likely would have seen a 10-25% increase in the various GPU benchmarks (on the lower end for those that are a bit more CPU biased). That changes AMD from being slightly better overall in GPU performance to a commanding lead.
On the CPU side of things, many of the Intel wins were on workloads with a lot of memory performance needed. Going from 2400 to 3200 would probably have only resulted in the AMD chip moving up 3-5% in many workloads (20-40% in the more memory subsystem dependent SPEC INT tests), but that would have still evened the playing field a lot more.
Going to 3766 like the Intel chip would have just been even more of the same.
Zen 2 and much higher memory bandwidth can't come soon enough for AMD.
It's not about binning, they couldn't support that memory and keep within their desired TDP because they would have to run infinity fabric at a higher speed. They could have used faster memory and lower CPU and/or GPU speed but this is the compromise they settled on.
AMD make/design for a client what that client wants, in this case, MSFT as "well known" for making sure to get (hopefully pay much for) what they want, for only reasons that they can understand.
this case, AMD really cannot say "we are not doing that" as this would mean loss of likely into the millions (or more) vs just saying "not a problem, what would you like?"
MSFT is very well known for catering to INTC and NVDA whims (they have, still do, even if it cost everyone many things)
still they AMD and MSFT should have "made sure" to not hold back it's potential performance by using "min spec" memory speed, instead choosing the highest speed they know (through testing) it will support.
I imagine AMD (or others) could have chosen to use LP memory selection as I call BS on others saying AMD would have no choice but to rearchitecture their design to use the LP over standard power memory, seeing as the LP is likely very little changes need to be done (if any compared to ground up for an entirely different memory type)
they should have "upped" to the next speed levels however instead of 2400 baseline, 2666, 2933, 3000, 3200 as power draw difference is "negligible" with proper tuning (which MSFT likely would have made sure to do...but then again is MSFT whom pull stupid as heck all the time, so long it keeps their "buddies happy" who care about the consumers themselves)
LPDDR4/LPDDR4X is not related to DDR4. It's a upgraded LPDDR3 which is also not related to DDR3.
LPDDR family is just like GDDR family and are total different type of DRAM standard. They almost draw 0 watt when not in use. And in active ram access they do not draw less power significantly compare to DDR4.
LPDDR4 was first shipped with iPhone 6s in 2015 and it takes Intel 4 years to finally catch up. BTW this article has a intentional typo: LPDDR4 3733 on Intel is actually quad channel because each channel is half width 32bit instead of DDR4 64bit.
AMD make/design for a client what that client wants, in this case, MSFT as "well known" for making sure to get (hopefully pay much for) what they want, for only reasons that they can understand.
this case, AMD really cannot say "we are not doing that" as this would mean loss of likely into the millions (or more) vs just saying "not a problem, what would you like?"
MSFT is very well known for catering to INTC and NVDA whims (they have, still do, even if it cost everyone many things)
still they AMD and MSFT should have "made sure" to not hold back it's potential performance by using "min spec" memory speed, instead choosing the highest speed they know (through testing) it will support.
I imagine AMD (or others) could have chosen to use LP memory selection as I call BS on others saying AMD would have no choice but to rearchitecture their design to use the LP over standard power memory, seeing as the LP is likely very little changes need to be done (if any compared to ground up for an entirely different memory type)
they should have "upped" to the next speed levels however instead of 2400 baseline, 2666, 2933, 3000, 3200 as power draw difference is "negligible" with proper tuning )
LP memory uses half size channels compared to regular DDR among other things, if you haven't designed your controller for it, it won't work, and since LP memory has only been used in the mobile sector earlier there was no reason for AMD to design the controller with that in mind. Extra transistors that could be used better elsewhere.
Whoa, nice surprise for a Friday morning! Thanks for the review.
Results are about what I expected, and I guess this was the risk for AMD pushing out their previous gen APU for a high profile product. I hope they get the Zen 2 APUs into products soon.
Honestly, I have no earthly clue why MS makes most of the choices it does with its hardware products. They launched the Surface Studio with a Maxwell GPU right after Pascal came out... then waited 2 years until after Turing was out to update the machine with the Pascal GPUs they should have had at launch. For a machine that is theoretically for "creators" (though, let's be honest, is probably mostly found on the desk of the boss of whoever is actually doing the creating).
They did the same thing with the ARM-based Surface Pro X with a custom chip based on the already-year-old Qualcomm 8cx. Honestly, the bigger surprise is that the updated Surface Pro series actually uses current-gen Intel chips soon after they launched.
Bet the Studio 3 comes out right before Ampere too. I agree, sometimes it seems like they're deliberately making themselves a bit less appealing not to piss off their partners.
Here is the issue: AMD's mantra is "wait until X", but when "X" arrives, it is outclassed by the competition, or comes out after the major holiday season. The rest of the world doesnt run on AMD's schedule, and this only hurts AMD. If MS didnt want to use zen+, the consumer model would have simply gone with ice lake.
Zen 2 APUs should have come out when the rest of zen 2 did.
More like they don't have the money to hold CPU releases for GPU integration
Chiplet arch should fix this and let AMD integrate within the same cycle -- this should, in theory, be the last cycle or second-to-last cycle where AMD ships a previous-generation "APU."
Don't think it was the money for mobile Zen 2, more likely lack of available wafers from TSMC, rather supply Desktop as enthusiast and gamers will then market the product while cashing in profits from Epyc.
What is amd waiting for with zen 2 for mobile? Are they prioritizing desktop and server chips first or is there a lack of capacity at tsmc with everyone hogging the 7nm process?
I think it's a mix of everything. Renoir is being worked on actively but AMD doesnt have enough resources to do all of Ryzen AM4, EPYC, and Ryzen Mobile at the same time. The desktop and server chips were prioritized first as those have clear wins over Intel solutions. Mobile however requires a lot more work with battery life and performance optimization. And yeah, we already see chip production issues with Ryzen 16 and 12-core as it is with most of the chips going to EPYC so Zen2 mobile would just add more capacity issues.
AMD should be releasing the Zen2 laptop chips at or around CES 2020(in another three weeks). Microsoft was probably preparing the Surface Laptop a good five months ago so AMD just didn't have samples ready for Microsoft at that time.
Realistically, TSMC's 7nm process (used in Zen 2's compute chiplets) is expensive and capacity-constrained. AMD gets around this on desktop by creating a separate IO die on an older, cheaper process, but that doesn't scale well to mobile.
To get Zen 2 on mobile, they'd likely need everything - compute, IO and GPU - to be on 7nm, and the chip would be considerably larger than (desktop) Zen 2's compute chiplet.
AMD basically had a choice - for a given number of 7nm wafers, they could get a few mobile APUs or a lot more desktop CPUs (and GPUs for their 7nm options). The margins are much better doing the latter, so that's what they did.
There's also the fact that we don't know what their agreement is with Global Foundries on 14/12nm wafers - we do know that AMD's agreement with them didn't require GloFo to finish its 7nm process. AMD has a commitment to buy some number of wafers from GloFo per year for a certain length of time, and it's likely that is a factor as well.
Disagreed -- cheaper IO dies have worked on mobile for years. The "cheaper" process is literally the process Zen+ shipped with.
Problem is entirely integration -- now that Navi (a major release) has shipped, AMD can integrate the current version into chiplet form, then port those integrations to the incremented version of Navi ready to ship next year. Since the chiplets provide a substantial level of decoupling, incremental upgrades should be reasonably easy to integrate.
AMD is a smaller company with fewer resources, as you can see with various things. The step by step Navi Launch (currently only Navi 14 is additionally available), delayed 3950X launch, the fragmented Threadripper launch, not until early 2020 Renoir with Zen2, but only combined with Vega (not Navi) and there's still the question when desktop APUs will hit the market. The new suface models presumably have been in development for about a year or even longer. There was no mobile Zen2 in sight ... and this also applies to Apple's Mac Pro and stupid demands like "it should have been Threadripper, not Xeon W". There was no appealing Threadripper available (the old 2000 series still has various flaws and limits and the Mac Pro is an even more complex platform). Additionally at this point, I would assume, that Microsofts strategy (with the two AMD and ARM models) is more about keeping Intel in check according to pricing and cooperation. The ARM Pro X has obvious limits and Microsoft also knew about AMDs comparability in advance. Additionally you can assume, that there is a good reason for Microsoft to offer channel consumers to the AMD 15" model (and to limit the Ice Lake model to business users), most likely, because they have to sell x K SKUs to make this experiment (or this Intel pressurizing model?) worthwhile. Maybe with next surface updates in 2021 AMD will be more competitive.
There's a lot more money in server and retail desktop CPU's than there are in sub-$1000 mobile parts. Why would AMD send perfect good Zen 2 cores to the mobile market if they can sell them as something with better margins?
There's no way to recommend an AMD laptop yet. Intel has all the numbers both in terms of performance and in terms of the various models by different OEMs. Lot to catch up for AMD. This market is Intel's to lose and by the looks of the integrated graphics in this, seems to be a tough task for AMD.
Zen 2 based APU's (Renoir) will easily catch up to Intel's mobile CPU lineup. We can assume that graphics and storage performance will be much better since NVMe can use the PCIe 4.0 bus and AMD's iGPU was always ahead of Intel. We will know soon since AMD is scheduled to release Renoir on CES in January. Microsoft should have waited for Renoir before putting it into its Surface.
Looking at the current high power draw of PCIe 4.0 I can't see it being attractive for mobile just yet, especially as battery life is generally more important than 5GB/s storage speeds in laptops.
If you can use half the lanes I can totally see it being attractive - just as a quad-core may use less power than an octa-core. The main issue on the desktop is wanting the same number of faster lanes.
It doesn't pull the full required power at all times, it's fine for burst. Still think it doesn't really make sense though as you don't need the speed in any laptops right now as no drive can properly max it.
Anybody care to mak any predictions on how much better the rRenoir Igpu will be compared to the 2400/3400g .. I'm thinking maybe a 25% uplift but maybe more it its NAVI. However everyone seems to be foregtting they have done most of the work , in getting out the Head Canyon NUC for Intel. they can announce in Jan with general avail in 2,3,6 months ??
There was a window where Intel was still mostly shipping dual-core i5s while AMD shipped the 2500U and 2700U for around the same price. Made for an interesting situation where the AMD units were effectively light-weight desktop replacements, while the Intel units served as ultralights, all while theoretically serving the same market.
Point being, there already was a moment where AMD laptop recommendations have made sense.
Also worth noting that Intel's H-series laptops would be a great place for AMD to compete. I'm still unconvinced AMD's core design scales down that well, and afaik Ryzen Mobile still uses the desktop process nodes? So this might be an area where Intel still has a manufacturing advantage -- midrange power consumption.
Zen 2 mobile is a low priority when server and HPDT have much higher margins. AMD needs to offer something competitive when they can afford to, but their margins is key to their financial recovery. Semi-custom CPU's for consoles are low margin, but at least in that case the volumes are so high and they're gaining a potential advantage in gaming by pushing game developers to optimize for RDNA.
