Archive Home arrow Site News arrow SC07 SuperComputing Coverage - NVIDIA
SC07 SuperComputing Coverage - NVIDIA E-mail
News - Featured Website News
Written by Olin Coles   
Tuesday, 13 November 2007

On day number one of SC07, the SuperComputing exposition hall was officially opened to the media and select visitors for open coverage. Many of those in attendance gravitated to some of the largest names in the high performance computing industry such as Sun Microsystems, Microsoft, and Intel. However Benchmark Reviews spent some quality time at a few exhibits by manufacturers with roots to the enthusiast segment. In this article we brief you on the Tesla project from NVIDIA.

SC07 SuperComputing Coverage - NVIDIA & Koolance

The CPU and operating system powering the modern PC solve an incredibly difficult problem in computing. As you use the computer, the operating system tracks all your activities, communicates in the background, and organizes the information you use while you're listening to music, browsing the Web, and reading e-mail. Even though the CPU works on separate tasks one at a time, it has enough speed so these serial tasks appear to operate simultaneously. With new multi-core CPUs, each core can handle an additional task with true simultaneity. A different class of computing problem, parallel computing, has until recently remained the realm of large server clusters and exotic supercomputers. Standard CPU architecture excels at managing many discrete tasks, but is not particularly efficient at processing tasks that can be divided into many smaller elements and analyzed in parallel. This is exactly the type of problem solved by graphics processing units (GPUs).

tesla_main.jpg

The GPU has great potential for solving such problems quickly and inexpensively. GPU computing makes supercomputing possible with any PC or workstation and expands the power of server clusters to solve problems that were previously not possible with existing CPU clusters. The goal of computing with GPUs is to apply the tremendous computational power inherent in the GPU to solve some of the most difficult and important problems in high performance computing.

sc07_nvidia_tesla_01.jpg

Revolutionary NVIDIA Tesla high performance computing (HPC) solutions arm scientists, engineers and other technical professionals with the power to solve previously unsolvable problems. A dedicated, high performance GPU computing solution, Tesla brings supercomputing power to any workstation or server and to standard, CPU-based server clusters.

Key elements include:

  • Industry's first massively multi-threaded architecture with a 128-processor computing core
  • World's only C-language development environment for the GPU
  • A suite of developer tools (C-compiler, debugger, performance profiler, optimized libraries)
  • Largest ISV development community for GPU Computing applications
  • Seamlessly able to fit into existing HPC environments

tesla_gpu_header.jpg

Power to Solve Complex Problems
Scientists and researchers benefit from the power of the massively parallel, computing architecture in Tesla solutions. This availability of supercomputing will unlock the answers to previously unsolvable problems in systems ranging from a workstation to server clusters.

sc07_nvidia_tesla_03.jpg

Massively Parallel Performance
A massively multi-threaded processor mines large datasets to extract information and reach answers fast. Key features available on Tesla GPUs include the Thread Execution Manager to coordinate the concurrent execution of thousands of parallel computing threads and the Parallel Data Cache to enable those threads to share data easily.

sc07_nvidia_tesla_02.jpg

Developer Community
NVIDIA is the catalyst for the largest GPU computing developer community. NVIDIA's interactive, on-line GPU developer community provides access to forums, educational materials, and additional resources and tools.

C for the GPU
The world's only C-language development environment for the GPU, the NVIDIA CUDA software development kit includes a standard C compiler, hardware debugger tools, and a performance profiler for simplified application development.

sc07_nvidia_tesla_04.jpg

Compatible Solutions
As an industry-standard solution, Tesla easily fits into existing HPC environments. Available products include a GPU for users to upgrade their existing workstation, a deskside supercomputer to add additional performance alongside a workstation, and a 1U GPU server for deployment within an enterprise data center.

Related Articles:


Related Articles:
 

Comments have been disabled by the administrator.

Search Benchmark Reviews
QNAP Network Storage Servers

Follow Benchmark Reviews on FacebookReceive Tweets from Benchmark Reviews on Twitter