Intel's 90nm Pentium M 755: Dothan Investigated
by Anand Lal Shimpi on July 21, 2004 12:05 AM EST- Posted in
- CPUs
Final Words
Banias was revolutionary; it changed the shape of the mobile CPU world and it changed how we perceive mobile chip design to be pursued. It would be nice if the same mobile-oriented design methodology spread to companies like ATI and NVIDIA. Maybe then we would truly be able to game while on the go.Dothan, as an update to Banias, leverages Intel's 90nm process and some architecture tweaks to not stun us, but to keep us happy with the Pentium M platform. At 2.0GHz, the Pentium M 755 is easily the best chip that money can buy for a notebook. Clock for clock, it's faster and uses less power than Banias - there's really no reason not to like it.
If you're using your notebook for office applications and even for content creation applications, there's no beating the new Pentium M. Even developers will find short compile times, competitive with that of a similarly clocked Athlon 64, possible with the Pentium M at lower power usage as well.
What the Pentium M isn't, however, is as strong of a media encoding or workstation processor as either the Athlon 64 or Pentium 4. Partially constrained by its 400MHz FSB and single channel memory interface, the Pentium M is not the successor to the Pentium 4 that many will make it out to be. In many ways, the Pentium M is much like Intel's version of the Athlon 64, sans memory controller. The short pipelined design and accurate branch predictor give the Pentium M a nice and high IPC, similar to the Athlon 64, which give it strength in areas like business applications and are also responsible for its high performance in our compile tests.
With Intel's vision for the future being centered on media encoding and content creation, the Pentium M is the last thing that Intel would want to build their future desktop CPUs around. That being said, the Pentium M is quite possibly the best solution that Intel has for the office PC. It would be nice if Intel used the Pentium M as the basis for its desktop Celeron processor, instead of just a castrated Pentium 4.
The Pentium M will continue to be an excellent mobile chip, and the next incarnation of the Pentium M should be yet another revolution as it may even receive multiple cores as a part of Intel's overall threading strategy.
On the desktop, we may see features like micro ops fusion and some of the Pentium M's power saving technologies make their way into the Pentium 4's eventual successor. Dropping Prescott and moving towards Dothan would be a step back, using each architecture for their strengths and borrowing technologies for use in another would make much more sense for the future of Intel's microprocessors.
In the more short term, we may see some motherboard manufacturers make Pentium M based motherboards for desktop usage. Shuttle, in particular, is currently working on a small form factor box based on the Pentium M. That should be out by the end of this year.
Facebook
Twitter
RSS
28 Comments
View All Comments
eBauer - Friday, November 19, 2004 - link
I'd like to see a P4 3.2-3.4 w/ dual channel DDR3200 used instead of single channel DDR2700 that was used for the P4 testbed - as a rough idea where a Dell 9100/XPS, for example, would fall in line with the Dothan.bhtooefr - Saturday, July 24, 2004 - link
Something else I noticed - where's the Banias 1.7 and the Dothan 735? It would make comparing Banias to Dothan so much easier to do that...Also, what motherboards did you use on the desktop chips?
#25: Take a look at http://cpu-museum.de/forum/viewtopic.php?t=1089. PowerLeap actually DID have something in the works, and was going to release in Quarter 1, but they cancelled it because of (well, at least this is what they said - I wouldn't be surprised if it was because Intel held them at gunpoint) LGA775 coming out (which can't really work with an adaptor) and the fact that the P-M wasn't running at 2.0GHz (I told them about Dothan, though).
Zebo - Saturday, July 24, 2004 - link
What are you talking about #4 and #17?Intels "dominace" in encoding? Those days are history my friends. If they ever were.
Anandtech's article published just before this one was proves it. Here's a quote:
"It was difficult to resist being a little sensationalist in this 939 roundup and titling the review, "Who needs 925X?" That would have been a fair title, however, since you can clearly see that all of the Socket 939/FX53 boards completely outperform Intel's top 560 on the top 925X motherboard. Even Media Encoding, the last bastion of Intel dominance, has fallen in benchmarks with our new AutoGK benchmark. "
As a matter of a fact it's always been a mismoner if you look at other suites/Codecs such as:
Codecs: xVID
Consumer programs: Ulead VideoStudio 4,5,6 and 7. Roxio videowave 4,5,6 and 7. Pinnacle studio 7.
MPEG2 Encoders: Canopus procoder, Ligos, bbMPEG.
DVD transcoding: DVD2AVI, VirtualDubMod
Freeware: VirtualDub.
Streaming: Quicktime.