But it would be good for the consumer to get some serious computing power with energy efficiency in the mobile space that has otherwise been stagnant for a while.
The current Lice likes are consider lower watt models and replacements for higher end IceLake are probably in early 2020 - likely 6 and 8 core models and higher end GPU's.
" higher end IceLake are probably in early 2020 - likely 6 and 8 core models and higher end GPU's. " um yea ok Hstewart, and you have proof of this how ?? my guess. you dont and this is just more of your pro intel opinions.
Well it just base on Intel road maps with 10nm - it might not be Ice Lake and could be next version. There is no difference in this than people saying wait for Zen 3. So do you expect that Intel will be always on 14nm and current processor will only be on 10nm. Intel has expected roadmaps on 7nm and even far less, I believe and it just a feeling that this battle with chip size will be over next year. Just remember what Intel did during the Pentum 4 days with Frequency wars especial with Intel came out with I series. Until the Zen came out Intel was really look like a monopoly but in reality it is not just AMD was doing so poorly But AMD did Intel a blessing, competition helps company from being lazy and ineffective
But I just and older profession who send technology from the early days including original IBM PC with Intel 8088 cpu.
HStewart.. and yet to STILL believe the lies intel tells you.. that is why you are thought of as an intel fanboy... until intel delivers on what they claim, and actually have products out, and no, the limited products on 10nm, that max out at quad core and lower frequencies then 14, does NOT count, anything they say, should be treated as BS... going by your post.. you have NO proof of your claims... and they are in fact, just your opinions...
Absolutely this. A friend's 6600k just died, and wanted to go with AMD. I'm up to date with most computer stuff, so he asked my opinion.
He was thinking a 3400g,because APU + 3rd gen Ryzen. It was weird to tell him that AMD stupidly named the 1st gen APUs with the 2nd gen nomenclature … then continued this stupidity for no reason. All chips 1xxx are Zen, all 2xxx chips are Zen+, all 3xxx chips are ZEN 2 … well, except the APUs, which are all back a generation.
I have high hopes for the 4xxx APUs; would be nice if they fixed the naming by pushing them back to 3xxx. However, even better would be Zen 3 cores in mobile; 4xxx naming would then make sense, they could get some volume out on the 7nm+ process that already works, use ASMedia/VIAs new higher efficiency chipset slated for Zen 3, and they'd likely decidedly obliterate Intel's offerings on CPU and GPU performance (probably not on power though; Intel is doing some good stuff there). One can dream.
I disagree, AMD is winning over a lot of the big businesses with their EPYC chips. They need the cash more than they need laptop revenue at the moment and 7nm capacity is very limited. It's best to start with the highest-profit maker first.
Get the business on their side, get more money to grow their team and then they'd be able to do more stuff at the same time like Mobile, Server, Desktop, Gaming, and so on.
All this means is that we can hope to look forward to see what Surface Laptop 4 can do with AMD's Zen 2 chips when it comes out next year.
Why? So AMD can delay their most valuable products (Epyc and Threadripper) so they can attempt to make some low margin mobile sales? Ridiculous. Mobile is one area that Intel is still fairly competitive in. There's no reason to give up high margin markets where AMD has a strong advantage, for a market where AMD would be forced to compete head on with Intel 10nm for low margin sales.
It is interesting how you did the test. I understand ur playing at "Value" because its a laptop. But if its getting 100FPS why couldn't you turn up the graphics a bit to get around 60, to show what people should expect? Because at "Value" you are purposelly making the 18% IPC affect the FPS in the situation (720p is "CPU Bound" at low)...
I also noticed you used alot of "real world" names for your cpu tests, even synthetics. While the gpu synthetics you just called "gpu tests"...
... You put the Ice Lake system in the best possible light is what I"m saying. Gave it every advantage. Why?
Because it is the thirteenth. They are wildly and insanely obviously in the pocket of Intel on odd days. Come back on an even day if you want an AMD-biased test, because those are the days AMD bought.
So it's not about "Not wanting to see the truth" It appears either anandtech's truth for temps is based on light workloads where the cpu isn't stressed at all, or possibly just wrong.
And this is supposed to be a "Showdown" between Ryzen and Intel. Not a "spotlight" or "sponsored" intel piece. So the way things were presented is very biased.
I fully expect Intel's CPU to beat Ryzen, thats for sure. And graphics on higher than value setting, Ryzen scores about 5-10 fps higher. But on this review, It's intentionally skewed for Intel.
You should've just made a separate article for each if you were planning on favoring one company. And show their positives and negatives. Not focus on every possible positive of Intel and hide everything negative.
It's just not right, where people expect you to give a non-biased review to base their purchase off of.
And how it actually performs in gaming, when your not using the IPC of Ice Lake to increase performance. The guy has plenty other videos if you care to watch.
1) Anandtech posted multiple graphs, I just linked one. Please actually read the article and look at all the data before posting here.
2) Having additional data on a graph does not distort the temperatures that were measured.
3) You can't compare one laptop's temperature to another completely different laptop's temperature and have any conclusions that are valid (your link to a Dell XPS is not compariable to this article on a Microsoft Surface). This is because each laptop has completely different cooling systems, fan speeds, etc.
I didn't notice that graph, your right. It's sad that the thermal throttling is so bad on Ice Lake. Look at how fast it hits 100c and has to lower its frequency to 2.5Ghz. If it could control its temps and didnt use so much power, it would be a beast of a chip.
So for servers if scaling is pretty close, its using 5 watts per core at 2.5 GHZ. u can expect 56 Cores in a 280W TDP. At Epyc's TDP of 225W, you can fit 45 cores.
So if Ice Lake is aiming for 280W TDP and, 38? cores. You should be able to hit around 3 GHZ with TDP judging by graph of 30watts being pulled around 3GHZ. (Not counting memory controller etc.) with IPC boost the server part might be decent, albeit with alot less cores.
You will never see Ice Lake desktop product. Compare the watt graph to Ryzen's wat graph below. See the way the Ice Lake increases substantially wattage to pull boost clocks? It's probably already near the silicon limit judging by the graph. Like where they plot the voltage vs frequency, showing that the last few hundred mhz takes an insane amount of power. Ice Lake graph shows this in the 3 GHz
Cpu performance and power effiency is much more important than slightly faster gpu than ryzen has. It is not like people buy these things to play games. And the difference is too small anyway. Ice lake wins in some games
"despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP"
Could you please remind readers how long those benchmarks run? Because if it's less than 10 seconds, the TDP and better heat dissipation of desktops won't provide any benefits on single-threaded workloads and it's misleading to mention it.
In fact, I would really like a comparison of various laptop chips versus desktop chips, in scenarios that require prolonged CPU load, such as compiling the Linux kernel, rendering something, etc. Scenarios where the laptop chips will be undergo throttling and show their weakness or lack thereof compared to desktop chips. Not those 1-second benchmarks such as those that are in AIDA64.
They are pretty long -- they don't mention the runtime of the single thread tests specifically but they do mention the multi thread tests took 4-6 hours. The multi thread tests are the same as the ST ones, just running multiple (8) copies. (This is all talking about the SPEC benchmarks)
"despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP"
This statement is contradictory. "Single-Threaded" Spec, then "Massively higher tdp" 9900k is 8 cores. which reflects the TDP. Single-Thread is one core, which is completely different. If you scale up the frequency / wattage graph, a Ice Lake 3.5GHz at 8 cores would be about 100 watts. Which im pretty sure is really close to a 9900k off the top of my head. Except this is at 3.5 and 9900k is hitting 5GHz.
Shame AMD didn't have 7nm Zen 2 for mobile ready in time for this. Maybe that's when the partnership will make more sense, until then I wish the consumer 15" was also Ice Lake.
The difference is too great. This is almost an i3 to i7 comparison, while the price difference is confusingly little. Though the Intel version could be hard to acquire, whoever's interested in this model should aim for the Intel version, even through unofficial channels.
Interesting! Guess Chipzilla is (still) king of the hill in mobile; quite impressive, as this IL i7 even comes close to desktop i7 speeds. AMD really needs to get 7 nm Zen2 or Zen2+ into mobile APUs pronto - they're leaving a lot of money on the table. Now is the time, before chipzilla gets their manufacturing problems worked out!
Less impressive when you consider the amount of money and number of defective CPU's required by Intel fabs to get to this point. When you factor in how much money Intel has dumped into their 10nm process over the last decade the money they're making off of 10 nm is sad.
Interesting to me is the slow speed of the memory with the Ryzen. Ryzen does better with faster memory. To bad we can't equalize the memory to see the effect.
I would never in a million years accuse AnandTech writers of being shills.
That said, you guys didn't notice the Ice Lake CPU was being allowed to peak and sustain a solid 8W higher than the Ryzen CPU was?
Yes, even with the higher ceiling on Ice Lake, it's able to bring its idle draw down way lower than Ryzen's, so the bursty nature of garden variety on-battery web surfing is going to heavily favor Ice Lake.
But plugged in, full bore benchmarks, the Ice Lake unit is at a clear advantage for power budget and that's concerning. It's not really as fair a fight when the Ryzen is being kneecapped, and while I don't think it would change the overall results, I *do* think the CPU tests wouldn't be the bloodbath that they are, and the GPU tests (and especially gaming tests) would be a clearer cut win for AMD.
I mentioned this in the text but the data logging tool is monitoring CPU core power on Ryzen and SoC package power on Core so the data isn't 100% comparable just due to what each company exposes to be monitored. If you look at the CPU temperatures on the Cinebench R20 test as an example the CPU temperatures are very similar so I don't think the power budget is as far off as you'd think, although Intel is pretty aggressive with boost.
Zen+ APU is also a 216mm die vs Intel's 125mm die. So it has double the surface area, which should mean almost double the heat dissipation (maybe less). So Intel's numbers being twice as high in power draw actually makes sense when you consider the area that is dissipating the heat.
That came out wrong re-reading it. sorry its late. AMD's die is twice the size which should let it cool better. and apu ccx is 60mm squared, intel's for 4 cores is almost the same on ice lake. So yes a similar power for similar heat would make sense. We don't know what's coolin the inside on both of these either though, that could make a big difference. However, I still think it's questionable when Intel boosts and hits 100 degrees C, just looks like they're pushing the CPU as far as possible to make it look as good as it can, knowing the new APU's are coming out soon.
And yes, the fact that it can draw 42 watts on CPU at boost definitely seems like an unfair advantage. The temps also reflect that in 25 seconds it overheats due to it too, forcing to ramp down.
And I looked into the IO. It should pull 10-15W, including graphics idle, putting Ryzen on par with Ice Lake around 20-25 watts. Which would make more sense.
Personal opinion. The review is very obviously biased towards Intel. The tests are carefully sellected to demonstrate Intel dominance, which in the real world is non present anymore. The only advantage Intel still has over AMD in the mobile space, is power consumption. As we see, that advantage is shrinking fast.
