An Overview of the Intel TFLOPS Supercomputer
作者： Timothy G. Mattson
DOI:
关键词:
摘要: Computer simulations needed by the U.S. Department of Energy (DOE) greatly exceed capacity world’s most powerful supercomputers. To satisfy this need, DOE created Accelerated Strategic Computing Initiative (ASCI). This program accelerates development new scalable supercomputers and will lead to a supercomputer early in next century that can run at rate 100 trillion floating point operations per second (TFLOPS). Intel built first computer program, ASCI Option Red Supercomputer (also known as TFLOPS supercomputer). system has over 4500 nodes, 594 Gbytes RAM, two independent 1 Tbyte disk systems. Late spring 1997, we set MP LINPACK world record 1.34 TFLOPS. In paper, give an overview Supercomputer. The motivation for building is presented hardware software views machine are described detail. We also briefly discuss what it like use machine.
parallel.ru 参考文章(9)
Greg Henry, Pat Fay, Ben Cole, Timothy G Mattson, The Performance of the Intel TFLOPS Supercomputer ,(1998)
Bradley Mitchell, Scalable Platform Services on the Intel TFLOPS Supercomputer ,(1998)
J. J. Dongarra, C. B. Moler, G. W. Stewart, J. R. Bunch, LINPACK Users' Guide ,(1987)
S.L. Lillevik, The Touchstone 30 Gigaflop DELTA Prototype distributed memory computing conference. pp. 671- 677 ,(1991) , 10.1109/DMCC.1991.633353
E. Barszcz, One year with an iPSC/860 COMPCON Spring '91 Digest of Papers. pp. 283- 288 ,(1991) , 10.1109/CMPCON.1991.128819
Wheat, Maccabe, Riesen, van Dresser, Stallcup, PUMA: an operating system for massively parallel systems hawaii international conference on system sciences. ,vol. 2, pp. 56- 65 ,(1994) , 10.1109/HICSS.1994.323279
Jack J. Dongarra, Performance of various computers using standard linear equations software in a Fortran environment international conference on advances in system simulation. ,vol. 49, pp. 51- 62 ,(1987) , 10.1177/003754978704900202
Y. Abei, K. Itakura, T. Boku, H. Nakamura, K. Nakazawa, Performance improvement for matrix calculation on CP-PACS node processor ieee international conference on high performance computing data and analytics. pp. 672- 677 ,(1997) , 10.1109/HPC.1997.592229
Gronbech-Jensen, Tamayo, Lomdahl, Beazley, 50 GFlops molecular dynamics on the Connection Machine-5 IEEE. pp. 520- 527 ,(1993) , 10.1109/SUPERC.1993.1263501