Note: The clusters receive regular maintenance and upgrades, so at times these CPU and GPU numbers may lag behind the latest changes. To see the most up-to-date information, ssh into your cluster of interest and use the "snodes" command.
|
Cluster |
Processor Speed |
Nodes | Cores per Node |
Memory per Node |
Total Cores | Interconnect | Theoretical FP64 Performance |
|---|---|---|---|---|---|---|---|
| Nobel Dell Linux Server |
2.6 GHz Intel Ice Lake | 2 | 64 | 1 TB | 128 | N/A | 10.6 TFLOPS |
| Adroit Dell Linux Cluster |
2.6 GHz Intel Skylake | 9 | 32 | 384 GB | 288 | FDR Infiniband | 24 TFLOPS |
| Adroit (adroit-h11g1) GPU node |
2.6 GHz Intel Skylake |
1 |
40 4 GPU/node |
770 GB 32 GB/GPU |
40 | N/A | 3.3 TFLOPS 28 TFLOPS |
| Adroit (adroit-h11g2) GPU node |
2.8 GHz Intel Ice Lake 1.41 GHz A100 |
1 | 48 4 GPU/node |
1 TB 40 GB/GPU |
48 | N/A | 4.3 TFLOPS 39 TFLOPS |
| Adroit (adroit-h11g3) GPU node |
2.6 GHz Intel Skylake |
1 |
56 4 GPU/node |
770 GB 32 GB/GPU |
56 | N/A | 3.3 TFLOPS 28 TFLOPS |
| Tigressdata Dell Linux Server |
2.4 GHz Gold 6148 Xeon |
1 | 40 | 768 GB | 20 | N/A | 0.4 TFLOPS |
| Della Dell Linux Cluster |
2.4 GHz Intel Broadwell |
96 |
28 |
128 GB |
6784 | FDR/EDR Infiniband | 486 TFLOPS |
| Della GPU nodes |
2.8 GHz Intel Ice Lake 1.41 GHz A100 |
70 |
48 4 GPU/node |
1 TB 80 GB/GPU |
3360 | HDR100 Infiniband | 301 TFLOPS 2716 TFLOPS |
| Della GPU nodes |
2.6 GHz AMD EPYC Rome 1.41 GHz A100 |
20 |
128 2 GPU/node |
768 GB 40 GB/GPU |
2560 | HDR100 Infiniband | 106 TFLOPS 388 TFLOPS |
| TigerCPU HPE Linux Cluster |
2.4 GHz Intel Skylake | 408 | 40 |
192 GB (40 w/768 GB) |
16320 | Omnipath | >1103 TFLOPS |
| Stellar Intel nodes |
2.9 GHz Intel Cascade Lake | 296 | 96 | 1 TB | 28416 | HDR100 Infiniband | 2600 TFLOPS |
| Stellar AMD nodes |
2.6 GHz AMD EPYC Rome | 187 | 128 | 512 GB | 23936 | HDR100 Infiniband | 996 TFLOPS |
| Stellar GPU nodes |
2.6 GHz AMD EPYC Rome 1.41 GHz A100 |
6 |
128 2 GPU/node |
500 GB 40 GB/GPU |
768 | HDR100 Infiniband | 32 TFLOPS 116 TFLOPS |
| Traverse IBM Linux Cluster |
2.7 GHz IBM POWER9 1530 MHz V100 GPU |
46 | 128 4 GPU/node
|
256 GB 32 GB/GPU |
8960 184 GPUs |
EDR Infiniband | 1435 TFLOPS |
The A100 has a FP64 performance of 9.7 TFLOPS while the V100 (PCIe) provides 7 TFLOPS and V100 (NVLink) gives 7.8 TFLOPS.