High Performance Computing
ACM-Chem Cluster is a 272 core parallel computational cluster at the University of South Carolina for use by researchers at USC and especially the College of Art and Science. The ACM-Chem cluster was built using resources from the start-up funds of Dr. Sophya Garaschuk (Chemistry) and Dr. Qi Wang (Mathematics), NSF EPSCOR Track II bridge grant, and the NanoCenter fund.
|Number of Processors||272 (compute)|
|Total Memory||1056 TB (DDR2)|
|Peak Performance||2.5 TFLOPS|
|Total Disk||11 TB (global)|
For more information about using ACM-CHEM, see the ACM-CHEM user guide.
Planck Cluster combines a mixture 264 cpu cores(Intel Xeon Westmere 2.8 Ghz) and and 57 M2070/M1060 Nvidia GPGPU accelerators boards to form a hydbrid supercomputer with a theoretical peak performance of 59 Teraflops. The cluster was aquired using NSF EPSCOR Track I funds.
|Number of CPU Processors||228 (compute)|
|Number of GPU Processors|| 48 NVIDIA M1060 (11520 CUDA cores)|
9 NVIDIA M2070 (4032 CUDA cores)
4 NVIDIA M2090 (2048 CUDA cores)
|Total Memory||528 GB (DDR3)|
|CPU Peak Performanc||e 5.9 TFLOPS (double precision)|
|GPU Peak Performance|| 11 TFLOPS (double precision)|
61 TFLOPS (single precision)
|Total Peak Performance||16.9 TFLOPS (double precision)|
|Total Disk||11 TB (shared)|
For more information about using Planck, see the Planck user guide.
Pople Cluster has 204 cpu cores (Intel Xeon Westmere 2.8 Ghz) with a theoretical peak performance of 4.9 Teraflops double precision. Pople is an educational shared resource that permits research, research training, and research outreach opportunities for USC and neighboring colleges such as Claflin University, USC Aiken, Newberry College, Allen University and Presbyterian College. The cluster was aquired using NSF CHE funds. Researchers should acknowledge "USC HPC clusters" in publications.
|Number of Processors||204 (compute)|
|Total Memory||408 GB|
|Peak Performance||4.9 TFLOPS|
|Total Disk||8 TB (shared)|