|
nVIDIA – WORLD LEADER IN GPU COMPUTING TECHNOLOGIES
General Purpose Graphical Processing Units (GPGPU) has been making tremendous inroads into HPC applications. The GPU based Personal Supercomputers (PSC) are providing staggering levels of computational performance to tackle large scale problems far faster than a small server cluster at rivalling prices.
With GPU technologies refreshed at a rapid pace, the leaps in sheer processing power over conventional CPUs have opened up new opportunities for standalone and clustered computing solutions using GPUs.
Multi-Teraflops performance is now packaged into small enclosures of 1U and 4U form factors.
NVIDIA, founded in 1993, invented the GPU technology. Since the invention of GPU in 1999, NVIDA has been setting standards in performance and price points and made Personal Supercomputing compact and highly affordable at individual, departmental and organizational levels while saving space and power consumption .
Presently NVIDIA offers two major GPU products.
Tesla 10
240 GPU cores
933 GFlops single precision and 78 GFlops double precision performance
Tesla 20 - FERMI
Next generation GPU
480+ GPU cores
ECC memory support
1 TF single precision and 515 GFlops double precision performance
Both Tesla 10 and Tesla 20 are supported on Microsoft Windows and Linux platforms
CUDA
CUDA is NVIDIA’s parallel computing architecture that enables dramatic increases in computing performance by harnessing the power of the GPU. CUDA is the computing engine in Tesla GPUs that is accessible to software developers through industry standard programming languages such as C.
For more information, please visit
www.nvidia.com
Hinditron offers total GPU solutions including application porting and support services.
For details, please contact us at gpu@hinditron.com
|
Information Technology
