HP delivers high-performance accelerators
- By Rutrell Yasin
- Nov 01, 2007
Hewlett-Packard has teamed up with software and hardware vendors to deliver new high-performance computing accelerators that speed the analysis of large amounts of data and perform complex simulations.
The HPC Accelerator and enhanced Multi-Core Optimization Programs are the first in a set of initiatives HP plans to address critical challenges facing users with intensive computing requirements.
Accelerators are alternative technologies ' such as general-purpose graphical processing units, field-programmable gate arrays and custom application-specific integrated circuits ' that help applications run faster. The technology is well suited for users in the areas of oil and gas, financial services, government and defense, life and material science and digital content creation, HP officials said.
'As individual processor cores are reaching their physical limits of complexity and speed, we can no longer rely on processor technology alone to meet users' constant demand for performance growth,' said Winston Prather, vice president and general manager of high performance computing at HP.
HP has partnered with vendors to ensure that qualified accelerator products can be integrated into HP ProLiant and BladeSystem servers. Partners include:
- ClearSpeed Technology, which offers Advance e620 accelerator cards that bring 64-bit, low-power acceleration to HP ProLiant and BladeSystem users;
- Celoxica, which has a RCHTX acceleration board using AMD Torrenza technology on the HTX channel in HP ProLiant DL145 G3 servers; and
- NVIDIA, AMD, RapidMind and Mitrionics, which are contributing innovative programming technologies and tools to simplify development of accelerated applications.
The HP Multi-Core Optimization Program, launched in June, now includes new partners and a toolkit that shows how to effectively use multicore technology. The toolkit is a set of products, reference material, HP best-practices, white papers and links to HP and partner Web sites.
New members of the program include Allinea, Interactive Supercomputing, The Portland Group, RapidMind, Stanford Pervasive Parallelism Lab and Visual Numerics.
Rutrell Yasin is is a freelance technology writer for GCN.