NVIDIA Tesla GPU Computing Module

The NVIDIA Tesla Computing Modules enable seamless integration of GPU computing with HP ProLiant servers for scale-out deployments. The Tesla GPUs have a capacity of up to 6 GB of GDDR5 memory with ECC, and peak performance of 655 double-precision Gflops or over 1.3 single-precision Tflop.

The Tesla M2070Q combines Tesla's high performance computation with the NVIDIA Quadro® professional-class visualization in the same GPU.

The HP GPU Ecosystem includes the CUDA programming interface and third-party CUDA development tools, as well as HP-supported GPU-aware cluster software such as the HP Cluster Management Utility (CMU). CMU will monitor and display GPU health such as temperature and will also install and provision the GPU drivers and the CUDA software. In addition, HP HPC Linux Value Pack includes a GPU-enhanced version of Platform LSF, with the capability of scheduling jobs based on GPU requirements.

Ideal for...

High performance computing and large data centers offering leading performance with minimal power consumption, for research and engineering

Introduction of the NVIDIA Tesla M2075 as a lower-power version of the latest NVIDIA passive-module computational accelerator architecture.

The Industry's Fastest Multi-core Processors, Integrated into HP ProLiant Servers.

• May dramatically reduce computational time for parallel applications. Reduces time to solution, thus enabling businesses to solve problems previously limited by CPU speed.
• The high density of performance decreases the physical footprint required for timely solutions.
• The combination of GPU processors – delivering high performance per watt - and HP ProLiant servers – the industry's most power-efficient computers – dramatically reduces the carbon footprint required for timely solutions.
• Enables solving problems not otherwise solvable within the customer's budget.
• With the Tesla M2070Q, you can do high performance computing and advanced visualization with a single GPU card. This allows rapid graphical analysis of large datasets generated through high-end simulations on the same server without inefficiency of moving the data to a graphics server.