High-Density Computing: A 240-Processor Beowulf in One Cubic Meter

We present results from computations on Green Destiny, a 240-processor Beowulf cluster which is contained entirely within a single 19-inch wide 42U rack. The cluster consists of 240 Transmeta TM5600 667-MHz CPUs mounted on RLX Technologies motherboard blades. The blades are mounted side-by-side in an RLX 3U rack-mount chassis, which holds 24 blades. The overall cluster contains 10 chassis and associated Fast and Gigabit Ethernet switches. The system has a footprint of 0.5 meter 2 (6 square feet), a volume of 0.85 meter 3 (30 cubic feet) and a measured power dissipation under load of 5200 watts (including network switches). We have measured the performance of the cluster using a gravitational treecode N-body simulation of galaxy formation using 200 million particles, which sustained an average of 38.9 Gflops on 212 nodes of the system. We also present results from a three-dimensional hydrodynamic simulation of a core-collapse supernova.

[1]  Leslie Greengard,et al.  A fast algorithm for particle simulations , 1987 .

[2]  Alan H. Karp.  Speeding up N-body Calculations on Machines without Hardware Square Root , 1992, Sci. Program..

[3]  Michael S. Warren,et al.  Astrophysical N-body simulations using hierarchical tree data structures , 1992, Proceedings Supercomputing '92.

[4]  Thomas L. Sterling,et al.  BEOWULF: A Parallel Workstation for Scientific Computation , 1995, ICPP.

[5]  David M. Beazley,et al.  Avalon: an Alpha/Linux cluster achieves 10 Gflops for $15k , 1998, SC '98.

[6]  Michael S. Warren,et al.  A portable parallel particle program , 1995 .

[7]  Thomas L. Sterling,et al.  Pentium Pro Inside: I. A Treecode at 430 Gigaflops on ASCI Red, II. Price/Performance of $50/Mflop on Loki and Hyglac , 1997, ACM/IEEE SC 1997 Conference (SC'97).

[8]  John K. Salmon,et al.  Large-scale structure after COBE: Peculiar velocities and correlations of cold dark matter halos , 1994 .

[9]  Michael S. Warren,et al.  The Fast Solution of Three-Dimensional Fluid Dynamical N-Body Problems Using Parallel Tree Codes: Vortex Element Method and Boundary Element Method , 1995, PPSC.

[10]  Wu-chun Feng,et al.  Honey, I shrunk the Beowulf! , 2002, Proceedings International Conference on Parallel Processing.

[11]  Michael S. Warren,et al.  Skeletons from the treecode closet , 1994 .

[12]  Michael S. Warren,et al.  1 3 J un 2 00 2 Modeling Core-Collapse Supernovae in 3-Dimensions , 2002 .

[13]  Piet Hut,et al.  A hierarchical O(N log N) force-calculation algorithm , 1986, Nature.

[14]  M. S. Warren,et al.  A parallel hashed Oct-Tree N-body algorithm , 1993, Supercomputing '93.

[15]  Michael S. Warren,et al.  Dark halos formed via dissipationless collapse. I: Shapes and alignment of angular momentum , 1992 .

[16]  Michael S. Warren,et al.  Using Parallel Computers for Very Large N-Body Simulations: Shell Formation Using 180 K Particles , 1990 .

[17]  Michael S. Warren,et al.  Fast Parallel Tree Codes for Gravitational and Fluid Dynamical N-Body Problems , 1994, Int. J. High Perform. Comput. Appl..

[18]  Gordon Bell,et al.  High Performance Computing: Crays, Clusters, and Centers. What Next? , 2001 .