ATI Radeon HD 4890 vs. NVIDIA GeForce GTX 275
by Anand Lal Shimpi & Derek Wilson on April 2, 2009 12:00 AM EST- Posted in
- GPUs
The Widespread Support Fallacy
NVIDIA acquired Ageia, they were the guys who wanted to sell you another card to put in your system to accelerate game physics - the PPU. That idea didn’t go over too well. For starters, no one wanted another *PU in their machine. And secondly, there were no compelling titles that required it. At best we saw mediocre games with mildly interesting physics support, or decent games with uninteresting physics enhancements.
Ageia’s true strength wasn’t in its PPU chip design, many companies could do that. What Ageia did that was quite smart was it acquired an up and coming game physics API, polished it up, and gave it away for free to developers. The physics engine was called PhysX.
Developers can use PhysX, for free, in their games. There are no strings attached, no licensing fees, nothing. Now if the developer wants support, there are fees of course but it’s a great way of cutting down development costs. The physics engine in a game is responsible for all modeling of newtonian forces within the game; the engine determines how objects collide, how gravity works, etc...
If developers wanted to, they could enable PPU accelerated physics in their games and do some cool effects. Very few developers wanted to because there was no real install base of Ageia cards and Ageia wasn’t large enough to convince the major players to do anything.
PhysX, being free, was of course widely adopted. When NVIDIA purchased Ageia what they really bought was the PhysX business.
NVIDIA continued offering PhysX for free, but it killed off the PPU business. Instead, NVIDIA worked to port PhysX to CUDA so that it could run on its GPUs. The same catch 22 from before existed: developers didn’t have to include GPU accelerated physics and most don’t because they don’t like alienating their non-NVIDIA users. It’s all about hitting the largest audience and not everyone can run GPU accelerated PhysX, so most developers don’t use that aspect of the engine.
Then we have NVIDIA publishing slides like this:
Indeed, PhysX is one of the world’s most popular physics APIs - but that does not mean that developers choose to accelerate PhysX on the GPU. Most don’t. The next slide paints a clearer picture:
These are the biggest titles NVIDIA has with GPU accelerated PhysX support today. That’s 12 titles, three of which are big ones, most of the rest, well, I won’t go there.
A free physics API is great, and all indicators point to PhysX being liked by developers.
The next several slides in NVIDIA’s presentation go into detail about how GPU accelerated PhysX is used in these titles and how poorly ATI performs when GPU accelerated PhysX is enabled (because ATI can’t run CUDA code on its GPUs, the GPU-friendly code must run on the CPU instead).
We normally hold manufacturers accountable to their performance claims, well it was about time we did something about these other claims - shall we?
Our goal was simple: we wanted to know if GPU accelerated PhysX effects in these titles was useful. And if it was, would it be enough to make us pick a NVIDIA GPU over an ATI one if the ATI GPU was faster.
To accomplish this I had to bring in an outsider. Someone who hadn’t been subjected to the same NVIDIA marketing that Derek and I had. I wanted someone impartial.
Meet Ben:
I met Ben in middle school and we’ve been friends ever since. He’s a gamer of the truest form. He generally just wants to come over to my office and game while I work. The relationship is rarely harmful; I have access to lots of hardware (both PC and console) and games, and he likes to play them. He plays while I work and isn't very distracting (except when he's hungry).
These past few weeks I’ve been far too busy for even Ben’s quiet gaming in the office. First there were SSDs, then GDC and then this article. But when I needed someone to play a bunch of games and tell me if he noticed GPU accelerated PhysX, Ben was the right guy for the job.
I grabbed a Dell Studio XPS I’d been working on for a while. It’s a good little system, the first sub-$1000 Core i7 machine in fact ($799 gets you a Core i7-920 and 3GB of memory). It performs similarly to my Core i7 testbeds so if you’re looking to jump on the i7 bandwagon but don’t feel like building a machine, the Dell is an alternative.
I also setup its bigger brother, the Studio XPS 435. Personally I prefer this machine, it’s larger than the regular Studio XPS, albeit more expensive. The larger chassis makes working inside the case and upgrading the graphics card a bit more pleasant.
My machine of choice, I couldn't let Ben have the faster computer.
Both of these systems shipped with ATI graphics, obviously that wasn’t going to work. I decided to pick midrange cards to work with: a GeForce GTS 250 and a GeForce GTX 260.
Facebook
Twitter
RSS
294 Comments
View All Comments
san1s - Thursday, April 2, 2009 - link
why didn't you post a page on how useless dx 10.1 is? I bet that there will be even less difference in gameplay with dx10.1 on compared to dx10 than physxAnandThenMan - Friday, April 3, 2009 - link
10.1 can bring some meaningful performance boosts.http://img7.imageshack.us/img7/1757/hd489064.jpg">http://img7.imageshack.us/img7/1757/hd489064.jpg
SiliconDoc - Monday, April 6, 2009 - link
ati can't do physx at all - so uhh, no performance boost there, EVER.Same with cuda.
