The super-computing power of several teraflops is now with Tesla™ GPUs
in small clusters reachable. This option offers MEGWARE now completely
preconfigured cluster based on the NVIDIA® Tesla™ S1070 GPU computing
system, with four T10 Tesla™ GPU performance of 4 teraflops and
achieved only one height unit (1U) needed in the rack.
Each GPU is based on innovative parallel NVIDIA® CUDA™ processor
architecture and has more than 240 parallel processing units. In
combination with server processors from Intel or AMD are obtained now a
32-Teraflop and a 64-Teraflop cluster by MEGWARE. In addition to the
massively parallel computing power these systems are energy- and
cost-saving because they have less electric power consumption and small
cooling. MEGWARE provides these two systems, like all other cluster,
ready with preinstalled software - directly ready for use.
More information »