More TFLOPS, Fewer WATTS Microway delivers the fastest and greenest floating point throughput in history 2. 5 TFLOPS
Enhanced GPU Computing with Tesla Fermi
480 Core NVIDIA® Tesla™ Fermi GPUs deliver 1.2 TFLOP
single precision & 600 GFLOP double precision performance!
New Tesla C2050 adds 3GB ECC protected memory
New Tesla C2070 adds 6GB ECC protected memory
Tesla Pre-Configured Clusters with S2070 4 GPU servers
WhisperStation - PSC with up to 4 Fermi GPUs
OctoPuter™ with up to 8 Fermi GPUs and 144GB memory
New Processors
; 12 Core AMD Opterons with quad channel DDR3 memory
8 Core Intel Xeons with quad channel DDR3 memory
Superior bandwidth with faster, wider CPU memory busses
Increased efficiency for memory-bound floating point algorithms
Configure your next Cluster today!
www.microway.com/quickquote
508-746-7341
45 TFLOPS
Fas Tree 864 GB/sec
Bi-sectional Bandwidth
Fas Tree™ QDR InfiniBand Switches and HCAs 36 Port, 40 Gb/s, Low Cost Fabrics Compact, Scalable, Modular Architecture Ideal for Building Expandable Clusters and Fabrics MPI Link-Checker™ and InfiniScope™ Network Diagnostics Achieve the Optimal Fabric Design for your Specific MPI Application with ProSim™ Fabric Simulator Now you can observe the real time communication coherency of your algorithms. Use this information to evaluate whether your codes have the potential to suffer from congestion. Feeding observed data into our IB fabric queuing-theory simulator lets you examine latency and bi-sectional bandwidth tradeoffs in fabric topologies.
10 TFLOPS
5 TFLOPS
GSA Schedule
Contract Number:
GS-35F-0431N