The results are no surprise. It’s Intel‘s newest architecture and manufacturing node against AMD‘s year old Zen1+ on 12nm. It’s too bad that AMD take their time updating mobile parts and APUs. A Zen2 + Navi on 7nm should be much more competitive. This is the inverse situation we see in desktop parts where there is no sunny cove and still 14nm for Intel
Lets see what AMD brings to the table with zen 2. Intel does have a better ipc with ice lake, even compared to zen 2, but it has a handicap in the fact that it is built on 10nm and they have only 4 cores in 15w tdp. I am pretty confident that AMD can afford a 6 core in the same tdp, like comet lake.
The boost behavior is very erratic on the AMD. I assume this ramping up/down under load is quite inefficient and buts a high emphasis on how fast the CPU can ramp up and down.
Looks very erratic to the point of being defective. Multi-core loads though are fine. Does this mean when a low-threaded task is ran the scheduler bunny hops the task on all cores? This used to be a problem on Ryzen1 and thought it was solved with an AMD driver, but it came back maybe on this MS product.
I suspect a few % performance were left on the table, but like most here I assume the GPU is bandwidth-starved and high speed memory can't come soon enough for AMD. Quite surprising how poor they are with feeding the GPU, lack of dev on the GPU side in the last few years has hurt a lot.
Awesome result for the Ice Lake chip. But let's keep expectations in check here; the Ryzen part is based on a uArch from 2017 with slight optimisations, and also has half the L3 cache size due to a significant process disadvantage. While a very valid product v product comparison it is not particularly 'fair' as Intel's development (sunk) costs for the ICL and 10nm node are astronomically higher than Picasso's.
I consider this 'flavour' until AMD can finally pull its finger out of its rear and get Zen2 mobile cores on 7nm, which should match or beat this. Finally, the memory limitation on the Picasso part is severely holding it back. I have no idea they wouldn't have opted for DDR4-2933 (what Picasso supports natively), which can lower its clock rate in less demanding tasks for higher efficiency. In some of these tests, this could have allowed AMD the win (which would have been particularly impressive given the huge aforementioned uArch/Node advantages Intel has). The bandwidth would have especially helped in the graphics tests. If the AMD surface design is significantly cheaper (I am willing to bet AMD can give these chips for 25% of what Intel is selling the ICL for) it could slot nicely in as a lower performing, but also much lower cost alternative.
I will wait patiently for Renoir to see what 7nm Zen2 can do in mobile; arguablly where it is most needed.
Then why are there results from desktop processors ESPECIALLY on the SINGLE threaded performance, but no results on the multi threaded performance?
Maybe to advertise the single threaded performance that Ice Lake offers, compared to Zen 2? Maybe because 16 Zen 2 cores obliterate Intel chips on multithreaded tests?
The fact that you avoid in the article to explain why you only post single threaded, but no multithreaded results from desktop processors explains much. Hides nothing.
One "Tom's Hardware" is enough. We don't need another.
Because it's relevant on single-threaded but irrelevant on multi-threaded. If you're after desktop CPU reviews we have those as well. If we were going to run SPEC multi-thread on a desktop CPU we'd not be running it at Rate 8. Rate 8 was run on these two processors because they have 8 threads.
I think 7nm Ryzen mobile will be a lot more competitive than Picasso. Which is 'ok' for a last generation 12nm effort, but unlikely to hold a candle to a 7nm Zen2 APU. So next year should be a really interesting one as far as notebooks are concerned.
Looking at the price, spec and performance, Intel is still the winner in the laptop areana. Why would anyone save 100 dollars for a laptop that is win home, slower and less battery life?
Two additional comments: 1. The Ryzen version is overpriced; as is, it would still be an attractive option if the price would be $300 or so less. Right now, the value proposition is on the side of the Ice Lake i7 version. 2. Many comments here complain about how this comparison is unfair to AMD's APU. Why not complain about AMD not shipping a mobile APU based on Zen2 in 7 nm, preferably with Navi graphics on board, more L3 cache, and LPDDR4 support? Isn't that the likely reason for Ice Lake taking the cake here, and trouncing AMD?
As is, I view the current Ryzen mobile APUs as good choices for mid-level machines suitable for occasional e-sports gaming, not for running a premium ultraportable $ 2000+ laptop.
"2. Many comments here complain about how this comparison is unfair to AMD's APU. Why not complain about AMD not shipping a mobile APU based on Zen2 in 7 nm, preferably with Navi graphics on board, more L3 cache, and LPDDR4 support? Isn't that the likely reason for Ice Lake taking the cake here, and trouncing AMD?"
Because all Zen 2 production is being channeled into HIGH profit server and HEDT CPU sales. While Intel fans get a moral victory in the Surface 3 review, their 10nm CPU parts are making Intel very little money using a process node they have burned billions trying to get working. AMD doesn't have the convenience of diverting valuable 7nm parts to low margin mobile sales. Congrats to Intel, they've won a much needed but very hollow victory.
It is as fair as comparing 5 year old Skylake with latest Zen2 desktop. The only difference is Skylake actually beats Zen2 if we compare “equal core count” parts. But here Zen+ is simply getting slaughtered by Ice Lake.
I thought the article was very fair, given what is available from AMD and Intel in the mobile space.
While these benchmarks show convincingly in favor of Intel, I was wondering if the performance difference was noticeable in "normal" user. Seems like the battery life is likely to be the most significant difference between the two (that the user would notice).
Have the authors been able to use these systems day to day to see what the responsiveness is like, whether you notice a difference loading office software, doing presentations, running excel, chrome, safari, etc?
I have a Dell mobile workstation for work, and just started testing a desktop workstation, and while I'm sure the desktop is faster in tests like these, I honestly don't notice a performance difference at all, except in the heaviest CAD work. Therefore I am curious of a user could really notice a difference between these two systems in typical use.
@Samus: It is sad that some people would buy the inferior Ryzen 15inch without even knowing that there is a better Intel 15inch version available and you don’t have be a business person to buy it. Pathetic anti-consumer attempt from Microsoft.
I guess for some people, the extra cpu processing power is not needed however the extra storage can be a handy thing? processors are quite powerful these days. Of course for me I would choose higher processing speed over storage, if it were for my mother i would choose storage over speed.
The business model with Intel is $100 more but comes with Windows 10 Pro which is a $100 upgrade over home, so for the hardware, it's the same money, but the business version is overall $100 more expensive.
@Brett: Agreed. In addition it comes with a better and more expensive ram. so the intel version is not just insanely faster, it is also the better value overall.
Considering Microsoft is just giving Windows Home in AMD offer and Pro + better ram on Intel one, they must be getting the intel cpu at a tremendously affordable price or up pricing the AMD offer considerably.
The pricing borders on the ridiculous for the AMD version of the SL3, on top of the Ice Lake variant not being available for consumers. I'm all for competition in the mobile chip arena but Ryzen Picasso is uncompetitive at $2000. Microsoft really made a mess this time. But then again, the 15" SL3 is a niche of a niche product, so they could be using this as an experiment for AMD on the next Surface Pro.
The only reason Microsoft is using leftover Zen Picasso parts is because Intel's 10nm is so pathetic that they can't get enough working Intel parts to supply demand. So Microsoft got a cheap deal on old Zen+ parts to help fill the void. Low margin mobile CPU markets are not what AMD is focused on. AMD needs money and the profits are in the server market and HEDT market.
"On purely CPU based tasks, Ice Lake really stretched its legs, and despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP.
Sunny Cove cores on Desktop are going to very good if Intel can get the darn things out.
Maybe somebody at AT or here knows, but my favorite suspicion why MS even had a Ryzen mobile system on offer despite it being a full generation behind is that they didn't trust Intel getting their 10 nm Ice Lake fabbing straightened out in time, and didn't want to risk getting caught without pants. In other words, was this MS playing it safe with the AMD APU as second option just in case Intel came up short again? What's the word?
Intel HASN'T gotten their 10nm Ice Lake "straightened out". Hence why Intel is suddenly mum on Ice Lake server parts. They've totally disappeared from presentations. Why anybody thought that Intel would get ICL "straightened out" now, out of the blue, is beyond me. 10nm is essentially dead.
As long as the Ryzen U chips don't get competetive idle power consumption, they will have a severe handicap in the laptop market. But for small fanless barebones systems (e.g., something like the Zotac ZBOX CI series), they should be quite appropriate. I wonder why we don't see them in that capacity.
10th Gen Comet Lake CPUs are better than Ryzen 3k APUs. Of course Ice Lake is going to dominate. Zen 3 better be decent to compete with Icelake on desktop.
Agreed. Even the outlook for 2020 server ICL has vanished from Intel's presentations. If they can't get working server parts out, there definitely won't be a 10nm desktop part in 2020.
That is why I am more upset with AMD not putting their best foot forward on mobile. A mobile APU based on Zen2 in 7 nm with Navi-based graphics and an LPDDR4 or LPDDR5 capable memory controller would have given Intel's Ice Lake a real challenge. Instead, AMD is a full generation behind. Ironic, as it's the other way around for desktops.
Why would a cash-strapped AMD "put their best foot forward" in the mobile segment. It's the lowest profit market of the three. They need profit margins and server/HEDT are FAR more profitable than mobile.
Your argument is irrelevant. Both are using the best possible RAM they supported. It is AMD’s fault to recycle old chip with outdated spec. We are comparing mobile platform (laptop), not desktop. Everything is highly integrated with CPU influences GPU performance. As both laptops are house in the same cooling solution it is a fair comparison to see which platform is faster in various task.
$2000, you've got to be kidding me! Why would anyone waste there money on these? I can get a $1000 desktop then can blow these away, and get a couple hundred dollar tablet to do the tablet stuff. Heck, I could probably get both for under $1000 if I shop around and am smart with my choices. People waste FAR too much money on unnecessary tech.
They didn't "drop the ball". AMD is still financially weak. They can't afford to waste 7nm on mobile parts, which have low profit margins. All their focus are on the highly profitable server market and very profitable high end desktop market. Hence why every new Zen release is strongly focus on SERVER and HEDT. After all the money Intel has thrown away trying to get 10nm to work, they're making very little money selling 10nm parts for low margin laptops.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
174 Comments
Back to Article
kgardas - Friday, December 13, 2019 - link
Thanks for the great article! Intel surprised me a bit since this Ice Lake wonder looks like Zen2 or even better. Now, the only question is when we can get Ice Lake cores in Xeon E/W form?UglyFrank - Friday, December 13, 2019 - link
Coffee Lake & Zen+ were very close in IPC terms & Ice Lake & Zen 2 both had similar IPC increases from their predecessorsphilehidiot - Sunday, December 15, 2019 - link
We all knew what the CPU result was going to be, and really this is a great foot in the door for AMD for when they have a competitive mobile product. Given the relentless execution after execution, I'm hopeful for something good here. I suspect they've left laptops until last as that's where a lot of Intel's dirty tactics are in play and will be the hardest market to crack. Working with MS will undoubtedly give AMD some ideas of where they can optimise for Windows on their end and also get MS to optimise on theirs. Given that AMD are actually competitive now, I suspect Microsoft are more interested in dealing with them.The surprise for me was the GPU leap from Intel. I was expecting AMD's GPU option to be far better and Intel's to just be the usual perfunctory stuff, good only for multimedia and maybe Minecraft. That.... could be a sign of things to come, methinks.
