During the spring of 2006, we find ourselves reflecting on the “early days” when NERSC moved to Berkeley a decade ago. We and many of the NERSC staff can now look back over a full ten years of our professional lives connected to NERSC. Memories of 1996, from the first system installation in a not yet completed computer room to the dedication of the Cray T3E, come back and remind us how far NERSC has progressed since then. So many exciting events have happened during that time, it would be impossible to list them all here.

Instead we will proudly repeat what Secretary of Energy Samuel Bodman recently said about NERSC: “NERSC has a well-earned reputation for providing highly reliable systems, fast turnaround on critical projects, and dedicated support for users.” He made these comments when announcing a renewed allocation of NERSC computer resources to the U.S. Army Corps of Engineers to simulate Gulf Coast hurricanes. In the aftermath of hurricane Katrina, the Corps plans to use the NERSC cycles for simulations of the Texas, Louisiana, and Mississippi coastlines. These simulations will allow researchers to produce more accurate models for calculating the effects of future hurricanes. “Because these simulations could literally affect the lives of millions of Americans, we want to ensure that our colleagues in the Corps of Engineers have access to supercomputers which are up to the task,” the Secretary stated, giving NERSC credit for its proven record of delivering highly reliable production supercomputing services.

We intentionally chose not to commemorate ten years in Berkeley with a big celebration this year. With rapid new developments expected in the next couple of years, our focus must be on the future, not on reminiscing. In 2005 NERSC set the stage for the next five years of its development through a series of important planning activities. Through the well-established Greenbook process, our active user community provided their input to the planning process. NERSC management then developed a new five-year plan for 2006 to 2010, which is discussed in detail in this annual report. This plan was then thoroughly reviewed in a programmatic review by DOE. We can be proud about the outcome of this review of our plans, which were fully endorsed: “NERSC is a strong, productive, and responsive science-driven center that possesses the potential to significantly and positively impact scientific progress.… NERSC is extremely well run with a lean and knowledgeable staff.” This strong endorsement, together with the continued support of our program office at DOE and the good budget news for 2007 and beyond, gives us the most positive outlook we have had in the last several years.

One of the reasons for excitement in the near future is the upcoming introduction of the NERSC-5 system in 2006 and 2007. As we eagerly await the completion of our contract negotiations, we can confidently state that NERSC-5 will be a major step forward in providing a new capability for our users, unlike any of the upgrades in the past half decade. The new system will provide 16 teraflop/s sustained performance on our benchmarks (more than 100 teraflop/s peak), which is about an order of magnitude more sustained performance than NERSC offers today in early 2006.

In addition to our increase of capability computing in the near future, NERSC added significant capacity in 2005 by introducing two new clusters, named “Jacquard” and “Bassi.” This increase in capacity was highly welcomed by the NERSC community. Our users make the most progress in computational science with exactly this type of system that provides high performance in a reliable, predictable environment. One of our users, Robert Duke of the University of North Carolina, commented on the two systems as follows: “I have to say that both of these machines are really nothing short of fabulous. While Jacquard is perhaps the best-performing commodity cluster I have seen, Bassi is the best machine I have seen, period.” By delivering this quality, NERSC makes it possible for its users to concentrate on their science. Not surprisingly, our user community enjoys continued and unparalleled scientific productivity. In 2005 we were able to list more than 1100 scientific publications that were written on the basis of simulations carried out at NERSC.

We also saw a long-term goal accomplished that will significantly set apart the NERSC production environment from other centers. In early 2006, NERSC deployed the NERSC Global Filesystem (NGF) into production, providing seamless data access from all of the Center’s computational and analysis resources. NGF is intended to facilitate sharing of data between users and between machines. NGF’s single unified namespace makes it easier for users to manage their data across multiple systems. Users no longer need to keep track of multiple copies of programs and data, and they no longer need to copy data between NERSC systems for pre- and post-processing. NGF provides several other benefits as well: storage utilization is more efficient because of decreased fragmentation, and computational resource utilization is more efficient because users can more easily run jobs on an appropriate resource.

Thus today NERSC users have the benefit of scalable high-end capability computing in Seaborg (and the soon-to-be-added NERSC-5), along with reliable capacity machines, in an integrated environment that offers a global filesystem, analytics support, and a seemingly infinite mass storage system (now close to 40 petabytes). Our five-year plan will move this integrated system environment to the near petaflop/s performance level, which we will reach in 2010 with the planned introduction of NERSC-6. Scalability to tens of thousands of processors, both for applications and systems software, managing petabytes of data, and at the same time continuing the expected level of support, reliability, and quality of service, will be the big challenges ahead. Thanks to the ongoing support from our program management at the Office of Advanced Scientific Computing Research at DOE, the NERSC budget has been set at a level that makes these ambitious plans feasible. Thus we are confidently looking forward to another year of both scientific and computing accomplishments at NERSC. As always, this progress would not be possible without the NERSC staff, who continue to tirelessly dedicate their time, skill, and effort to make NERSC the best scientific computing resource in the world. Our special thanks to all of you.

Horst D. Simon
NERSC Center Division Director

William T. C. Kramer
Division Deputy and NERSC Center General Manager