A Plan for Powering Up Processors

While today's microprocessors are vastly more powerful than those of even a few years ago, there are some computing challenges that make even the fastest computer seem like the 1950s' Univac. Many of these challenges are related to national security issues such as weapon system simulations and processing of satellite images. The government's reliance on clusters of commercial-off-the-shelf components falls far short of solving these security issues.

A new plan from the High-End Computing Revitalization Task Force (HECRTF), a group chartered by the National Science and Technology Council (an executive-branch group that coordinates science, space, and technology research and development), lays out its vision for ensuring U.S. leadership in high-end computing.

The plan recommends "a coordinated, sustained research program over 10-15 years to overcome major technology barriers that limit effective use of high-end computer systems." These barriers include poor system reliability and restrictions on processor and memory performance.

Federal government resources for high-end computing are limited. Some agencies that have science and technology missions have no access to high-end computers, and available resources are "heavily oversubscribed." Another resource issue is the shortage of powerful computing systems. The plan suggests that a "robust investment in high-end computing" will accelerate research and "send a strong signal" about the government's commitment to high-end computing.

The plan also suggests pilot projects to improve the efficiency of the procurement process, including benchmarking, developing new models for total cost of ownership, and sharing procurement processes across agencies.

@ The HECRTF's plan