Lolimaster - Monday, December 16, 2019 - link
Vega11 is the same as the Ryzen 2400G. Zen2+ mobile will be a major upshift.philehidiot - Tuesday, December 17, 2019 - link
I really hope you're right. It's great to finally see competition. I am actually considering upgrading my main PC for the first time in years. What I don't want is Intel abusing its market position to cane AMD in the GPU market with an inferior product.YB1064 - Tuesday, December 17, 2019 - link
I did not expect such a big improvement in IPC from Intel either. Looks like they were holding back. A desktop equivalent is going to snatch the CPU crown back from AMD. I just hope the platform price is not insane.cheshirster - Wednesday, December 18, 2019 - link
Problem is that 10nm can't hit high enough clocks.They are now backporting it to 14nm, and the result will have a questionable ... everithing other than performance in games.
eva02langley - Friday, December 13, 2019 - link
Really interesting. It just proves that AMD needs to upper their game in laptops, still it is jsut telling me that I should wait for Zen 2 mobile at this point.At least, i know what I will buy in the future.
Targon - Friday, December 13, 2019 - link
7nm AMD laptop chips are expected to show up in January at CES, potentially even shipping at that point. With the 15% IPC boost, that will help a lot. If Zen3 brings another 15% IPC boost(something that has been rumored), then in January 2021 we might see AMD beating Intel in the laptop space(since Intel doesn't improve IPC all that often).RSAUser - Friday, December 13, 2019 - link
I think it will probably be higher once Zen3 ships as node will either have had a process improvement or shrunk to 5nm.Current AMD mobile is pretty bad, hope they get competitive and bring laptop prices down.
Cliff34 - Saturday, December 14, 2019 - link
I agree. AMD won't be able to compete w Intel until AMD focuses on building cpus for laptops.generalako - Monday, December 16, 2019 - link
Ehhh, Sunny Cove's successor will bring an equal IPC increase, from what Intel has stated.AMD has themselves to blame. Why tf would you delay the architecture of both GPU and CPU and process node as well like that? I mean, the money lost from doing that, as OEMs will have little reason to move away from Intel's superior products, outweighs the money "saved" from this delay, no? Not like a Zen 2 APU won't come anyway, so why not do it before rather than after?
If AMD are smart they'll jump one architecture ahead. Starting with Zen 4 in 1.5 year's time, they should be smart enough to jump straight from Zen 2 APU to Zen 4.
Korguz - Monday, December 16, 2019 - link
generalako " Ehhh, Sunny Cove's successor will bring an equal IPC increase, from what Intel has stated. " and you believe intel ???cheshirster - Tuesday, December 31, 2019 - link
"Sunny Cove's successor will bring an equal IPC increase, from what Intel has stated"They never stated that.
Qasar - Wednesday, January 1, 2020 - link
heh... yea right.. until its proven, just another lie from intel to keep their investors and shareholders happy....azazel1024 - Monday, December 16, 2019 - link
I don't think it is just the IPC boost. The two chips in the comparison here had memory bandwidth differences of 80%!Simply looking at a few of the SPEC INT tests, that was very clear that if AMD even on Zen had similar memory bandwidth to the Ice Lake chip, it likely would have been spitting distance. GPU workloads are also very heavily memory bandwidth constrained and AMD's Vega here was slightly ahead of Intel. If it had 80% higher memory bandwidth it probably would have been 20-30% faster in many of those games/benchmarks.
Even some of the not heavily memory constrained workloads like Handbrake, faster memory does improve performance. That 80% memory bandwidth difference may well have been a 5-8% performance hit to AMD.
BUT AMD shipped it with only DDR4-2400 support. It is the chip they brought to the fight.
If their Zen 2 manages to both have lower platform power (doesn't need to be parity with Intel or better, but 10% better would go a huge long way towards making it less of a decision for a lot of people), brings its 15% better IPC and if it keeps its clock speeds AND manages to bring DDR4L-3000/DDR4x-3760 or whatever compatibility and suspect Intel is done for in the laptop space.
That should give AMD several more wins in CPU performance, bring it to parity or near parity in most of the others leaving Intel with only a few wins in that. For GPU performance, AMD wouldn't even really need to update the iGPU. Just give it that extra bandwidth and Zen 2 CPU behind it and it likely would be kicking Intel's butt by 20-40%. Upgraded on top of having that, yes please.
qap - Saturday, December 14, 2019 - link
I don't think it is about IPC or CPU power anymore. Yes, AMD is slower, but not in a way that would bother me. It is more about power management that is and always was sore spot of AMD in notebooks.And no - 7nm is not the solution unfortunately. It may help under load, but under standby it can actually hurt battery life (leakage is higher). Most improvements in battery life are done by architectural changes (more specialized units, powering down unused parts etc).
nico_mach - Tuesday, December 17, 2019 - link
Well, this driver is still under development and in the first test, they were running cooler than Intel, so likely there is tuning to be done, still. They just haven't done much laptop work in recent years. They're going to be behind on more than the hardware, which is why Microsoft's involvement is really promising, really. Of course they still have some minor hardware matters to sort out, as you point out.Rezurecta - Friday, December 13, 2019 - link
Do you think a lot of help came from the sizeable memory speed differences? Do you think it benefitted cpu or gpu more? Is there a way to underclock the Intel memory so you can see the differences that the memory brings?Thank you.
ikjadoon - Friday, December 13, 2019 - link
It's perplexing: AMD created an entirely new SKU for Microsoft (the AMD RYZEN™ 7 3780U Microsoft Surface® Edition Processor), but AMD still left this CU 11-equipped iGPU at the barebones 2400 MHz speed.Why not rate this MS-only SKU at 2933MHz or 3200MHz? My only thought: high-speed DDR4 was just going to exacerbate AMD's already-too-large power consumption.
TheinsanegamerN - Friday, December 13, 2019 - link
It isnt just speed, the intel chip uses LPDDR4X. That's an entirely different beat from LPDDR4, let alone normal DDR4.AMD would need to redesign their memory controller, and they have just...not done it. The writing was on the wall, and I have no idea why AMD didnt put LPDDR4X compatibility in their chips, hell I dont know why intel waited so long. The sheer voltage difference makes a huge impact in the mobile space.
You are correct, pushing those speeds at normal DDR4 voltage levels would have tanked battery life.
ikjadoon - Friday, December 13, 2019 - link
Sigh, it is just speed. DDR4-2400 to DDR4-3200 is simply speed: there is no "entirely new controller" needed. The Zen+ desktop counterpart is rated between DDR4-2666 to 2933.LPDDR4X is almost identical to LPDDR4: "LPDDR4X is identical to LPDDR4 except additional power is saved by reducing the I/O voltage (Vddq) to 0.6 V from 1.1 V." Whoever confused you that LPDDR4X is "an entirely different beat" from LPDDR4 is talking out of their ass and I caution you to believe anything else they ever say.
And, no: DDR4-3200 vs DDR4-2400 would've tanked battery life, but simply made it somewhat worse. DDR4-3200 can still run on the stock 1.2V that SO-DIMM DDR4 relies on, but it's pricier and you'd still pay the MHz power penalty.
I don't think RAM speed/voltage has ever "tanked" a laptop's battery life: shaking my head here...
mczak - Friday, December 13, 2019 - link
I'm quite sure you're wrong here. The problem isn't the memory itself (as long as you get default 1.2V modules, which exist up to ddr4-3200 itself), but the cpu. Zen(+) cpus require higher SoC voltage for higher memory speeds (memory frequency is tied to the on-die interconnect frequency). And as far as I know, this makes quite a sizeable difference - not enough to really matter on the desktop, but enough to matter on mobile. (Although I thought Zen+ could use default SoC voltage up to ddr4-2666, but I could be wrong on that.)Byte - Friday, December 13, 2019 - link
Ryzen had huge problems with memory speed and even compatibility at launch. No doubt they had to play it safe on laptops. They should have it mostly sorted out with Zen 2 laptop, it is why the notebooks are a gen behind where as intel notebook are usually a gen ahead.ikjadoon - Saturday, December 14, 2019 - link
We both agree it would be bad for battery life and a clear AMD failure. But, the details...more errors:1. Zen+ is rated up to DDR4-2933. 3200 is a short jump. Even then, AMD couldn't even rate this custom SKU to 2666 (the bare minimum of Zen+). AMD put zero work into this custom SKU (whose only saving grace is graphics and even that was neutered). It's obviously a low-volume part (relative to what AMD sells otherwise) or such a high-profile design win.
2. If AMD can't rate (= bin) *any* of its mobile SoC batches to support even 2666MHz at normal voltages, I'd be shocked.
For any random Zen+ silicon, sure, it'd need more voltage. The whole impetus for my comments are that AMD created an entire SKU for Microsoft and seemed to take it out of oven half-baked.
Or, perhaps they had binned the GPU side too much that very few of those CU 11 units could've survived a second binning on the memory controller.
azazel1024 - Monday, December 16, 2019 - link
So all that being said, yes it had a huge impact. GPU based workloads are heavily memory speed dependent. Going from 2400 to 3200MHz likely would have seen a 10-25% increase in the various GPU benchmarks (on the lower end for those that are a bit more CPU biased). That changes AMD from being slightly better overall in GPU performance to a commanding lead.On the CPU side of things, many of the Intel wins were on workloads with a lot of memory performance needed. Going from 2400 to 3200 would probably have only resulted in the AMD chip moving up 3-5% in many workloads (20-40% in the more memory subsystem dependent SPEC INT tests), but that would have still evened the playing field a lot more.
Going to 3766 like the Intel chip would have just been even more of the same.
Zen 2 and much higher memory bandwidth can't come soon enough for AMD.
Zoolook - Saturday, December 21, 2019 - link
It's not about binning, they couldn't support that memory and keep within their desired TDP because they would have to run infinity fabric at a higher speed.They could have used faster memory and lower CPU and/or GPU speed but this is the compromise they settled on.
Dragonstongue - Friday, December 13, 2019 - link
AMD make/design for a client what that client wants, in this case, MSFT as "well known" for making sure to get (hopefully pay much for) what they want, for only reasons that they can understand.this case, AMD really cannot say "we are not doing that" as this would mean loss of likely into the millions (or more) vs just saying "not a problem, what would you like?"
MSFT is very well known for catering to INTC and NVDA whims (they have, still do, even if it cost everyone many things)
still they AMD and MSFT should have "made sure" to not hold back it's potential performance by using "min spec" memory speed, instead choosing the highest speed they know (through testing) it will support.