Kinda likewise with this cool Vreveal clean up video thingie.
Same with the badaboom converter compared to ati's err(non mentioned terrible implementation)...lol hush hush!!! doesn't matter ! doesn't matter ! nothing going on there !
tamalero - Thursday, April 9, 2009 - link
hu, gpu phsyx gpu aceleration only helps when theres heavy physx caltulations.almost no game uses that heavy calculations nowadays.
besides, if you wanted physix to run, dont you need a second card to run physX while the other does the graphics?
I suspect thers a slowdown as well if the same graphic card does the work.
SiliconDoc - Friday, April 24, 2009 - link
Since you suspect there's a slowdown with PhysX enabled it points out two things to me : 1. you have no clue if there is because you don't have an nvidia card, indicating your red rooster issue.2. That's why you didn't get my point when the other poster linked to the other review and listed the various settings and I laughed while pointing out the NV said PhysX enabled.
_______
It's funny how your brain farts at just the wrong time, and you expose your massive experience weakness:
"you suspect" - you don't know.
Go whine at someone else, or don't at all. At least bring a peashooter to the gunfight.
Ever played War Monger or Mirror's Edge ? lol
No of course not! YOU CAN'T, until, you know.
yacoub - Thursday, April 2, 2009 - link
Given that you label the price for the GTX260-216 as $205, and in reality it's closer to $175, can we expect the 275 will be closer to $215 in short order?yacoub - Thursday, April 2, 2009 - link
Of course ATi has a hard launch of this product - its hardware appears from your description to be identical to existing hardware just with a slight clockspeed boost, where as the GTX-275 is actually a break between the 285 and the 260.Also the 275 is much more appealing given that it has actual hardware improvements over the 260 for just a bit more cash.
chrnochime - Thursday, April 2, 2009 - link
Did you actually bothered to read other comments wrt the fact that the RV790 is a respin, not just bump in gpu/mem clkspeed??And the actual hardware improvements that are just cut-down from the GTX295. Big deal. Appealing to the NV fanboys, sure.
7Enigma - Friday, April 3, 2009 - link
It's possible he saw the article before Anand/Gary corrected the table. Originally the table had identical numbers between the 4870 and 4890. And in the text they mentioned that it was basically an OC'd 4870. This could lead one to assume it was pretty much the same card. In the article (on the next page) they did mention quickly the high transistor count, but it was brushed over quickly and they didn't really go into detail about the differences (still waiting to get a response about the cooling solution changes).As for the rest of his post, he IS clearly an Nvidia fanboy, because the 4890 is clearly the better product in just about every case (not even looking at the OC'ing potential which seems to be very nice).
SiliconDoc - Monday, April 6, 2009 - link
Hmm, better in every case, without the overclocking potential ? lol He's a fanboy ?Is it better with Cuda, curing cancer with Folding, Vreveal clean up video recoding, forced game profiles, dual GPU game forcing ? Any of those have an as good equivalent ? NO NOT ONE.
So you know dude, he's not the fanboy...
The thing is, ati did a good job with the ring around the gpu 3m transistors to cut down frazzly electric - and gain a good overclock.
That they did well. They also added capacitors and voltage management to the card - an expense left - not mentioned in expense terms - including the larger die cost.
So, on the one hand we have a rebranded overclock that merely used the same type of core reworking that goes into a shrink, but optimized for clocks with a transistor band around the outside.
Not a core rework, but a very good refinement.
I knew the intricacies would be wailed about by the red fans, but not a one is going to note that the G80 is NOT the same as the G92b - the refinements happened there as well in the die shrink, and in between, just like they do on cpu revisions.
Since ATI was making and overclocking upgrade, they needed to ring the core - and make whatever rearrangements were neccessary to do that.
Purely rebrand ? Ahh, not really, but downclocking it to the old numbers may (likely) reveal it's identical anyway.
At that point rebrand is tough to get away from, since the nvidia rebrands offered core revision and memory/clock differences, as well.
I'll give ati a very slight edge because of the ring capacitors, which is interesting, and may be due to the ddr5, that made their core viable for competition to begin with, instead of just a 9800X equivalent, the 4850 - minus the extra capabilities - cuda, better folding, physx, forced sli, game profiles - etc... vreveal... and on and on - evga game drivers on release day - etc. - oh the uhh.. ambient occlusion and fear + many other game mods for it...
Anyway, tell me none of that matters with a straight face - and that face will be so red you'll have to pay in wampum at the puter store.