IBM®
Skip to main content
    Country/region [change]    Terms of use
 
 
 
    Home    Products    Services & solutions    Support & downloads    My account    

IBM Technical Journals
Special report: Celebrating 50 years of the IBM Journals
All topics > Computing System Architectures >

Overview of the Blue Gene/L system architecture

 Award plaque by A. Gara,
M. A. Blumrich,
D. Chen,
G. L.-T. Chiu,
P. Coteus,
M. E. Giampapa,
R. A. Haring,
P. Heidelberger,
D. Hoenicke,
G. V. Kopcsay,
T. A. Liebsch,
M. Ohmacht,
B. D. Steinmacher-Burow,
T. Takken,
and P. Vranas

The Blue Gene®/L computer is a massively parallel supercomputer based on IBM system-on-a-chip technology. It is designed to scale to 65,536 dual-processor nodes, with a peak performance of 360 teraflops. This paper describes the project objectives and provides an overview of the system architecture that resulted. We discuss our application-based approach and rationale for a low-power, highly integrated design. The key architectural features of Blue Gene/L are introduced in this paper: the link chip component and five Blue Gene/L networks, the PowerPC® 440 core and floating-point enhancements, the on-chip and off-chip distributed memory system, the node- and system-level design for high reliability, and the comprehensive approach to fault isolation.

Originally published:

IBM Journal of Research and Development, Volume 49, Issue 2/3, pp. 195-212 (2005).

Significance:

The Blue Gene/L system was ranked as the top supercomputer in the world at the 2006 International Supercomputer conference, based on the Linpack benchmark. The system achieving this rank was a joint development of IBM and the Department of Energy's National Nuclear Security Administration. In 2005, five of the top 10 supercomputer systems were Blue Gene systems.

The Blue Gene architecture for massively parallel supercomputers was used most famously in the Deep Blue machine which defeated chess champion Gary Kasparov in 1997. The performance of Blue Gene systems has been increased by a factor of over 1000 since then, and these systems have been used in a variety of applications, from modeling turbulence to determining how proteins fold. This paper describes the Blue Gene project objectives and provides a high-level description of the Blue Gene/L architecture.

Comments:


    About IBM    Privacy    Contact