The HPC Linux computing resource consists of a total 16 compute nodes in one cluster. The charts below describe the various Linux computing partitions available on HPC, their names, and information on the partitions’ computing power.
These are general nodes available to all HPC researchers:
Partition Names | Node Range | # of Nodes | Mem size | Cores per Node | CPU Speed GHz | CPU Type | GPUs per Node | /tmp Size |
---|---|---|---|---|---|---|---|---|
low | node1 ~ node6 | 6 | 125 G | 24 | 2.5 | Xeon | - | 50 G |
These are personal nodes available to partial HPC researchers:
Partition Names | Node Range | # of Nodes | Mem size | Cores per Node | CPU Speed GHz | CPU Type | GPUs per Node | /tmp Size |
---|---|---|---|---|---|---|---|---|
high | node7 ~ node8 | 2 | 125 G | 24 | 2.5 | Xeon | - | 50 G |
Addnode | node9 ~ node14 | 6 | 125 G | 24 | 2.5 | Xeon | - | 50 G |
zhao | node15 ~ node16 | 2 | 125 G | 24 (node15) 28 (node16) |
2.5 | Xeon | - | 50 G |
Students could have the permission of additional nodes if they’re belong to teachers mentioned below.
high: Yan Yin; Addnode: Xiaoyan Ma, Chao Liu; zhao: Tianliang Zhao
Some nodes are privately owned and are not available to all users.
In addition, nodes are sometimes reserved for specific projects or purposes and may be temporarily unavailable.