Shot out to alex;)
T8000 - Thursday, July 22, 2004 - link
Since the Pentium M shares its bus protocol with the Pentium 4, I think companies like Powerleap should be more then able to provide an adaptor for desktop use.This would likely result in an "Intel unknown" detection, along with a lack of power save options, but previous experience with the "Tualatin" P3(1400) learned that this could work and even offer full performance compared to a full recognition.
Also, I think the limited memory bandwith did not hold Athlon 64 back, because I still remember seeing Athlon FX (dual channel) and 64 (sinle channel) perform almost equal for the same clockspeed. However, the Pentium 4 would benefit from more bandwith, as happened before on every FSB bump it got. So I think adding faster memory would only make a noticeable difference for the Pentium 4.
SKiller - Thursday, July 22, 2004 - link
I didn't say it would make up *all* the performance, but enough to make it competetive. For example: the P4 is essentially faster than A64 in those tasks but the A64 is still sufficiently close to be a viable altermative. Dothan doesn't have to beat the P4 in media encoding or content creation, just come close enough to be a good desktop CPU. It should be possible by increasing the FSB.Xentropy - Thursday, July 22, 2004 - link
#3 - "As you'd expect from 90nm, the leakage power has shot right up, meaning that in its low power states, the CPU is draining a great deal more power than Banias."Oh? Link to more info on this, please.
TrogdorJW - Wednesday, July 21, 2004 - link
One comment I forgot to mention: even with the talk of thermal design targets and clock speed limitations, I imagine that a Dothan CPU in a desktop motherboard with a large copper heatsink would have quite a lot of overclocking headroom. I know there was a French site that claimed to have overclocked a Dothan to 2.4 GHz, and the performance was quite impressive (if true). I would really like to see overclocking results for Dothan (and Banias) on a desktop system. With processors designed to generate 1/4 the amount of heat of a P4 3.2+, it could be interesting. Here's hoping some motherboard manufacturers will accommodate us! :)TrogdorJW - Wednesday, July 21, 2004 - link
Overall, I thought this was a great article. Those complaining about various configuration issues need to stop whining. Never once did AT actually give out any numbers for battery life or low-power performance, or claim that the Dothan was beating a desktop Athlon 64/P4. All we're looking at is what the various can do in typical *laptop* configurations. Getting a P4 and A64 laptop would make them into real laptops, but then we would have the laptop manufacturer's configuration, and likely it wouldn't be the same as what they had for the Dothan system. What we've got is three platforms running the same RAM, Hard Drive, and graphics card. Yes, it's limiting what some of the CPUs could do, but it's about as fair as you can get.Of course, if history is any indication, I'm fairly confident that Dothan is going to dominate other mobile architectures in battery life - when using the same screen, hard drive, and battery. Too bad it's so difficult to actually meet those criteria. Screens differ quite a bit, and many P4 and A64 laptops are shipping with 90+ kWhrs batteries, while Banias and Dothan laptops often get by with 60 to 70 kWhrs. Anyway, the Banias was generally the best laptop CPU before (i.e. most efficient while still providing good performance), so why shouldn't the Dothan be similar if not better? The results are hardly surprising.
I really hope that we one day see some desktop boards designed for Dothan, though. I imagine that getting the FSB up to 200 MHz quad-pumped should be possible, although even 133 or 166 would be helpful. Combine a Dothan CPU with all the other desktop accoutrements, and it would likely be a formidable gaming platform. Of course, it would really only be about the same as Athlon 64, and Intel is currently milking the Dothan/Banias line for all they can. $600+ for a processor that probably costs Intel less money to create than their P4 chips.
Again, great article, Anand. (And for the interested, Jon "Hannibal" Stokes over at Ars Technica put together as much information as I've seen anywhere about how the internals of the Banias/Dothan function. Not much detail in comparison to other CPU comparision articles he's written, but there is some additional information about what ops can benefit from the fusion technique, IIRC.)
rocketbuddha - Wednesday, July 21, 2004 - link
Acer(for sure) & Emachines(AFAIK) use a 4200 RPM HDD not a 5400 RPM HDD.The problem is that the available Athlon laptops are castrated in some level
a) Poor Graphics card(This changed with A64).
b) Lower speed HDDs.(Of course Aleinware, FNW fix with high end components pricing goes to the roof.)
c) Only 4 from a list of 32 varying modes by PowerNow!
With these either of these three there no way would a mobile A64 laptop would be in its best form for comparision.
SnakeJG - Wednesday, July 21, 2004 - link
I would also like to see this review updated with an Athlon 64 with 1 meg L2 cache. It seems a shame to compare an expensive 2 meg L2 cache Dothan with the cheaper 512KB L2 cache Athlon 64's. In addition, by keeping Athlon 64's to the slower/single channel RAM, you are making the L2 cache more important than normal, and hindering the Athlon 64.