NVIDIA GeForce, Tesla and Quadro


NVIDIA Tesla card on motherboard

NVIDIA Tesla card on motherboard

People ask, what is the difference between NVIDIA GeForce, Tesla and Quadro:

in the nutshell:

GeForce: low price gaming card, still very powerful, rolling out newest GPU’s and architectures first like GTX295 double card;

Quadro: corporate pricing, better CAD support, testig is more through, more memory, basically using same GPU-s than GeForce (Quadro FX 5800 = GTX280 240 shader GPU)

Tesla: GPCPU / CUDA computing card, no video output. Tesla C1060 about same as GeForce GTX 280, exactly the same as Quadro FX 5800.

So, for home folding — GeForce is the way to go. F@H is low memory intensive task, additional memory of Tesla and Quadro does not give any advantage. Also the price is way cheaper

There is excellent article in Toms Harware covering the differences of Gaming and Corporate cards.


What is GT300 ?

Original article: http://www.brightsideofnews.com/news/2009/4/22/nvidias-gt300-specifications-revealed—its-a-cgpu!.aspx

What is GT300?

Even though it shares the same first two letters with GT200 architecture [GeForce Tesla], GT300 is the first truly new architecture since SIMD units first appeared in graphical processors.

GT300 architecture groups processing cores in sets of 32 – up from 24 in GT200 architecture. But the difference between the two is that GT300 parts ways with the SIMD architecture that dominate the GPU architecture of today. GT300 Cores rely on MIMD-similar functions – all the units work in MPMD mode, executing simple and complex shader and computing operations on-the-go. We’re not exactly sure should we continue to use the word “shader processor” or “shader core” as these units are now almost on equal terms as FPUs inside latest AMD and Intel CPUs.

GT300 itself packs 16 groups with 32 cores – yes, we’re talking about 512 cores for the high-end part. This number itself raises the computing power of GT300 by more than 2x when compared to the GT200 core. Before the chip tapes-out, there is no way anybody can predict working clocks, but if the clocks remain the same as on GT200, we would have over double the amount of computing power.
If for instance, nVidia gets a 2 GHz clock for the 512 MIMD cores, we are talking about no less than 3TFLOPS with Single-Precision. Dual precision is highly-dependant on how efficient the MIMD-like units will be, but you can count on 6-15x improvement over GT200.

This is not the only change – cluster organization is no longer static. The Scratch Cache is much more granular and allows for larger interactivity between the cores inside the cluster. GPGPU e.g. GPU Computing applications should really benefit from this architectural choice. When it comes to gaming, the question is obviously – how good can GT300 be? Please do bear in mind that this 32-core cluster will be used in next-generation Tegra, Tesla, GeForce and Quadro cards.

This architectural change should result in dramatic increase in Dual-Precision performance, and if GT300 packs enough registers – performance of both Single-Precision and Dual-Precision data might surprise all the players in the industry. Given the timeline when nVidia begun work on GT300, it looks to us like GT200 architecture was a test for real things coming in 2009.

Just like the CPU, GT300 gives direct hardware access [HAL] for CUDA 3.0, DirectX 11, OpenGL 3.1 and OpenCL. You can also do direct programming on the GPU, but we’re not exactly sure would development of such a solution that be financially feasible. But the point in question is that now you can do it. It looks like Tim Sweeney’s prophecy is slowly, but certainly – coming to life.

So, worth of waiting..

Nice game to understand protein folding

fold it

fold it


You may like to download and try out: http://fold.it/portal/

This game explains step-by-step basics of protein folding, including side-chain positioning, backbone packing, hydrogen bonding and other..


One small step for Stanford, one giant leap for Me …

I broke 100 000 PPD barrier. Actually my current rig generates in average 135 000 PPD.


Estonia Donates 100K PPD

Estonia Donates 100K PPD


Latest configuration: 20 GPUs = 17,9 TFLOPS

  • 4800 streaming processors + 14 CPU cores
  • 17,9  arithmetic TFLOPS (yes, 17 900 GFLOPS)
  • Max 140 000 PPD (Folding@home, Fahmon) 
Estonia Donates GPU supercomputer, 20 GPU-s 17,9 TFLOPS

Estonia Donates GPU supercomputer, 20 GPU-s 17,9 TFLOPS

GPU card for Folding to choose today

We know, that in game-playing world there are 2 titans — NVIDIA and ATI, pushing bar higher and higher. In Folding world the is one king only — NVIDIA. ATI cards are powerful on paper, but in real (folding)life they lag behind.

Here is quite useful table to choose right GPU card for folding:

NVIDIA folding gpu card PPDs

NVIDIA folding gpu card PPD's

Having ambitious plans, there is currently only one way to go — GTX295 double card. In the table only 1/2 card PPD performance is shown. Other cards are single GPU cards and allow max 4 GPU conf’s, as GTX295 doubles one rig capacity.

GPU news

Nvidia GT200 GT300 roadmap



Nvidia GT200 GT300 roadmap


How long rules current 55nm Nvidia GTX295, based on GT200 core? We know, that 2009 Q4 we’ll see lot of good news:

  • CUDA 3.0
  • GT300 core
  • DX11

GT300 will be possibly first 40nm card from Nvidia. Better power efficiency helps to get heat down from current near 300W readings per GTX295. The GT300’s architecture will be based on a new form of number-crunching machinery. While today’s NVIDIA GPUs feature a SIMD (single instruction multiple data) computation mechanism, the GT300 will introduce the GPU to MIMD (multiple instructions multiple data) mechanism. This is expected to boost the computational efficiency of the GPU many-fold. The ALU cluster organization will be dynamic, pooled, and driven by a crossbar switch. Once again, NVIDIA gets to drop clock-speeds and power consumptions, while achieving greater levels of performance than current-generation GPUs. With GT300, NVIDIA will introduce the next major update to CUDA. With the new GPUs being built on the 40nm silicon fabrication process, transistor counts are expected to spiral-up. NVIDIA’s GT300 is expected to go to office in Q4 2009, with its launch schedule more or less dependent on that of Microsoft’s Windows 7 operating system that brings in DirectX 11 support (source: Harware-Infos).

CUDA Roadmap

CUDA Roadmap

CUDA 3.0 is mainly good news to developers, but we will see more and more GPGPU capale applications like Adobe CS3 and other…