LCC system overview
The Lipscomb Compute Cluster (LCC) is a high-performance computing system used for large-scale scientific workloads. The system provides CPU and GPU compute resources connected through high-speed networking and shared storage.
Admin Nodes (motherships)
Administrative nodes manage cluster services such as scheduling, monitoring, and system management.
Intel Processor Number | Processor Class | Cores per node | Nodes In Cluster | Memory per node (GB) | Network | Node Names |
Intel(R) Xeon(R) CPU E5-2670 v3 @ 2.30GHz | Haswell | 24 | 2 | 128 | Infiniband EDR(100Gbps) | mothership[1-2] |
Login Nodes
Login nodes provide access points for users to log in, compile software, and submit jobs to the scheduler.
Intel Processor Number | Processor Class | Cores per node | Nodes In Cluster | Memory per node (GB) | Network | Node Names |
E5-2670 | Haswell | 24 | 6 | 128 | Infiniband EDR(100Gbps) | login[001-006] |
Compute Nodes
Compute nodes perform the primary scientific workloads on the cluster. Different node types are available to support CPU workloads, GPU acceleration, and specialized compute tasks.
Node Type | Intel Processor Number | Processor Class | Cores per node | Nodes In Cluster | Total Cores in Cluster | Memory per node (GB) | GPU Type | Total GPUs | GPU RAM | Network | Node Names |
Skylake Nodes | 6130 | Skylake | 32 | 56 | 1,792 | 192 |
|
|
| Infiniband EDR(100Gbps) | skylake[001-056] |
SKYLAKE with NVIDIA P100 cards | 6130 | Skylake | 32 | 2 | 64 | 192 | P100 | 8 | 16GB | Infiniband EDR(100Gbps) | gpdnode[001-002] |
SKYLAKE with NVIDIA P100 cards | 6130 | Skylake | 32 | 10 | 320 | 192 | P100 | 40 | 12GB | Infiniband EDR(100Gbps) | gphnode[001-010] |
SKYLAKE with NVIDIA V100 cards | 6130 | Skylake | 32 | 6 | 192 | 192 | V100 | 24 | 32GB | Infiniband EDR(100Gbps) | gvnode[001-006] |
CASCADE Nodes | 6252 | Cascade | 48 | 52 | 2496 | 192 |
|
|
| Infiniband EDR/2(50Gbps) | cascade[001-052] |
CASCADE Nodes | 6252 | Cascade | 48 | 60 | 2880 | 192 |
|
|
| Infiniband EDR(100Gbps) | cascadeb[001-060] |
CASCADE with NVIDIA V100 cards | 6230 | CASCADE | 40 | 12 | 240 | 192 | V100 | 48 | 32GB | Infiniband EDR(100Gbps) | gvnodeb[001-012] |
Icelake with NVIDIA A100 cards | 6330 | Icelake | 56 | 4 | 224 | 256 | A100 | 8 | 80GB | Infiniband EDR(100Gbps) | ganode[001-004] |
Sapphire Rapids with NVIDIA H200 cards | 8480 | Sapphire Rapids | 112 | 2 | 224 | 2048 | H200 | 16 | 144GB | Infiniband EDR(100Gbps) | ghnode[001-002] |
Data Transfer Node
Data transfer nodes are used for moving data into and out of the cluster efficiently.
Node Type | Intel Processor Number | Processor Class | Cores per node | Nodes In Cluster | Total Cores in Cluster | Memory per node (GB) | Network | Node Names |
DTN Node | 6152 | Cascade | 44 | 1 | 44 | 192 | Ethernet (40 Gbps external) EDR(100 Gbps Internal) | dtn |
Storage
The LCC cluster uses the Lenovo GPFS parallel filesystem for shared storage across compute nodes.
Lenovo GPFS (DSS-G) parallel file system -1: 1.3PB Usable (1.9PB RAW)
Lenovo GPFS (DSS-G) parallel file system -2: 1.6PB Usable (2.2PB RAW)
Related Information
For details on SLURM queue names see the SLURM Queues page.