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
CUDA - Oh there’s More
Oh I’m not done. Other than PhysX, NVIDIA is stressing CUDA as another huge feature that no other GPU maker on the world has.
For those who aren’t familiar, CUDA is a programming interface to NVIDIA hardware. Modern day GPUs are quite powerful, easily capable of churning out billions if not a trillion instructions per second when working on the right dataset. The problem is that harnessing such power is a bit difficult. NVIDIA put a lot of effort into developing an easy to use interface to the hardware and eventually it evolved into CUDA.
Now CUDA only works on certain NVIDIA GPUs and certainly won’t talk to Larrabee or anything in the ATI camp. Both Intel and ATI have their own alternatives, but let’s get back to CUDA for now.
The one area that GPU computing has had a tremendous impact already is the HPC market. The applications there lent themselves very well to GPU programming and thus we see incredible CUDA penetration there. What NVIDIA wants however is CUDA in the consumer market, and that’s a little more difficult.
The problem is that you need a compelling application and the first major one we looked at was Elemental’s Badaboom. The initial release of Badaboom fell short of the mark but over time it became a nice tool. While it’s not the encoder of choice for people looking to rip Blu-ray movies, it’s a good, fast way of getting your DVDs and other videos onto your iPod, iPhone or other portable media player. It only works on NVIDIA GPUs and is much faster than doing the same conversion on a CPU if you have a fast enough GPU.
The problem with Badaboom was that, like GPU accelerated PhysX, it only works on NVIDIA hardware and NVIDIA isn’t willing to give away NVIDIA GPUs to everyone in the world - thus we have another catch 22 scenario.
Badaboom is nice. If you have a NVIDIA GPU and you want to get DVD quality content onto your iPod, it works very well. But spending $200 - $300 on a GPU to run a single application just doesn’t seem like something most users would be willing to do. NVIDIA wants the equation to work like this:
Badaboom -> You buy a NVIDIA GPU
But the equation really works like this:
Games (or clever marketing) -> You buy a NVIDIA GPU -> You can also run Badaboom
Now if the majority of applications in the world required NVIDIA GPUs to run, then we’d be dealing in a very different environment, but that’s not reality in this dimension.
Facebook
Twitter
RSS
294 Comments
View All Comments
7Enigma - Thursday, April 2, 2009 - link
And just go and disregard everything I typed (minus the different driver versions). Xbit apparently underclocked the 4890 to stock speeds. So I have no clue how the heck their numbers are so significantly different, except they have this posted on system settings:ATI Catalyst:
Smoothvision HD: Anti-Aliasing: Use application settings/Box Filter
Catalyst A.I.: Standard
Mipmap Detail Level: High Quality
Wait for vertical refresh: Always Off
Enable Adaptive Anti-Aliasing: On/Quality
Other settings: default
Nvidia GeForce:
Texture filtering – Quality: High quality
Texture filtering – Trilinear optimization: Off
Texture filtering – Anisotropic sample optimization: Off
Vertical sync: Force off
Antialiasing - Gamma correction: On
Antialiasing - Transparency: Multisampling
Multi-display mixed-GPU acceleration: Multiple display performance mode
Set PhysX GPU acceleration: Enabled
Other settings: default
If those are set differently in Anand's review I'm sure you could get some weird results.
SiliconDoc - Monday, April 6, 2009 - link
LOL - set PhysX gpu accelleration enabled.roflmao
Yeah man, I'm gonna get me that red card... ( if you didn't detect sarcasm, forget it)
tamalero - Thursday, April 9, 2009 - link
good to know you blame everyone for "bad reading understanding"let's see
ATI Catalyst:
Smoothvision HD: Anti-Aliasing: Use application settings/Box Filter
Catalyst A.I.: Standard
Mipmap Detail Level: High Quality
Wait for vertical refresh: Always Off
Enable Adaptive Anti-Aliasing: On/Quality
Other settings: default
Nvidia GeForce:
Texture filtering – Quality: High quality
Texture filtering – Trilinear optimization: Off
you see the big "NVIDIA GEFORCE:" right below "other settings"?
that means the physX was ENABLED on the GEFORCE CARD.
you sir, are a nvidia fanboy and a big douché
SiliconDoc - Thursday, April 23, 2009 - link
More personal attacks, when YOU are the one who can't read, you IDIOT.Here are my first two lines: LOL - set PhysX gpu accelleration enabled.
roflmao
_____
Then you tell me it says PhySx is enabled - which is what I pointed out. You probably did not go see the linked test results at the other site, and put two and two together.
