Performance Comparison between the Earth Simulator & US-made Supercomputers
Dr. Leonid Oliker of NERSC/LBNL held the Seminar titled "Scientific Computations on Modern Parallel Vector Systems" at Miyoshi memorial hall in October 14, 2004.
He reported the result of performance evaluation of the benchmark code on plasma physics, material science, astrophysics, and magnetic fusion by using the Earth Simulator as the collaboration between NERSC and ESC in last December.In the past, they only evaluated the scalar type of HPC, however they achieved the high-performance which ever been achieved by the other machines in 10 days operation.The result of this performance evaluation was explained with comparing to HPC machines: IBM, SGI, and CRAY. TOP500 evaluates only by the Linpack HPC that is Large-scale matrix computation, where as this research covers various field.The result clearly indicates that the Earth Simulator achieves high score in peak performance ratio, and high scalability in highly parallelized computations in various applications.
- Date:
- 2004/10/14 14:00-15:00
- Place:
- Miyoshi Memorial Auditorium in 2nd floor, Conference Building, Yokohama Institute of Independent Administrative Institution, Japan Agency of Marine-Earth Science and Technology.
- Speaker:
- Dr. Leonid Oliker (web page: http://www.nersc.gov/~oliker/)
- Title:
- "Scientific Computations on Modern Parallel Vector Systems"
- Language:
- English
- Abstract:
- Computational scientists have seen a frustrating trend of stagnating application performance despite dramatic increases in the claimed peak capability of HPC systems. This trend has been widely attributed to the use of superscalar-based commodity components whose architectural designs are unbalanced for large-scale numerical computations. Recently, two innovative parallel-vector architectures have become operational: the Japanese Earth Simulator (ES) and the Cray X1. In order to quantify what these modern vector capabilities entail for the scientists that rely on modeling and simulation, it is critical to evaluate this architectural paradigm in the context of demanding computational algorithms. Our evaluation study examines four diverse scientific applications with the potential to run at ultrascale, from the areas of plasma physics, material science, astrophysics, and magnetic fusion.We compare performance between the vector-based ES and X1, with leading superscalar-based platforms: the IBM Power3/4 and the SGI Altix. Our research team was the first international group to conduct a performance evaluation study at the EarthSimulatorCenter. Results demonstrate that the vector systems achieve excellent performance on our application suite -- the highest of any architecture tested to date. However, vectorization of a particle-in-cell code highlights the potential difficulty of expressing irregularly structured algorithms as data-parallel programs.
- Briefing paper
- Scientific Computations on Modern Parallel Vector Systems (PDF)