SCIENCE

AMAX has announced that it is showcasing its high-performance GPGPU technologies at the SC10 conference. The latest AMAX offerings combine the massively parallel compute performance of NVIDIA GPUs with the fastest multi-core Intel and AMD processors. The high-density AMAX systems boast gold and platinum-level power supplies and feature InfiniBand and fiber channel interconnect options to limit bottlenecks in order to drive the peak performance necessary for mission-critical HPC applications.

AMAX's next-generation GPU solutions are developed around NVIDIA's CUDA programming environment and Tesla GPUs to fully maximize performance and efficiency. In many cases, processing tasks that are simply not possible on CPU-based clusters and workstations are being enabled by Tesla 20-series GPUs. When properly optimized using best-of-breed components within validated supercomputing configurations, AMAX's GPU platforms take full advantage of Tesla GPUs combined with Intel Xeon or AMD Opteron processor-based servers to deliver eight times the performance of CPU-only solutions. AMAX's GPU platforms are ideal for a wide range of visualization and compute intensive applications, revolutionizing industries such as oil and gas, finance, medical and life science. AMAX's GPGPU expertise is most evident when it comes to configuring clusters, achieving density as high as 16 Tesla M2050 or M2070 GPUs per 5U of rack space, or up to 128 Tesla cards in a 42U rack while still maintaining maximum flexibility and scalability to address current needs while planning for future potentials of the most demanding HPC projects.

"AMAX has a reputation for quality, capability and reliability," said Shanker Trivedi, Vice President, Worldwide PSG Sales for NVIDIA. "AMAX was one of the first certified Tesla preferred partners and they continue to deliver exceptional value to their global customers through excellent HPC solutions and services."

"AMAX continues to collaborate with NVIDIA to promote the latest GPU accelerator technologies for HPC computing," said Jean Shih, President at AMAX. "From individual servers to hundred-node clusters, our experience with integrating NVIDIA Tesla GPUs into our high-performance HPC Series products ensure that our customers are always deploying the most cutting-edge supercomputing solutions optimized dynamically to address the most difficult mission-critical technical computing challenges."

AMAX products featured at the show include:

• ServMax Highest-Density GPGPU Cluster Solution is comprised of an advanced 3U 16GPU base server and two 1U Intel or AMD host systems to deliver the highest-density 5U 16GPU cluster building block solution on the market today.
• StorMax Ultra-Dense JBOD Expansion Storage with dual SAS 2.0 module fill-over design and 4 SAS 6Gb/sec ports. Features dual redundant power supply and is an ideal building block to scale out NAS or scale up NAS.
• ServMax GPGPU Personal Supercomputer is powered by the massively parallel CUDA architecture and supports four NVIDIA Tesla C2050 or C2070 GPUs transforming a workstation into the equivalent of a small cluster at 5% the power consumption and 10% the cost. It delivers an astounding 4.12 Teraflops of performance in a single chassis.
• ServMax GPGPU Supercomputing Server with 2.06 Teraflops of computing horsepower is the fastest 1U server available. Powered by the massively parallel CUDA architecture and Tesla™ GPU computing processors, it can seamlessly fit into existing HPC environments and is ideal for financial services applications.

AMAX is investing heavily into GPU computing solutions and is helping customers such as China University of Petroleum, an important scientific research base for petroleum and petrochemical industries, and Tongji University, the leading university directly under the State Ministry of Education in China, with innovative ways to accelerate applications and shorten computational cycles. Both universities' systems were upgraded with NVIDIA Tesla GPUs, which greatly shortened the processing time required for calculating research data by one-third in comparison to previous CPU-based performance. As a result, AMAX's industry-leading GPU cluster solutions allowed the universities to fully utilize their IT infrastructure, optimize their supercomputing power and expedite their research processes to achieve faster, tangible results.

AMAX will be demonstrating its latest GPGPU innovations at SC10, which will be held at Ernest N. Morial Convention Center in New Orleans, LA in the booth #4546, November 15^th through 18^th, 2010.