Print 36 comment(s) - last by littlebrendon.. on Feb 19 at 4:57 PM

Company still tight on details

GPU giant NVIDIA is betting the future of the company on its GF100 graphics chips using the new Fermi architecture. The new chips started mass production last month, and are expected to be used in new video cards that will be launched around the end of February.

There still aren't a lot of specific details about the video cards that will use the new chips, but the company has just released information about the names of the first cards to use GF100 chips.

The GeForce GTX 480 will be the high end video card using the highest clocked speeds, while the GeForce GTX 470 will be a more affordable card with lower core and memory timings. Several defective CUDA cores are likely to be disabled in the GTX 470, which is going to be a big production problem for NVIDIA due to the large die size of the GF100.

NVIDIA seems to be almost entirely skipping the 300-series, with only the GeForce 310 occupying the space. That card is an OEM-only rebadge of the GeForce 210, a low-end DirectX 10.1 part.

Hopefully, this means that all cards in the 400-series will be DirectX 11 capable.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

RE: Nice start...
By jurassic512 on 2/4/2010 1:53:30 PM , Rating: 2
There is nothing on the AMD/ATi 2010 roadmap that shows any new CPU or GPU architectures. All AMD/ATi can do now (and word is they already are) is jack the clocks on the 5870 a la the 4890. But that's what a billion less transistors gets you. I've yet to hear of a mass amount of cards going as high as 90c melting, so heat is not an issue unless your case has space for only 2 80mm fans (or one), or you're just really picky and/or overclock.

The 5000 series isn't all that much different from the 4000 series. GF100 on the other hand is a from the ground up design with DX11 and CUDA in mind.

All AMD/ATi has now are the upcoming 45nm (still) Thurban, and Phenom II's with jacked clocks and TDP/ACP (140w on the soon to arrive, Phenom II 975 @ 3.6GHz) is coming out in 2010. Everything else is the same ol we've been seeing (K8 and K10 variants). All Intel has to do is wait until AMD shows signs of being a threat, and then release what they have been waiting to release. ie: 32nm quad, sexa, octocores. Word is Intel can do octocore natively with Nehelam, but AMD can only do 6 cores natively. So all that delay (First Gen Phenom's), just to have Intel do a native quad first and better.

AMD has been behind for a few years now (performance wise), but clock for clock, Phenom II is ancient. There is a CPU scaling review over at Tom's. C2Q vs Core i7 vs Phenom II @ 2.8GHz. It's not pretty, so don't say i didn't warn you.

"This is from the It's a science website." -- Rush Limbaugh

Most Popular Articles5 Cases for iPhone 7 and 7 iPhone Plus
September 18, 2016, 10:08 AM
Automaker Porsche may expand range of Panamera Coupe design.
September 18, 2016, 11:00 AM
Walmart may get "Robot Shopping Carts?"
September 17, 2016, 6:01 AM
No More Turtlenecks - Try Snakables
September 19, 2016, 7:44 AM
ADHD Diagnosis and Treatment in Children: Problem or Paranoia?
September 19, 2016, 5:30 AM

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki