Cluster Resources
Kamiak is a high-performance computing (HPC) cluster that follows a condominium model in which faculty investors as well as colleges purchase nodes that provide the resources for research computing. The nodes, which are high-powered computers, are grouped into partitions, each owned by a faculty or college, on which the owners receive non-preemptable access. There is also an overarching “backfill” partition that provides open access to idle resources on all nodes for use by the entire WSU research community. Note, however, that jobs submitted to the backfill partition are subject to preemption, i.e., cancel and requeue, by investor or college jobs if they need their nodes in order to be able to run.
Access to each investor or college partition is restricted to users who are associated with the owner of the resources. To gain access to a partition, please contact the owner for approval.
Below we describe the architecture of the backfill, college, and investor partitions.
Definitions:
Max CPU Cores per User: The maximum number of CPU cores across all running jobs of a user. Newly submitted jobs which would exceed this number are automatically queued, pending the completion of the user’s older jobs.
Max Memory per User: The maximum amount of memory across all running jobs of a user. Newly submitted jobs which would exceed this number are automatically queued, pending the completion of the user’s older jobs.
Scheduler Feature Tags: Tags which can be used in job submissions (with –constraint=tag) to specify the node a job should run on.
Backfill Partition
| Queue Name | Node Count | GPUs Total | CPU Cores Total | Memory Total | Wall Clock Limit per Job | Max CPU Cores per User | Max Memory per User |
|---|---|---|---|---|---|---|---|
| kamiak | 138 | 48 | 3860 | 38970 GB | 7-00:00:00 | 120 | 1.5TB |
The following nodes are contained in the “kamiak” backfill partition.
| Node Count | Accelerators per Node | CPU Model | CPU Cores per Node | Memory per Node | Scheduler Feature Tags |
|---|---|---|---|---|---|
| 40 | None | Intel Xeon E5-2660 v3 @ 2.60GHz | 20 | 128GB | haswell, e5-2660-v3-2.60ghz, avx2 |
| 33 | None | Intel Xeon E5-2680 v2 @ 2.80GHz | 20 | 256GB | ivybridge, e5-2680-v2-2.80ghz, avx |
| 10 | None | Intel Xeon E5-2660 v4 @ 2.00GHz | 28 | 128GB | broadwell, e5-2660-v4-2.00ghz, avx2 |
| 8 | 2 NVIDIA Tesla K80 (4 GPUs total) | Intel Xeon E5-2670 v3 @ 2.30GHz | 24 | 256GB | haswell, e5-2670-v3-2.30ghz, avx2 |
| 6 | None | Intel Xeon E5-2660 v3 @ 2.60GHz | 20 | 256GB | haswell, e5-2660-v3-2.60ghz, avx2 |
| 6 | None | Intel Xeon E5-2660 v4 @ 2.00GHz | 28 | 256GB | broadwell,e5-2660-v4-2.00ghz, avx2 |
| 3 | None | Intel Xeon Gold 6230 @ 2.10GHz | 40 | 384GB | cascadelake,gold-6230-2.10ghz,avx-512,connectX-6 |
| 5 | None | Intel Xeon Platinum 8368 CPU @ 2.40GHz | 76 | 1TB | icelake,platinum-8368-2.40ghz,avx-512,connectX-6,mem1TB,cores76 |
| 4 | None | Intel Xeon Gold 6230 @ 2.10GHz | 40 | 192GB | cascadelake,gold-6230-2.10ghz,avx-512,connectX-6 |
| 3 | None | Intel Xeon E5-2660 v4 @ 2.00GHz | 28 | 512GB | broadwell, e5-2660-v4-2.00ghz, avx2 |
| 3 | None | Intel Xeon Gold 6338 CPU @ 2.00GHz | 64 | 512GB | icelake,gold-6338-2.00ghz,avx-512,connectX-6,mem512GB,cores64 |
| 2 | None | Intel Xeon Gold 6138 @ 2.00GHz | 40 | 384GB | skylake, gold-6138-2.00ghz, avx2 |
| 2 | None | Intel Xeon Silver 4116 @2.10GHz | 24 | 192GB | skylake,silver-4116-2.10ghz,avx-512,connectX-5 |
| 1 | None | Intel Xeon Gold 6230 @ 2.10GHz | 40 | 384GB | cascadelake,gold-6230-2.10ghz,avx-512,connectX-6,nvme |
| 2 | None | Intel Xeon Gold 6138 @ 2.00GHz | 40 | 384GB | skylake,gold-6138-2.00ghz,avx-512,connectX-5 |
| 2 | None | Intel Xeon Gold 6338 @ 2.00GHz | 64 | 512GB | icelake,platinum-8368-2.40ghz,avx-512,connectX-6 |