I imagine AMD (or others) could have chosen to use LP memory selection as I call BS on others saying AMD would have no choice but to rearchitecture their design to use the LP over standard power memory, seeing as the LP is likely very little changes need to be done (if any compared to ground up for an entirely different memory type)
they should have "upped" to the next speed levels however instead of 2400 baseline, 2666, 2933, 3000, 3200 as power draw difference is "negligible" with proper tuning (which MSFT likely would have made sure to do...but then again is MSFT whom pull stupid as heck all the time, so long it keeps their "buddies happy" who care about the consumers themselves)
mikeztm - Friday, December 13, 2019 - link
LPDDR4/LPDDR4X is not related to DDR4.It's a upgraded LPDDR3 which is also not related to DDR3.
LPDDR family is just like GDDR family and are total different type of DRAM standard.
They almost draw 0 watt when not in use. And in active ram access they do not draw less power significantly compare to DDR4.
LPDDR4 was first shipped with iPhone 6s in 2015 and it takes Intel 4 years to finally catch up.
BTW this article has a intentional typo: LPDDR4 3733 on Intel is actually quad channel because each channel is half width 32bit instead of DDR4 64bit.
Dragonstongue - Friday, December 13, 2019 - link
AMD make/design for a client what that client wants, in this case, MSFT as "well known" for making sure to get (hopefully pay much for) what they want, for only reasons that they can understand.this case, AMD really cannot say "we are not doing that" as this would mean loss of likely into the millions (or more) vs just saying "not a problem, what would you like?"
MSFT is very well known for catering to INTC and NVDA whims (they have, still do, even if it cost everyone many things)
still they AMD and MSFT should have "made sure" to not hold back it's potential performance by using "min spec" memory speed, instead choosing the highest speed they know (through testing) it will support.
I imagine AMD (or others) could have chosen to use LP memory selection as I call BS on others saying AMD would have no choice but to rearchitecture their design to use the LP over standard power memory, seeing as the LP is likely very little changes need to be done (if any compared to ground up for an entirely different memory type)
they should have "upped" to the next speed levels however instead of 2400 baseline, 2666, 2933, 3000, 3200 as power draw difference is "negligible" with proper tuning )
IMO
Zoolook - Saturday, December 21, 2019 - link
LP memory uses half size channels compared to regular DDR among other things, if you haven't designed your controller for it, it won't work, and since LP memory has only been used in the mobile sector earlier there was no reason for AMD to design the controller with that in mind.Extra transistors that could be used better elsewhere.
RSAUser - Friday, December 13, 2019 - link
Remember this is Zen gen 1, when they still had the RAM frequency scaling issue vs infinity fabric.sorten - Friday, December 13, 2019 - link
Whoa, nice surprise for a Friday morning! Thanks for the review.Results are about what I expected, and I guess this was the risk for AMD pushing out their previous gen APU for a high profile product. I hope they get the Zen 2 APUs into products soon.
coder543 - Friday, December 13, 2019 - link
Microsoft really should have waited until mobile Zen 2.sing_electric - Friday, December 13, 2019 - link
Honestly, I have no earthly clue why MS makes most of the choices it does with its hardware products. They launched the Surface Studio with a Maxwell GPU right after Pascal came out... then waited 2 years until after Turing was out to update the machine with the Pascal GPUs they should have had at launch. For a machine that is theoretically for "creators" (though, let's be honest, is probably mostly found on the desk of the boss of whoever is actually doing the creating).They did the same thing with the ARM-based Surface Pro X with a custom chip based on the already-year-old Qualcomm 8cx. Honestly, the bigger surprise is that the updated Surface Pro series actually uses current-gen Intel chips soon after they launched.
tipoo - Thursday, December 19, 2019 - link
Bet the Studio 3 comes out right before Ampere too. I agree, sometimes it seems like they're deliberately making themselves a bit less appealing not to piss off their partners.TheinsanegamerN - Friday, December 13, 2019 - link
And miss all the holiday sales?Here is the issue: AMD's mantra is "wait until X", but when "X" arrives, it is outclassed by the competition, or comes out after the major holiday season. The rest of the world doesnt run on AMD's schedule, and this only hurts AMD. If MS didnt want to use zen+, the consumer model would have simply gone with ice lake.
Zen 2 APUs should have come out when the rest of zen 2 did.
Meteor2 - Friday, December 13, 2019 - link
For sure, but they don't have the money to do the development concurrently. Something had to give.lmcd - Friday, December 13, 2019 - link
More like they don't have the money to hold CPU releases for GPU integrationChiplet arch should fix this and let AMD integrate within the same cycle -- this should, in theory, be the last cycle or second-to-last cycle where AMD ships a previous-generation "APU."
RSAUser - Friday, December 13, 2019 - link
Don't think it was the money for mobile Zen 2, more likely lack of available wafers from TSMC, rather supply Desktop as enthusiast and gamers will then market the product while cashing in profits from Epyc.Zoolook - Saturday, December 21, 2019 - link
@RSAUser I think you nailed it, it's better for their bottom line to dedicate 7nm chips to products with higher margins.shabby - Friday, December 13, 2019 - link
What is amd waiting for with zen 2 for mobile? Are they prioritizing desktop and server chips first or is there a lack of capacity at tsmc with everyone hogging the 7nm process?RBD117 - Friday, December 13, 2019 - link
I think it's a mix of everything. Renoir is being worked on actively but AMD doesnt have enough resources to do all of Ryzen AM4, EPYC, and Ryzen Mobile at the same time. The desktop and server chips were prioritized first as those have clear wins over Intel solutions. Mobile however requires a lot more work with battery life and performance optimization. And yeah, we already see chip production issues with Ryzen 16 and 12-core as it is with most of the chips going to EPYC so Zen2 mobile would just add more capacity issues.Targon - Friday, December 13, 2019 - link
AMD should be releasing the Zen2 laptop chips at or around CES 2020(in another three weeks). Microsoft was probably preparing the Surface Laptop a good five months ago so AMD just didn't have samples ready for Microsoft at that time.sing_electric - Friday, December 13, 2019 - link
Realistically, TSMC's 7nm process (used in Zen 2's compute chiplets) is expensive and capacity-constrained. AMD gets around this on desktop by creating a separate IO die on an older, cheaper process, but that doesn't scale well to mobile.To get Zen 2 on mobile, they'd likely need everything - compute, IO and GPU - to be on 7nm, and the chip would be considerably larger than (desktop) Zen 2's compute chiplet.
AMD basically had a choice - for a given number of 7nm wafers, they could get a few mobile APUs or a lot more desktop CPUs (and GPUs for their 7nm options). The margins are much better doing the latter, so that's what they did.
There's also the fact that we don't know what their agreement is with Global Foundries on 14/12nm wafers - we do know that AMD's agreement with them didn't require GloFo to finish its 7nm process. AMD has a commitment to buy some number of wafers from GloFo per year for a certain length of time, and it's likely that is a factor as well.
lmcd - Friday, December 13, 2019 - link
Disagreed -- cheaper IO dies have worked on mobile for years. The "cheaper" process is literally the process Zen+ shipped with.Problem is entirely integration -- now that Navi (a major release) has shipped, AMD can integrate the current version into chiplet form, then port those integrations to the incremented version of Navi ready to ship next year. Since the chiplets provide a substantial level of decoupling, incremental upgrades should be reasonably easy to integrate.
RSAUser - Friday, December 13, 2019 - link
Doubt the I/O die size matters in this case, doesn't scale down well so not an issue.Most likely a supply issue and getting Navi to integrate.
scineram - Monday, December 16, 2019 - link
What Navi?AnGe85 - Friday, December 13, 2019 - link
AMD is a smaller company with fewer resources, as you can see with various things. The step by step Navi Launch (currently only Navi 14 is additionally available), delayed 3950X launch, the fragmented Threadripper launch, not until early 2020 Renoir with Zen2, but only combined with Vega (not Navi) and there's still the question when desktop APUs will hit the market.The new suface models presumably have been in development for about a year or even longer. There was no mobile Zen2 in sight ... and this also applies to Apple's Mac Pro and stupid demands like "it should have been Threadripper, not Xeon W". There was no appealing Threadripper available (the old 2000 series still has various flaws and limits and the Mac Pro is an even more complex platform).
Additionally at this point, I would assume, that Microsofts strategy (with the two AMD and ARM models) is more about keeping Intel in check according to pricing and cooperation.
The ARM Pro X has obvious limits and Microsoft also knew about AMDs comparability in advance. Additionally you can assume, that there is a good reason for Microsoft to offer channel consumers to the AMD 15" model (and to limit the Ice Lake model to business users), most likely, because they have to sell x K SKUs to make this experiment (or this Intel pressurizing model?) worthwhile. Maybe with next surface updates in 2021 AMD will be more competitive.
5080 - Friday, December 13, 2019 - link
Margins. EPYC, Threadripper have a much higher margin than mobile APU's.MBarton - Monday, December 30, 2019 - link
There's a lot more money in server and retail desktop CPU's than there are in sub-$1000 mobile parts. Why would AMD send perfect good Zen 2 cores to the mobile market if they can sell them as something with better margins?Teckk - Friday, December 13, 2019 - link
There's no way to recommend an AMD laptop yet. Intel has all the numbers both in terms of performance and in terms of the various models by different OEMs. Lot to catch up for AMD. This market is Intel's to lose and by the looks of the integrated graphics in this, seems to be a tough task for AMD.5080 - Friday, December 13, 2019 - link
Zen 2 based APU's (Renoir) will easily catch up to Intel's mobile CPU lineup. We can assume that graphics and storage performance will be much better since NVMe can use the PCIe 4.0 bus and AMD's iGPU was always ahead of Intel. We will know soon since AMD is scheduled to release Renoir on CES in January. Microsoft should have waited for Renoir before putting it into its Surface.smilingcrow - Friday, December 13, 2019 - link
Looking at the current high power draw of PCIe 4.0 I can't see it being attractive for mobile just yet, especially as battery life is generally more important than 5GB/s storage speeds in laptops.GreenReaper - Friday, December 13, 2019 - link
If you can use half the lanes I can totally see it being attractive - just as a quad-core may use less power than an octa-core. The main issue on the desktop is wanting the same number of faster lanes.MBarton - Monday, December 30, 2019 - link
Exactly. AMD showing up with Zen 2 and x570's PCIe 4.0 wouldn't have carried much weight if they sacrificed lane count in the process.RSAUser - Friday, December 13, 2019 - link
It doesn't pull the full required power at all times, it's fine for burst.Still think it doesn't really make sense though as you don't need the speed in any laptops right now as no drive can properly max it.
RSAUser - Friday, December 13, 2019 - link
Storage speed is relative, most would not notice the speedup.MASSAMKULABOX - Sunday, December 15, 2019 - link
Anybody care to mak any predictions on how much better the rRenoir Igpu will be compared to the 2400/3400g .. I'm thinking maybe a 25% uplift but maybe more it its NAVI. However everyone seems to be foregtting they have done most of the work , in getting out the Head Canyon NUC for Intel. they can announce in Jan with general avail in 2,3,6 months ??scineram - Monday, December 16, 2019 - link
No.lmcd - Friday, December 13, 2019 - link
There was a window where Intel was still mostly shipping dual-core i5s while AMD shipped the 2500U and 2700U for around the same price. Made for an interesting situation where the AMD units were effectively light-weight desktop replacements, while the Intel units served as ultralights, all while theoretically serving the same market.Point being, there already was a moment where AMD laptop recommendations have made sense.
