TY - JOUR
T1 - Overview of the QCDSP and QCDOC computers
AU - Boyle, Peter A.
AU - Chen, Dong
AU - Christ, Norman H.
AU - Clark, Michael A.
AU - Cohen, Saul D.
AU - Cristian, Calin
AU - Dong, Zhihua
AU - Gara, Alan
AU - Joó, Bálint
AU - Jung, Chulwoo
AU - Kim, Changhoan
AU - Levkova, Ludmila A.
AU - Liao, Xiadong
AU - Liu, Guofeng
AU - Mawhinney, Robert D.
AU - Ohta, Shigemi
AU - Petrov, Konstantin
AU - Wettig, Tilo
AU - Yamaguchi, Azusa
PY - 2005
Y1 - 2005
N2 - The QCDSP and QCDOC computers are two generations of multithousand-node multidimensional mesh-based computers designed to study quantum chromodynamics (QCD), the theory of the strong nuclear force. QCDSP (QCD on digital signal processors), a four-dimensional mesh-machine, was completed in 1998; in that year, it won the Gordon Bell Prize in the price/performance category. Two large installations - of 8,192 and 12,288 nodes, with a combined peak speed of one teraflops - have been in operation since. QCD-on-a-chip (QCDOC) utilizes a six-dimensional mesh and compute nodes fabricated with IBM system-on-a-chip technology. It offers a tenfold improvement in price/performance. Currently, 100-node versions are operating, and there are plans to build three 12,288-node, 10-teraflops machines. In this paper, we describe the architecture of both the QCDSP and QCDOC machines, the operating systems employed, the user software environment, and the performance of our application - lattice QCD.
AB - The QCDSP and QCDOC computers are two generations of multithousand-node multidimensional mesh-based computers designed to study quantum chromodynamics (QCD), the theory of the strong nuclear force. QCDSP (QCD on digital signal processors), a four-dimensional mesh-machine, was completed in 1998; in that year, it won the Gordon Bell Prize in the price/performance category. Two large installations - of 8,192 and 12,288 nodes, with a combined peak speed of one teraflops - have been in operation since. QCD-on-a-chip (QCDOC) utilizes a six-dimensional mesh and compute nodes fabricated with IBM system-on-a-chip technology. It offers a tenfold improvement in price/performance. Currently, 100-node versions are operating, and there are plans to build three 12,288-node, 10-teraflops machines. In this paper, we describe the architecture of both the QCDSP and QCDOC machines, the operating systems employed, the user software environment, and the performance of our application - lattice QCD.
UR - http://www.scopus.com/inward/record.url?scp=21044436873&partnerID=8YFLogxK
U2 - 10.1147/rd.492.0351
DO - 10.1147/rd.492.0351
M3 - Article
AN - SCOPUS:21044436873
SN - 0018-8646
VL - 49
SP - 351
EP - 365
JO - IBM Journal of Research and Development
JF - IBM Journal of Research and Development
IS - 2-3
ER -