Look in the mirror and see who can't read, YOU FOOL.
Better luck next time crowing barnyard animal.
"Cluckle red 'el doo ! Cluckle red 'ell doo !"
Let's see, I say PhySx is enabled, and you scream at me to point out it says PhysX is enabled, and call me an nvidia fan because of it - which would make you an nvidia fan as well - according to you, IF you knew what the heck you were doing, which YOU DON'T.
That makes you - likely a red rooster... I may check on that - hopefully you're not a noob poster, too, as that would reduce my probabilities in the discovery phase. Good luck, you'll likely need it after what I've seen so far.
7Enigma - Thursday, April 2, 2009 - link
Looked even closer and the drivers used were different.ATI Drivers:
Anand-9.4 beta
Xbit-9.3
Nvidia:
Anand-185
Xbit-182.08
ancient46 - Thursday, April 2, 2009 - link
I don't see the fun in shooting cloth and unrealistic non impact resistant windows in high rise buildings. The video with the cloth was distracting, it made me wonder why it was there. What was its purpose? My senior eyes did not see much of an improvement in the videos in the CUDA application.SiliconDoc - Monday, April 6, 2009 - link
Maybe someday you'll lose you're raging red fanboy bias, brakdown entirely, toss out your life religion, and buy an nvidia card. At that point perhaps Mirror's Edge will come with it, and after digging it out of the trash can (second thoughts you had), you'll try it, and like anand, really like it - turn it off, notice what you've been missing, turn it back on, and enjoy. Then after all that, you can crow "meh".I suppose after that you can revert to red rooster raging fanboy - you'll have to have your best red bud rip you from your Mirror's Edge addiction, but that's ok, he's a red and will probably smack you for trying it out - and have a clean shot with ow absorbed you'll be.
Well, that should rap it up.
poohbear - Thursday, April 2, 2009 - link
are the driver issues for AMD that significant that it needs to be mentioned in a review article? im asking in all honesty as i dont know. Also, this close developer relationship nvidia has w/ developers. does that show up in any games to significantly give a performance edge for nvidia vid cards? is there an example game out there for this? thanks.7Enigma - Thursday, April 2, 2009 - link
Look no further than this article. :) Here's the quote:"The first thing about Warmonger is that it runs horribly slow on ATI hardware, even with GPU accelerated PhysX disabled. I’m guessing ATI’s developer relations team hasn’t done much to optimize the shaders for Radeon HD hardware. Go figure."
But ATI also has some relations with developers that show an unusually high advantage as well (Race Driver G.R.I.D. for example). All in all, as long as no one is cheating by disabling effects or screwing with draw distances, it only benefits the consumer for the games to be optimized. The more one side pushes for optimizations, the more the other side is forced, or risk losing the benchmark wars (which ultimately decides purchases for most people).
SkullOne - Thursday, April 2, 2009 - link
In the conclusion mentions Nvidia's partners releasing OC boards but nothing about AMD. There is already two versions of the XFX HD4890 on Newegg. One is 850 core and the other is 875 core.The HD4890 is geared to open that SKU of "OC" cards for AMD. People with stock cooling and stock voltage can already push the card to 950+MHz. On the ASUS card you boost voltage to the GPU which has allowed people to get over 1GHz on their GPU. As the card matures seeing 1GHz cores on stock cooling and voltage will become a reality.
It seems like these facts are being ignored.