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
SiliconDoc - Monday, April 6, 2009 - link
Well thanks for stomping the red rooster into the ground, definitively, after proving, once again, that what an idiot blabbering pussbag red spews about without a clue should not be swallowed with lust like a loose girl.I mean it's about time the reds just shut their stupid traps - 6 months of bs and lies will piss any decent human being off. Heck, it pissed off NVidia, and they're paid to not get angry. lol
tamalero - Sunday, April 5, 2009 - link
arggh, lots of typoos."Mirrors Edge's PhysX in other hand does show indeed add a lot of graphical feel. " should have been : Mirrors Edge's physx in other hand, does indeed show a lot of new details.
lk7600 - Friday, April 3, 2009 - link
Can you please die? Prefearbly by getting crushed to death, or by getting your face cut to shreds with a
pocketknife.
I hope that you get curb-stomped, f ucking retard
Shut the *beep* up f aggot, before you get your face bashed in and cut
to ribbons, and your throat slit.
papapapapapapapababy - Saturday, April 4, 2009 - link
Yes, i love you too, silly girl.lk7600 - Friday, April 3, 2009 - link
Can you please remove yourself from the gene pool? Preferably in the most painful and agonizing way possible? Retard
magnetar68 - Thursday, April 2, 2009 - link
Firstly, I agree with the articles basic premise that lack of convincing titles for PhysX/CUDA means this is not a weighted factor for most people.I am not most people, however, and I enjoy running NVIDIA's PhysX and CUDA SDK samples and learning how they work, so I would sacrifice some performance/quality to have access to these features (even spend a little more for them).
The main point I would like to make, however, is that I like the fact that NVIDIA is out there pushing these capabilities. Yes, until we have cross-platform OpenCL, physics and GPGPU apps will not be ubiquitous; but NVIDIA is working with developers to push these capabilities (and 3D Stereo with 3D VISION) and this is pulling the broader market to head in this direction. I think that vision/leadership is a great thing and therefore I buy NVIDIA GPUs.
I realize that ATI was pushing physics with Havok and GPGPU programming early (I think before NVIDIA), but NVIDIA has done a better job of executing on these technologies (you don't get credit for thinking about it, you get credit for doing it).
The reality is that games will be way cooler when the you extrapolate from Mirror's Edge to what will be around down the road. Without companies like NVIDIA out there making solid progress on executing on delivering these capabilities, we will never get there. That has value to me I am willing to pay a little for. Having said that, performance has to be reasonable close for this to be true.
JarredWalton - Thursday, April 2, 2009 - link
Games will be better when we get better effects, and PhysX has some potential to do that. However, the past is a clear indication that developers aren't going to fully support PhysX until it works on every mainstream card out there. Pretty much it means NVIDIA pays people to add PhysX support (either in hardware or other ways), and OpenCL is what will really create an installed user base for that sort of functionality.If you're a dev, what would you rather do: work on separate code paths for CUDA and PhysX and forget about all other GPUs, or wait for OpenCL and support all GPUs with one code path? Look at the number of "DX10.1" titles for a good indication.
josh6079 - Thursday, April 2, 2009 - link
Agreed.NVidia has certainly received credit for getting accelerated physics moving, but its momentum stops when they couple it to CUDA when offering it to discrete graphics cards outside of the GeForce family.
Hrel - Thursday, April 2, 2009 - link
Still no 3D Mark scores, STILL no low-med resolutions.Thanks for including the HD4850, where's the GTS250??? Or do you guys still not have one? Well, you could always use a 9800GTX+ instead, and actually label it correctly this time. Anyway, thanks for the review and all the info on CUDA and PhysX; pretty much just confirmed what I already knew; none of it matters until it's cross-platform.
7Enigma - Friday, April 3, 2009 - link
3DMark can be found in just about every other review. I personally don't care, but realize people compete on the Orb, and since it's just a simple benchmark to run it probably could be included without much work. The only problem I see (and agree with) is the highly optimized nature both Nvidia and ATI put on the PCVantage/3DMark benchmarks. They don't really tell you much about anything IMO. I'd argue they not only don't tell you about future games (since to my knowledge no (one?) games have ever used an engine from the benchmarks), nor do they tell you much between cards from different brands since they look for every opportunity to tweak them for the highest score, regardless of whether it has any effect in realworld performance.What low-med resolution are you asking for? 1280X1024 is the only one I'd like to see (as that's what I and probably 25-50% of all gamers are still using), but I can see why in most cases they don't test it (you have to go to low end cards to have an issue with playable framerates on anything 4850 and above at that resolution). Xbitlabs' review did include 1280X1024, but as you'll see, unless you are playing Crysis:Warhead, and to a lesser extent Farcry2 with max graphics settings and high levels of AA you are normally in the high double to triple digits in terms of framerate. Any resolution lower than that, you've got to be on integrated video to care about!