Also worth noting that Intel's H-series laptops would be a great place for AMD to compete. I'm still unconvinced AMD's core design scales down that well, and afaik Ryzen Mobile still uses the desktop process nodes? So this might be an area where Intel still has a manufacturing advantage -- midrange power consumption.
Teckk - Friday, December 13, 2019 - link
Yes. Need more wins with HP and Dell, something like XPS. Not happening till mobile Zen2 at the very least.nathanddrews - Friday, December 13, 2019 - link
Ice Lake + G7 is giving me flashbacks to the Pentium M/Yonah era buzz.strahinja78 - Friday, December 13, 2019 - link
yeah,...and that wouldn't be good for AMDGreenReaper - Friday, December 13, 2019 - link
They need Zen 2 out on mobile. It'd sort at least half of the performance issues. Better memory would help there, on power, and with graphics as well.MBarton - Monday, December 30, 2019 - link
Zen 2 mobile is a low priority when server and HPDT have much higher margins. AMD needs to offer something competitive when they can afford to, but their margins is key to their financial recovery. Semi-custom CPU's for consoles are low margin, but at least in that case the volumes are so high and they're gaining a potential advantage in gaming by pushing game developers to optimize for RDNA.nathanddrews - Friday, December 13, 2019 - link
But it would be good for the consumer to get some serious computing power with energy efficiency in the mobile space that has otherwise been stagnant for a while.HStewart - Saturday, December 14, 2019 - link
The current Lice likes are consider lower watt models and replacements for higher end IceLake are probably in early 2020 - likely 6 and 8 core models and higher end GPU's.Not intended as gaming CPU's
Korguz - Sunday, December 15, 2019 - link
" higher end IceLake are probably in early 2020 - likely 6 and 8 core models and higher end GPU's. " um yea ok Hstewart, and you have proof of this how ?? my guess. you dont and this is just more of your pro intel opinions.HStewart - Friday, December 20, 2019 - link
Well it just base on Intel road maps with 10nm - it might not be Ice Lake and could be next version. There is no difference in this than people saying wait for Zen 3. So do you expect that Intel will be always on 14nm and current processor will only be on 10nm. Intel has expected roadmaps on 7nm and even far less, I believe and it just a feeling that this battle with chip size will be over next year. Just remember what Intel did during the Pentum 4 days with Frequency wars especial with Intel came out with I series. Until the Zen came out Intel was really look like a monopoly but in reality it is not just AMD was doing so poorly But AMD did Intel a blessing, competition helps company from being lazy and ineffectiveBut I just and older profession who send technology from the early days including original IBM PC with Intel 8088 cpu.
Korguz - Friday, December 20, 2019 - link
HStewart.. and yet to STILL believe the lies intel tells you.. that is why you are thought of as an intel fanboy... until intel delivers on what they claim, and actually have products out, and no, the limited products on 10nm, that max out at quad core and lower frequencies then 14, does NOT count, anything they say, should be treated as BS...going by your post.. you have NO proof of your claims... and they are in fact, just your opinions...
MBarton - Monday, December 30, 2019 - link
At this point Intel's roadmaps are nothing more than lip service for investors. Absolutely nobody takes them seriously.djayjp - Friday, December 13, 2019 - link
AMD should've led Zen 2 (third gen Ryzen) with mobile first. Oh well maybe next generation.djayjp - Friday, December 13, 2019 - link
Also the marketing name for the chips really doesn't help, in fact I think it works against them and makes them look bad.ChubChub - Sunday, December 15, 2019 - link
Absolutely this. A friend's 6600k just died, and wanted to go with AMD. I'm up to date with most computer stuff, so he asked my opinion.He was thinking a 3400g,because APU + 3rd gen Ryzen. It was weird to tell him that AMD stupidly named the 1st gen APUs with the 2nd gen nomenclature … then continued this stupidity for no reason. All chips 1xxx are Zen, all 2xxx chips are Zen+, all 3xxx chips are ZEN 2 … well, except the APUs, which are all back a generation.
I have high hopes for the 4xxx APUs; would be nice if they fixed the naming by pushing them back to 3xxx. However, even better would be Zen 3 cores in mobile; 4xxx naming would then make sense, they could get some volume out on the 7nm+ process that already works, use ASMedia/VIAs new higher efficiency chipset slated for Zen 3, and they'd likely decidedly obliterate Intel's offerings on CPU and GPU performance (probably not on power though; Intel is doing some good stuff there). One can dream.
MikhailT - Friday, December 13, 2019 - link
I disagree, AMD is winning over a lot of the big businesses with their EPYC chips. They need the cash more than they need laptop revenue at the moment and 7nm capacity is very limited. It's best to start with the highest-profit maker first.Get the business on their side, get more money to grow their team and then they'd be able to do more stuff at the same time like Mobile, Server, Desktop, Gaming, and so on.
All this means is that we can hope to look forward to see what Surface Laptop 4 can do with AMD's Zen 2 chips when it comes out next year.
MBarton - Monday, December 30, 2019 - link
Why? So AMD can delay their most valuable products (Epyc and Threadripper) so they can attempt to make some low margin mobile sales? Ridiculous. Mobile is one area that Intel is still fairly competitive in. There's no reason to give up high margin markets where AMD has a strong advantage, for a market where AMD would be forced to compete head on with Intel 10nm for low margin sales.Fataliity - Friday, December 13, 2019 - link
It is interesting how you did the test. I understand ur playing at "Value" because its a laptop. But if its getting 100FPS why couldn't you turn up the graphics a bit to get around 60, to show what people should expect? Because at "Value" you are purposelly making the 18% IPC affect the FPS in the situation (720p is "CPU Bound" at low)...I also noticed you used alot of "real world" names for your cpu tests, even synthetics. While the gpu synthetics you just called "gpu tests"...
... You put the Ice Lake system in the best possible light is what I"m saying. Gave it every advantage. Why?
Lord of the Bored - Saturday, December 14, 2019 - link
Because it is the thirteenth. They are wildly and insanely obviously in the pocket of Intel on odd days. Come back on an even day if you want an AMD-biased test, because those are the days AMD bought.Brett Howse - Saturday, December 14, 2019 - link
Because we use the data for future articles:https://www.anandtech.com/bench/Notebook/725
If I run at random settings then the data isn't useful for the future.
Also if anything the AMD system was put in the best light because it was running at 1280x720 instead of 1366x768, as mentioned in the article.
Fataliity - Friday, December 13, 2019 - link
And I dont see a single page about temps? Is that because it runs at 90-100 C?extide - Friday, December 13, 2019 - link
There is an entire page of graphs with temps...dullard - Friday, December 13, 2019 - link
Exactly, but looking at Fatality's post above, it did not appear like Fatality wanted to see the actual truth. The CPU temps were not much different.https://images.anandtech.com/doci/15213/PCMark10.p...
Fataliity - Friday, December 13, 2019 - link
Every other website shows temps here.https://openbenchmarking.org/embed.php?i=1911040-H...
Over 90 celsius.
That single graph isnt a good comparison for temps. That's why its littered with 'frequency' and 'power'. to distort it.
Fataliity - Friday, December 13, 2019 - link
So it's not about "Not wanting to see the truth" It appears either anandtech's truth for temps is based on light workloads where the cpu isn't stressed at all, or possibly just wrong.Fataliity - Friday, December 13, 2019 - link
And this is supposed to be a "Showdown" between Ryzen and Intel. Not a "spotlight" or "sponsored" intel piece. So the way things were presented is very biased.I fully expect Intel's CPU to beat Ryzen, thats for sure. And graphics on higher than value setting, Ryzen scores about 5-10 fps higher. But on this review, It's intentionally skewed for Intel.
You should've just made a separate article for each if you were planning on favoring one company. And show their positives and negatives. Not focus on every possible positive of Intel and hide everything negative.
It's just not right, where people expect you to give a non-biased review to base their purchase off of.
Fataliity - Friday, December 13, 2019 - link
https://www.phoronix.com/scan.php?page=news_item&a...Link to article where graph is from.
Fataliity - Friday, December 13, 2019 - link
And how it actually performs in gaming, when your not using the IPC of Ice Lake to increase performance. The guy has plenty other videos if you care to watch.https://www.youtube.com/watch?v=B9FgYx2z0NY&t=...
maroon1 - Saturday, December 14, 2019 - link
PointlessThis is apu vs apu. It should be compared the way it is made
dullard - Friday, December 13, 2019 - link
1) Anandtech posted multiple graphs, I just linked one. Please actually read the article and look at all the data before posting here.2) Having additional data on a graph does not distort the temperatures that were measured.
3) You can't compare one laptop's temperature to another completely different laptop's temperature and have any conclusions that are valid (your link to a Dell XPS is not compariable to this article on a Microsoft Surface). This is because each laptop has completely different cooling systems, fan speeds, etc.
4) The temps were done doing things like multi-threadded Cinebench. If that is too light of a load for you, feel free to do your own test and post them here. https://images.anandtech.com/doci/15213/CinbenchR2...
Fataliity - Friday, December 13, 2019 - link
I didn't notice that graph, your right. It's sad that the thermal throttling is so bad on Ice Lake. Look at how fast it hits 100c and has to lower its frequency to 2.5Ghz. If it could control its temps and didnt use so much power, it would be a beast of a chip.Fataliity - Friday, December 13, 2019 - link
Over 40 Watts at boost frequency! That's insane. more than the TDP of the product.Fataliity - Friday, December 13, 2019 - link
So for servers if scaling is pretty close, its using 5 watts per core at 2.5 GHZ. u can expect 56 Cores in a 280W TDP.At Epyc's TDP of 225W, you can fit 45 cores.
So if Ice Lake is aiming for 280W TDP and, 38? cores. You should be able to hit around 3 GHZ with TDP judging by graph of 30watts being pulled around 3GHZ. (Not counting memory controller etc.) with IPC boost the server part might be decent, albeit with alot less cores.
Fataliity - Friday, December 13, 2019 - link
And Lastly, look at that graph for TDP during boost.https://images.anandtech.com/doci/15213/CinbenchR2...
You will never see Ice Lake desktop product.
Compare the watt graph to Ryzen's wat graph below.
See the way the Ice Lake increases substantially wattage to pull boost clocks? It's probably already near the silicon limit judging by the graph. Like where they plot the voltage vs frequency, showing that the last few hundred mhz takes an insane amount of power. Ice Lake graph shows this in the 3 GHz
sorry no edit button.
