With the world's first teraflop many-core processor, NVIDIA®Tesla™ high performance computing (HPC) solutions enable the necessary transition to energy efficient parallel computing power. With 240 cores per processor and a standard C compiler that simplifies application development, Tesla™ scales to solve the world's most important computing challenges-more quickly and accurately.
Tesla is a dedicated, high performance GPU computing solution, bring supercomputing power to any workstation or server and to standard, CPU-based server clusters.
Key features:
Industry's first massively multi-threaded architecture with a 240-processor computing core per processor.
Many-core architecture delivers optimum scaling across HPC applications.
Scale to thousands of processor cores to solve large-scale problems by splitting the problem across multiple GPUs.
High-efficiency computing platform for energy-conscious organizations.
NVIDIA CUDA technology unlocks the power of Tesla many-core computing products.
Seamlessly able to fit into existing HPC environments.
Four Tesla processors (240 computing cores per processor, 960 cores total)
Delivers four teraflops in a 1U chassis with a typical energy footprint of only 700 watts
IEEE 754 single & double floating point precision
16 GB dedicated memory (organized as 4.0 GB per GPU)
4x 512-bit GDDR3 memory interface (organized as a 512-bit interface per GPU)
408 GB/sec memory bandwidth (102 GB/s per GPU to local memory)
Standard 1U rackmount chassis
Connects to host server via cabling to a PCI Express x16 host adapter card (included), optional PCI Express x8 host adapter card Up to three Tesla processors (240 computing cores per processor, 720 cores total)
Configuration:
Reference Number:
S1070-500
S1070-400
With 2 x PCI-E Gen2 links serving 4 GPUs
With 2 x PCI-E Gen2 links serving 4 GPUs
P/N: 5003178
P/N: 5003179
Copyright ® 2010 AMAX Information Technologies. All rights reserved.
1U Rackmount
