NVIDIA Fermi architecture released

Posted on Monday, January 18 2010 @ 15:02 CET by Thomas De Maesschalck
NVIDIA has paperlaunched its GF100 "Fermi" GPU, lots of details and some performance information are now available on the web. The company's upcoming DX11 flagship will feature a total of three billion transistors on a 40nm process, with 512 CUDA cores, 64 texture units, 48 ROPs, and a 384-bit GDDR5 memory bus.

The new architecture features some significant changes and promises improvements for GPGPU computing such as eight times higher peak double-precision computing performance, ECC support and C++ code support, but also improvements for gamers like tessellation, improved image quality (much better CSAA IQ), and faster performance. It's still unknown when Fermi will be available in retail stores, if things go as planned it may take another two months or so until these GPUs will be available in large quantities.

Lets take a look at a couple of previews to see what other sites think, first up is the one from AnandTech. They conclude NVIDIA has taken a big risk, and say we'll have to wait a bit longer to see if Fermi has what it takes to win.
NVIDIA has taken a big risk on GF100, first with its compute abilities for GPGPU use, then on its geometry abilities for gaming, and now the risk is time. Being 6 months late has hurt NVIDIA, and being 6 months late has hurt consumers through uncompetitive pricing from AMD. By no means is the situation dire, but we can quickly come up with some scenarios where it is if NVIDIA can’t convincingly beat AMD in gaming performance. NVIDIA has shown their cards, and they’re all in. Now in the next couple of months we’ll see if they’re bluffing or if they really have what it takes to win. Stay tuned.
Another preview can be found at PC Perspective, the reporter writes he's now far more positive in his outlook for the Fermi architecture when it comes to traditional gaming. He also notes NVIDIA did a lot of work to make the GF100 architecture more modular, meaning they can now create multiple products on the basis of these SM and GPC units, and predicts mid-range solutions based on Fermi may cause a lot of trouble for AMD.
After seeing what NVIDIA actually has in store, I am far more positive in my outlook for the Fermi architecture when it comes to traditional gaming, as well as upcoming DX11 gaming. Off the cuff I would guess that in current applications, the GF100 based high end card will be 15% to 20% faster overall than the HD 5870. It will not overtake the HD 5970 dual GPU card, but it looks to certainly trounce the single HD 5870. I think that NVIDIA is also doing tessellation right, and when games come out which will implement it in a more pervasive manner, NVIDIA will take a much larger performance lead.
There's also a preview over at HardOCP, if you still want to learn more about Fermi. The site concludes they've come away more excited about the GF100 than they've ever been, but at this point it's hard to tell whether this card will offer better value than the Radeon HD 5870.
After spending time with NVIDIA on GF100 I have actually come away feeling more positive about it actually performing quite well. When I saw the GF100 playing games, it was doing so at a very fast framerate. I think GF100 is going to be faster than 5870. Is GF100 going to be a better value? I am not so sure about that. This fact is the 5850 is the best value in "real" gaming right now and this is where NVIDIA is going to have to compete. I am not sure NVIDIA is going to be able to do that for quite a while.

About the Author

Thomas De Maesschalck

Thomas has been messing with computer since early childhood and firmly believes the Internet is the best thing since sliced bread. Enjoys playing with new tech, is fascinated by science, and passionate about financial markets. When not behind a computer, he can be found with running shoes on or lifting heavy weights in the weight room.

Loading Comments