Fataliity - Friday, December 13, 2019 - link
25Watts at 2.7-2.9 ghz. 47 watts at 3.7ghz.Brett Howse - Saturday, December 14, 2019 - link
That's how the Power Levels work. Processor gets extended boost level 2 for a certain amount of time. Please see here:https://www.anandtech.com/show/13544/why-intel-pro...
m53 - Sunday, December 15, 2019 - link
@Brett: Why feeding the troll? Just ignore Fataliity. I wish Anandtech has some moderation in the comment section against trolling and spamming.maroon1 - Friday, December 13, 2019 - link
Cpu performance and power effiency is much more important than slightly faster gpu than ryzen has. It is not like people buy these things to play games. And the difference is too small anyway. Ice lake wins in some gamesSo intel is better out of the two
ZoZo - Friday, December 13, 2019 - link
"despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP"Could you please remind readers how long those benchmarks run? Because if it's less than 10 seconds, the TDP and better heat dissipation of desktops won't provide any benefits on single-threaded workloads and it's misleading to mention it.
In fact, I would really like a comparison of various laptop chips versus desktop chips, in scenarios that require prolonged CPU load, such as compiling the Linux kernel, rendering something, etc. Scenarios where the laptop chips will be undergo throttling and show their weakness or lack thereof compared to desktop chips. Not those 1-second benchmarks such as those that are in AIDA64.
extide - Friday, December 13, 2019 - link
They are pretty long -- they don't mention the runtime of the single thread tests specifically but they do mention the multi thread tests took 4-6 hours. The multi thread tests are the same as the ST ones, just running multiple (8) copies. (This is all talking about the SPEC benchmarks)Andrei Frumusanu - Friday, December 13, 2019 - link
Correct. Between 3-5 hours each ST and MT.Fataliity - Friday, December 13, 2019 - link
"despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP"This statement is contradictory. "Single-Threaded" Spec, then "Massively higher tdp" 9900k is 8 cores. which reflects the TDP. Single-Thread is one core, which is completely different. If you scale up the frequency / wattage graph, a Ice Lake 3.5GHz at 8 cores would be about 100 watts. Which im pretty sure is really close to a 9900k off the top of my head. Except this is at 3.5 and 9900k is hitting 5GHz.
tipoo - Friday, December 13, 2019 - link
Shame AMD didn't have 7nm Zen 2 for mobile ready in time for this. Maybe that's when the partnership will make more sense, until then I wish the consumer 15" was also Ice Lake.s.yu - Friday, December 13, 2019 - link
The difference is too great. This is almost an i3 to i7 comparison, while the price difference is confusingly little. Though the Intel version could be hard to acquire, whoever's interested in this model should aim for the Intel version, even through unofficial channels.eastcoast_pete - Friday, December 13, 2019 - link
Interesting! Guess Chipzilla is (still) king of the hill in mobile; quite impressive, as this IL i7 even comes close to desktop i7 speeds. AMD really needs to get 7 nm Zen2 or Zen2+ into mobile APUs pronto - they're leaving a lot of money on the table. Now is the time, before chipzilla gets their manufacturing problems worked out!MBarton - Monday, December 30, 2019 - link
Less impressive when you consider the amount of money and number of defective CPU's required by Intel fabs to get to this point. When you factor in how much money Intel has dumped into their 10nm process over the last decade the money they're making off of 10 nm is sad.Consumer1 - Friday, December 13, 2019 - link
Interesting to me is the slow speed of the memory with the Ryzen. Ryzen does better with faster memory. To bad we can't equalize the memory to see the effect.Dustin Sklavos - Friday, December 13, 2019 - link
I would never in a million years accuse AnandTech writers of being shills.That said, you guys didn't notice the Ice Lake CPU was being allowed to peak and sustain a solid 8W higher than the Ryzen CPU was?
Yes, even with the higher ceiling on Ice Lake, it's able to bring its idle draw down way lower than Ryzen's, so the bursty nature of garden variety on-battery web surfing is going to heavily favor Ice Lake.
But plugged in, full bore benchmarks, the Ice Lake unit is at a clear advantage for power budget and that's concerning. It's not really as fair a fight when the Ryzen is being kneecapped, and while I don't think it would change the overall results, I *do* think the CPU tests wouldn't be the bloodbath that they are, and the GPU tests (and especially gaming tests) would be a clearer cut win for AMD.
Brett Howse - Saturday, December 14, 2019 - link
I mentioned this in the text but the data logging tool is monitoring CPU core power on Ryzen and SoC package power on Core so the data isn't 100% comparable just due to what each company exposes to be monitored. If you look at the CPU temperatures on the Cinebench R20 test as an example the CPU temperatures are very similar so I don't think the power budget is as far off as you'd think, although Intel is pretty aggressive with boost.Fataliity - Saturday, December 14, 2019 - link
Zen+ APU is also a 216mm die vs Intel's 125mm die. So it has double the surface area, which should mean almost double the heat dissipation (maybe less). So Intel's numbers being twice as high in power draw actually makes sense when you consider the area that is dissipating the heat.Fataliity - Saturday, December 14, 2019 - link
That came out wrong re-reading it. sorry its late. AMD's die is twice the size which should let it cool better. and apu ccx is 60mm squared, intel's for 4 cores is almost the same on ice lake. So yes a similar power for similar heat would make sense. We don't know what's coolin the inside on both of these either though, that could make a big difference. However, I still think it's questionable when Intel boosts and hits 100 degrees C, just looks like they're pushing the CPU as far as possible to make it look as good as it can, knowing the new APU's are coming out soon.And yes, the fact that it can draw 42 watts on CPU at boost definitely seems like an unfair advantage. The temps also reflect that in 25 seconds it overheats due to it too, forcing to ramp down.
Fataliity - Saturday, December 14, 2019 - link
And I looked into the IO. It should pull 10-15W, including graphics idle, putting Ryzen on par with Ice Lake around 20-25 watts. Which would make more sense.pifaa - Saturday, December 14, 2019 - link
Personal opinion. The review is very obviously biased towards Intel. The tests are carefully sellected to demonstrate Intel dominance, which in the real world is non present anymore. The only advantage Intel still has over AMD in the mobile space, is power consumption. As we see, that advantage is shrinking fast.m53 - Saturday, December 14, 2019 - link
Personal opinion. The earth is flat.GruenSein - Saturday, December 14, 2019 - link
The results are no surprise. It’s Intel‘s newest architecture and manufacturing node against AMD‘s year old Zen1+ on 12nm. It’s too bad that AMD take their time updating mobile parts and APUs. A Zen2 + Navi on 7nm should be much more competitive. This is the inverse situation we see in desktop parts where there is no sunny cove and still 14nm for Intelyeeeeman - Saturday, December 14, 2019 - link
Lets see what AMD brings to the table with zen 2. Intel does have a better ipc with ice lake, even compared to zen 2, but it has a handicap in the fact that it is built on 10nm and they have only 4 cores in 15w tdp. I am pretty confident that AMD can afford a 6 core in the same tdp, like comet lake.dragosmp - Saturday, December 14, 2019 - link
The boost behavior is very erratic on the AMD. I assume this ramping up/down under load is quite inefficient and buts a high emphasis on how fast the CPU can ramp up and down.Looks very erratic to the point of being defective. Multi-core loads though are fine. Does this mean when a low-threaded task is ran the scheduler bunny hops the task on all cores? This used to be a problem on Ryzen1 and thought it was solved with an AMD driver, but it came back maybe on this MS product.
I suspect a few % performance were left on the table, but like most here I assume the GPU is bandwidth-starved and high speed memory can't come soon enough for AMD. Quite surprising how poor they are with feeding the GPU, lack of dev on the GPU side in the last few years has hurt a lot.
AshlayW - Saturday, December 14, 2019 - link
Awesome result for the Ice Lake chip. But let's keep expectations in check here; the Ryzen part is based on a uArch from 2017 with slight optimisations, and also has half the L3 cache size due to a significant process disadvantage. While a very valid product v product comparison it is not particularly 'fair' as Intel's development (sunk) costs for the ICL and 10nm node are astronomically higher than Picasso's.I consider this 'flavour' until AMD can finally pull its finger out of its rear and get Zen2 mobile cores on 7nm, which should match or beat this. Finally, the memory limitation on the Picasso part is severely holding it back. I have no idea they wouldn't have opted for DDR4-2933 (what Picasso supports natively), which can lower its clock rate in less demanding tasks for higher efficiency. In some of these tests, this could have allowed AMD the win (which would have been particularly impressive given the huge aforementioned uArch/Node advantages Intel has). The bandwidth would have especially helped in the graphics tests. If the AMD surface design is significantly cheaper (I am willing to bet AMD can give these chips for 25% of what Intel is selling the ICL for) it could slot nicely in as a lower performing, but also much lower cost alternative.
I will wait patiently for Renoir to see what 7nm Zen2 can do in mobile; arguablly where it is most needed.
scineram - Monday, December 16, 2019 - link
I think they worried about power with faster memory.MBarton - Monday, December 30, 2019 - link
profit margin limited and not enough leftover Zen 2 after satisfying Epyc orders, THR3/Ryzen 3k, and console orders.
yannigr2 - Saturday, December 14, 2019 - link
Nice. Let's run the single threaded SPEC2017 test to show how good Intel is against the new star, the Ryzen 9 3950.Let's run the multi threaded SPEC2017 test WITHOUT the desktop CPUs to avoid showing how bad Intel is against the Ryzen 9 3950.
yannigr2 - Saturday, December 14, 2019 - link
Thats way we can title both results as "Intel having an advantage"Where is the good old Anandtech?
Brett Howse - Saturday, December 14, 2019 - link
The comparison is 15W AMD to 15W Intel in the same laptop. Please refer to the title of the article.yannigr2 - Sunday, December 15, 2019 - link
Then why are there results from desktop processors ESPECIALLY on the SINGLE threaded performance, but no results on the multi threaded performance?Maybe to advertise the single threaded performance that Ice Lake offers, compared to Zen 2? Maybe because 16 Zen 2 cores obliterate Intel chips on multithreaded tests?
The fact that you avoid in the article to explain why you only post single threaded, but no multithreaded results from desktop processors explains much. Hides nothing.
One "Tom's Hardware" is enough. We don't need another.
Brett Howse - Sunday, December 15, 2019 - link
Because it's relevant on single-threaded but irrelevant on multi-threaded. If you're after desktop CPU reviews we have those as well. If we were going to run SPEC multi-thread on a desktop CPU we'd not be running it at Rate 8. Rate 8 was run on these two processors because they have 8 threads.Maxiking - Saturday, December 14, 2019 - link
Lol, amd trash brigading again.Haawser - Saturday, December 14, 2019 - link
I think 7nm Ryzen mobile will be a lot more competitive than Picasso. Which is 'ok' for a last generation 12nm effort, but unlikely to hold a candle to a 7nm Zen2 APU. So next year should be a really interesting one as far as notebooks are concerned.Cliff34 - Saturday, December 14, 2019 - link
Looking at the price, spec and performance, Intel is still the winner in the laptop areana. Why would anyone save 100 dollars for a laptop that is win home, slower and less battery life?eastcoast_pete - Saturday, December 14, 2019 - link
Two additional comments:1. The Ryzen version is overpriced; as is, it would still be an attractive option if the price would be $300 or so less. Right now, the value proposition is on the side of the Ice Lake i7 version.