| 1 | None | Intel Xeon E7-4880 v2 @ 2.50GHz | 60 | 2TB | ivybridge, e7-4880-v2-2.50ghz, avx |
| 1 | 4 NVIDIA Tesla K80 (8 GPUs total) | Intel Xeon E5-2670 v3 @ 2.30GHz | 24 | 256GB | haswell, e5-2670-v3-2.30ghz, avx2 |
| 1 | 2 NVIDIA Tesla K80 (4 GPUs total) | Intel Xeon E5-2660 v4 @ 2.00GHz | 28 | 256GB | broadwell, e5-2660-v4-2.00ghz, avx2 |
| 1 | Intel Xeon Phi coprocessor | Intel Xeon E5-2670 v3 @ 2.30GHz | 24 | 256GB | haswell, e5-2670-v3-2.30ghz, avx2 |
| 1 | None | Intel Xeon Silver 4116 @ 2.10GHz | 24 | 192GB | skylake, silver-4116-2.10ghz, avx2 |
| 1 | None | Intel Xeon Gold 6138 @ 2.00GHz | 40 | 384GB | skylake,gold-6138-2.00ghz,avx-512,connectX-5,nvme |
| 1 | 1 NVIDIA V100 Tensor Core CPU | Intel Xeon Gold 6138 @ 2.00GHz | 40 | 192GB | skylake,gold-6138-2.00ghz,avx-512,v100,volta |
| 1 | None | Intel Xeon Gold 6338 CPU @ 2.00GHz | 64 | 1TB | icelake,gold-6338-2.00ghz,avx-512,connectX-6,mem1TB,cores64 |
| 1 | 4 NVIDIA A100 | Intel Xeon Platinum 8368 CPU @ 2.40GHz | 76 | 2TB | icelake,platinum-8368-2.40ghz,avx-512,connectX-6,a100,ampere,mem2TB,cores76 |
College Partitions
| Queue Name | Node Count | CPU Cores per Node | Memory per Node | Accelerators per Node | Wall Clock Limit per Job | Max CPU Cores per User | Max Memory per User |
|---|---|---|---|---|---|---|---|
| cahnrs | 11 | 20 | 256GB | 7-00:00:00 | 120 | 1.5TB | |
| cahnrs_bigmem | 1 | 60 | 2TB | 7-00:00:00 | |||
| cahnrs_gpu | 1 | 24 | 256GB | 2 NVIDIA Tesla K80 (4 GPUs total) | 7-00:00:00 | ||
| cas | 10 | 20 | 256GB | 7-00:00:00 | |||
| cas | 5 | 76 | 1TB | 7-00:00:00 | |||
| cas | 1 | 76 | 2TB | 4 NVIDIA A100 | 7-00:00:00 | ||
| cas | 1 | 64 | 1TB | 7-00:00:00 | |||
| coe | 1 | 40 | 384GB | 7-00:00:00 | |||
| free | 3 | 24 | 256GB | 2 NVIDIA Tesla K80 (4 GPUs total) | 7-00:00:00 | ||
| vcea | 3 | 20 | 256GB | 7-00:00:00 |
Investor Partitions
| Queue Name | Node Count | CPU Cores per Node | Memory per Node | Accelerators per Node | Wall Clock Limit per Job | Max CPU Cores per User | Max Memory per User |
|---|---|---|---|---|---|---|---|
| adam | 1 | 40 | 384GB | 14-00:00:00 | |||
| adam | 2 | 28 | 128GB | 14-00:00:00 | |||
| awn | 1 | 40 | 384GB | 7-00:00:00 | |||
| beckman | 35 | 20 | 128GB | 7-00:00:00 | |||
| catalysis_gpu | 1 | 24 | 256GB | 2 NVIDIA Tesla K80 (4 GPUs total) | 7-00:00:00 | ||
| catalysis_long | 3 | 20 | 256GB | 7-00:00:00 | |||
| catalysis_long | 5 | 40 | 384GB | 7-00:00:00 | |||
| clark | 1 | 28 | 256GB | 7-00:00:00 | |||
| clark | 3 | 20 | 128GB | 7-00:00:00 | |||
| clark | 1 | 64 | 512GB | 7-00:00:00 | |||
| cook | 1 | 24 | 256GB | 4 NVIDIA Tesla K80 (8 GPUs total) | 7-00:00:00 | ||
| fernandez | 1 | 40 | 192GB | 7-00:00:00 | |||
| ficklin | 1 | 40 | 192GB | 1 NVIDIA V100 Tensor Core GPU | 30-00:00:00 | ||
| ficklin | 5 | 24 | 256GB | 2 NVIDIA Tesla K80 (4 GPUs total) | 30-00:00:00 | ||
| hipps | 5 | 248 | 2176GB | 7-00:00:00 | |||
| hpc_club | 1 | 28 | 256GB | 2 NVIDIA Tesla K80 (4 GPUs total) | 7-00:00:00 | ||
| hpc_club | 3 | 28 | 256GB | 7-00:00:00 | |||
| katz | 1 | 20 | 256GB | 7-00:00:00 | |||
| lee | 1 | 28 | 256GB | 7-00:00:00 | |||
| lee | 2 | 28 | 128GB | 7-00:00:00 | |||
| lofgren | 1 | 20 | 128GB | 14-00:00:00 | |||
| lofgren | 2 | 24 | 192GB | 14-00:00:00 | |||
| mainlab | 2 | 20 | 256GB | 7-00:00:00 | |||
| neibergs | 1 | 40 | 192GB | 7-00:00:00 | |||
| pacbio | 3 | 40 | 384GB | 7-00:00:00 | |||
| pddms | 5 | 28 | 128GB | 7-00:00:00 | |||
| peters | 1 | 28 | 256GB | 7-00:00:00 | |||
| peters | 1 | 28 | 512GB | 7-00:00:00 | |||
| popgenom | 3 | 20 | 256GB | 14-00:00:00 | |||
| rajagopalan | 1 | 40 | 384GB | 7-00:00:00 | |||
| ssl | 2 | 20 | 128GB | 7-00:00:00 | |||
| stockle | 2 | 20 | 256GB | 7-00:00:00 | |||
| storfer | 2 | 28 | 512GB | 7-00:00:00 | |||
| tanner | 1 | 24 | 192GB | 7-00:00:00 |