GPUs Deliver 25X Speed Up for Physics and Energy Research at University of Maryland

University of Maryland is using GPUs to understand ionized plasma under thermal nuclear conditions for physics and energy research, research that is aiding the design of new energy generation devices. The university is using large GPU clusters to provide much more science for the dollar, while reducing calculation times from a day to an hour.

Category:

Science & Technology

Tags:

• NVIDIA
• Tesla
• CUDA
• GPU
• parallel
• computing
• supercomputing
• Maryland
• Dorland
• astrophysics
• thermal
• nuclear

[MrJama1ca]

That guy is smart AS fuck damn it i wish i had did math in college

[shineclot]

I love cuda

[daomingjin]

The AMD HD6990 is just over 4 TeraFlops of double precision. So i actually think you could use ATI cards if you wanted to. However Nvidia cards seem to be much more efficient in terms of Electrical Power Efficiency

[thedevo01]

Let's say you and I wanted to build a render farm!

If the 4TeraFlops of the Tesla S1070 is 205 times more than the 20437MegaFlops of the i7 980x, does it mean that you would need 205 CPUs (i7 980x) just to have equal performance to that of a Tesla S1070?

Or is this calculation false because the CPUs would perform better/worse when synchronized?

[wwyk1993]

@halfpol No one uses AMD/ATi GPUs in supercomputers.

[thedevo01]

0:48

[GameDoctor21]

this guy need to laugh

[aronjakob1]

@halfpol mabe, but can your ATI do this!

[halfpol]

4890 x2 pwn3d 260x3

[halfpol]

no!very bad!!!!