2. Many comments here complain about how this comparison is unfair to AMD's APU. Why not complain about AMD not shipping a mobile APU based on Zen2 in 7 nm, preferably with Navi graphics on board, more L3 cache, and LPDDR4 support? Isn't that the likely reason for Ice Lake taking the cake here, and trouncing AMD?
As is, I view the current Ryzen mobile APUs as good choices for mid-level machines suitable for occasional e-sports gaming, not for running a premium ultraportable $ 2000+ laptop.
MBarton - Monday, December 30, 2019 - link
"2. Many comments here complain about how this comparison is unfair to AMD's APU. Why not complain about AMD not shipping a mobile APU based on Zen2 in 7 nm, preferably with Navi graphics on board, more L3 cache, and LPDDR4 support? Isn't that the likely reason for Ice Lake taking the cake here, and trouncing AMD?"Because all Zen 2 production is being channeled into HIGH profit server and HEDT CPU sales. While Intel fans get a moral victory in the Surface 3 review, their 10nm CPU parts are making Intel very little money using a process node they have burned billions trying to get working. AMD doesn't have the convenience of diverting valuable 7nm parts to low margin mobile sales. Congrats to Intel, they've won a much needed but very hollow victory.
maroon1 - Sunday, January 12, 2020 - link
You mean LPDDR4XThere is big difference between LPDDR4 and LPDDR4X in speed
Llawehtdliub - Saturday, December 14, 2019 - link
Latest Intel CPU vs AMD's year old CPU.Doesn't seem very fair.
Brett Howse - Saturday, December 14, 2019 - link
And yet this is what is available as current gen products.m53 - Saturday, December 14, 2019 - link
It is as fair as comparing 5 year old Skylake with latest Zen2 desktop. The only difference is Skylake actually beats Zen2 if we compare “equal core count” parts. But here Zen+ is simply getting slaughtered by Ice Lake.0ldman79 - Tuesday, December 17, 2019 - link
This is Ice Lake, not Skylake.Not a 5 year old design.
This is, however, the two products that Microsoft has available to choose from.
Korguz - Tuesday, December 17, 2019 - link
m53. i doubt skylake can beat zen2, clock for clock, or core for core....0ldman79 but is based on a 5 year old design, so basically the same thing, hence why is 10th generation, not 1st generation
MBarton - Monday, December 30, 2019 - link
Wow, such a fanboy statement from a guy that whined about that fatality guy being a fanboy.maroon1 - Tuesday, December 17, 2019 - link
WRONGRyzen 7 3780U came out this year.
It is based on 12nm Zen+ but AMD low power mobile is almost 1 year behind their desktop parts
Zen2 APU will be out in 2020, but intel will also have Tiger lake-U in 2020.
AnnoyedGrunt - Saturday, December 14, 2019 - link
I thought the article was very fair, given what is available from AMD and Intel in the mobile space.While these benchmarks show convincingly in favor of Intel, I was wondering if the performance difference was noticeable in "normal" user. Seems like the battery life is likely to be the most significant difference between the two (that the user would notice).
Have the authors been able to use these systems day to day to see what the responsiveness is like, whether you notice a difference loading office software, doing presentations, running excel, chrome, safari, etc?
I have a Dell mobile workstation for work, and just started testing a desktop workstation, and while I'm sure the desktop is faster in tests like these, I honestly don't notice a performance difference at all, except in the heaviest CAD work. Therefore I am curious of a user could really notice a difference between these two systems in typical use.
-AG
Samus - Saturday, December 14, 2019 - link
Ouch, well that's embarrassing.Why the hell did Microsoft even bother with a Ryzen model when its so inferior?
m53 - Saturday, December 14, 2019 - link
@Samus: It is sad that some people would buy the inferior Ryzen 15inch without even knowing that there is a better Intel 15inch version available and you don’t have be a business person to buy it. Pathetic anti-consumer attempt from Microsoft.0ldman79 - Tuesday, December 17, 2019 - link
Pretty sure it's because Intel is having issues supplying the chips to the OEMs.Microsoft, like most OEM, are hedging their bets and designing systems to run AMD chips.
MBarton - Monday, December 30, 2019 - link
It's sad that anyone would buy an overpriced Microsoft Surface.MBarton - Monday, December 30, 2019 - link
Because Intel 10nm yields are so woefully pathetic that Microsoft had to source AMD's old Zen parts to help make up for the lack of parts.Polacott - Saturday, December 14, 2019 - link
I guess for some people, the extra cpu processing power is not needed however the extra storage can be a handy thing? processors are quite powerful these days. Of course for me I would choose higher processing speed over storage, if it were for my mother i would choose storage over speed.Brett Howse - Saturday, December 14, 2019 - link
You can get the same storage in both.Polacott - Sunday, December 15, 2019 - link
For the dame money?Brett Howse - Sunday, December 15, 2019 - link
The business model with Intel is $100 more but comes with Windows 10 Pro which is a $100 upgrade over home, so for the hardware, it's the same money, but the business version is overall $100 more expensive.m53 - Sunday, December 15, 2019 - link
@Brett: Agreed. In addition it comes with a better and more expensive ram. so the intel version is not just insanely faster, it is also the better value overall.Polacott - Saturday, December 14, 2019 - link
Considering Microsoft is just giving Windows Home in AMD offer and Pro + better ram on Intel one, they must be getting the intel cpu at a tremendously affordable price or up pricing the AMD offer considerably.serendip - Saturday, December 14, 2019 - link
The pricing borders on the ridiculous for the AMD version of the SL3, on top of the Ice Lake variant not being available for consumers. I'm all for competition in the mobile chip arena but Ryzen Picasso is uncompetitive at $2000. Microsoft really made a mess this time. But then again, the 15" SL3 is a niche of a niche product, so they could be using this as an experiment for AMD on the next Surface Pro.MBarton - Monday, December 30, 2019 - link
The only reason Microsoft is using leftover Zen Picasso parts is because Intel's 10nm is so pathetic that they can't get enough working Intel parts to supply demand. So Microsoft got a cheap deal on old Zen+ parts to help fill the void. Low margin mobile CPU markets are not what AMD is focused on. AMD needs money and the profits are in the server market and HEDT market.Jhlot - Sunday, December 15, 2019 - link
"On purely CPU based tasks, Ice Lake really stretched its legs, and despite this being a 3.9 GHz chip, in single-threaded SPEC 2017, it managed to come very close to a 5.0 GHz Core i9-9900K with a massively higher TDP.Sunny Cove cores on Desktop are going to very good if Intel can get the darn things out.
eastcoast_pete - Sunday, December 15, 2019 - link
Maybe somebody at AT or here knows, but my favorite suspicion why MS even had a Ryzen mobile system on offer despite it being a full generation behind is that they didn't trust Intel getting their 10 nm Ice Lake fabbing straightened out in time, and didn't want to risk getting caught without pants. In other words, was this MS playing it safe with the AMD APU as second option just in case Intel came up short again? What's the word?heffeque - Sunday, December 15, 2019 - link
This sounds very plausible.MBarton - Monday, December 30, 2019 - link
Intel HASN'T gotten their 10nm Ice Lake "straightened out". Hence why Intel is suddenly mum on Ice Lake server parts. They've totally disappeared from presentations. Why anybody thought that Intel would get ICL "straightened out" now, out of the blue, is beyond me. 10nm is essentially dead.AntonErtl - Sunday, December 15, 2019 - link
As long as the Ryzen U chips don't get competetive idle power consumption, they will have a severe handicap in the laptop market. But for small fanless barebones systems (e.g., something like the Zotac ZBOX CI series), they should be quite appropriate. I wonder why we don't see them in that capacity.Farfolomew - Sunday, December 15, 2019 - link
10th Gen Comet Lake CPUs are better than Ryzen 3k APUs. Of course Ice Lake is going to dominate. Zen 3 better be decent to compete with Icelake on desktop.scineram - Monday, December 16, 2019 - link
What Icelake on desktop?maroon1 - Tuesday, December 17, 2019 - link
We won't see icelake on desktop even next yearMBarton - Monday, December 30, 2019 - link
Agreed. Even the outlook for 2020 server ICL has vanished from Intel's presentations. If they can't get working server parts out, there definitely won't be a 10nm desktop part in 2020.deil - Monday, December 16, 2019 - link
I just wonder why...16 GB Dual-Channel DDR4-2400 16 GB Dual-Channel LPDDR4X-3733
AMD gains A LOT from faster ram, what they get? cheappest....
Brett Howse - Monday, December 16, 2019 - link
That's all that Picasso supportseastcoast_pete - Wednesday, December 18, 2019 - link
That is why I am more upset with AMD not putting their best foot forward on mobile. A mobile APU based on Zen2 in 7 nm with Navi-based graphics and an LPDDR4 or LPDDR5 capable memory controller would have given Intel's Ice Lake a real challenge. Instead, AMD is a full generation behind. Ironic, as it's the other way around for desktops.MBarton - Monday, December 30, 2019 - link
Why would a cash-strapped AMD "put their best foot forward" in the mobile segment. It's the lowest profit market of the three. They need profit margins and server/HEDT are FAR more profitable than mobile.MutualCore - Monday, December 16, 2019 - link
Why nothing about SSD performance? This is a critical point for people that need to copy large amounts of data.derstef - Wednesday, December 18, 2019 - link
"DDR4-2400" vs "LPDDR4X-3733" ... fair comparision on the GPU side of things? I dont think so.Butterfish - Friday, December 20, 2019 - link
Your argument is irrelevant. Both are using the best possible RAM they supported. It is AMD’s fault to recycle old chip with outdated spec. We are comparing mobile platform (laptop), not desktop. Everything is highly integrated with CPU influences GPU performance. As both laptops are house in the same cooling solution it is a fair comparison to see which platform is faster in various task.maroon1 - Sunday, January 12, 2020 - link
AMD does not support LPDDR4XAnsley11 - Wednesday, December 18, 2019 - link
$2000, you've got to be kidding me! Why would anyone waste there money on these? I can get a $1000 desktop then can blow these away, and get a couple hundred dollar tablet to do the tablet stuff. Heck, I could probably get both for under $1000 if I shop around and am smart with my choices. People waste FAR too much money on unnecessary tech.sheh - Thursday, December 19, 2019 - link
The x264 conclusion is wrong:"x264, was also run. Here we see that once again Ice Lake has a significant performance advantage"
In fact, combining the result for both passes it's:
AMD: 11.79 FPS
Intel: 11.03 FPS
peevee - Monday, December 23, 2019 - link
Laptop CPUs needed the efficiency gains of Zen 2, Navi and 7nm the most. AMD had obviously dropped the ball here.MBarton - Monday, December 30, 2019 - link
They didn't "drop the ball". AMD is still financially weak. They can't afford to waste 7nm on mobile parts, which have low profit margins. All their focus are on the highly profitable server market and very profitable high end desktop market. Hence why every new Zen release is strongly focus on SERVER and HEDT. After all the money Intel has thrown away trying to get 10nm to work, they're making very little money selling 10nm parts for low margin laptops.Aviraj_21 - Saturday, January 4, 2020 - link
nice