Tesla 20-Series Features and Advantages

[mipko]

Hello eyalhir,

I dont understand this:

“NVIDIA GPUs will be available to applications running as a Windows service (i.e. in Session 0) on systems running the TCC drivers” → does this

mean that my application must run as a service? I can’t run it as a Windows GUI application?

No, it means that applications in session 0 will be able to access nV hardware.

Release notes and new features in new software looks generated with some buzzword generator :) Bunch of terms poorly explained and hardly understandable as a whole :)

Regarding RemoteDesktopFX, no joy - no cuda devices found under RDP. Call me stupid but didn’t understand what is so revolutionary in RemoteFX :)

For time being I give up totally on remote dev and deployment.

regards

Mirko

[SPWorley]

Tesla 20-series (Fermi-based) products are designed for high performance

scientific and technical GPU computing.

They thus have features, testing, and support over and above our consumer

GeForce GTX 470 and 480 (Fermi-based) products such as:

This is a great enumeration of Tesla vs GeForce feature differences.

Which of these features are also enabled for the GF100 Quadros?

I’m especially interested in the performance differences of DP rate, dual DMA, and ECC.

Conversely, which Quadro features are enabled on Tesla? Does Tesla have the same fast OpenGL wireframe speeds?

[SPWorley]

Tesla 20-series (Fermi-based) products are designed for high performance

scientific and technical GPU computing.

They thus have features, testing, and support over and above our consumer

GeForce GTX 470 and 480 (Fermi-based) products such as:

This is a great enumeration of Tesla vs GeForce feature differences.

Which of these features are also enabled for the GF100 Quadros?

I’m especially interested in the performance differences of DP rate, dual DMA, and ECC.

Conversely, which Quadro features are enabled on Tesla? Does Tesla have the same fast OpenGL wireframe speeds?

[Simon_Green]

Yes, Fermi-based Quadros have fast double precision, dual DMA engines and ECC:

NV_DS_QUADRO_6000_Oct10_US_LR.pdf

http://www.nvidia.com/docs/IO/40049/Dual_copy_engines.pdf

Tesla has basically the same graphics feature set as GeForce. There’s a (somewhat outdated) description of the Quadro graphics advantages here:
http://www.nvidia.com/object/quadro_geforce.html

[Simon_Green]

Yes, Fermi-based Quadros have fast double precision, dual DMA engines and ECC:

NV_DS_QUADRO_6000_Oct10_US_LR.pdf

http://www.nvidia.com/docs/IO/40049/Dual_copy_engines.pdf

Tesla has basically the same graphics feature set as GeForce. There’s a (somewhat outdated) description of the Quadro graphics advantages here:
http://www.nvidia.com/object/quadro_geforce.html

[SPWorley]

Yes, Fermi-based Quadros have fast double precision, dual DMA engines and ECC:

http://www.nvidia.com/docs/IO/40049/NV_DS_QUADRO_6000_Oct10_US_LR.pdf

http://www.nvidia.com/docs/IO/40049/Dual_copy_engines.pdf

Tesla has basically the same graphics feature set as GeForce. There’s a (somewhat outdated) description of the Quadro graphics advantages here:

Page Not Found | NVIDIA

What’s the difference between the Quadro 6000 and Tesla C2070?

They have the same MSRP, and seem to have the same hardware specs. (GF100, 448SP, 1150GHz) But the Quadro has the fast workstation OpenGL performance and the Tesla C2070 does not.