The 2021-22 rates for the Great Lakes Cluster have been approved. These rates represent cost recovery for the Great Lakes Cluster, and do not include any support your unit may choose to provide.
Partition | Rate Per Minute | Rate Per Month | CPU Cores | Memory Unit | GPU Unit |
---|---|---|---|---|---|
standard/debug/viz | $0.000250463 | $10.82 | 1 | 7 gigabytes | N/A |
largemem | $0.000770370 | $33.28 | 1 | 41.75 gigabytes | N/A |
gpu | $0.002739120 | $118.33 | 20 | 90 gigabytes | 1 |
spgpu | $0.001807870 | $78.10 | 4 | 48 | 1 |
The monthly rate is based on a 30-day month; per-minute rates are included for reference. Charges are based on the percentage of the machine your job requests (in terms of the amount of cores, memory, and GPUs), and its actual runtime.
Estimating Maximum Job Costs
The command can estimate the maximum cost of job based on cpu, memory, gpu, and partition selection. The amount returned is the most a job could cost and does not account for any unit or other subsidies you have access to.
my_job_estimate --help my_job_estimate --cores 20 --memory 10gb --partition standard
College of Engineering Funding
The College of Engineering (CoE) cost shares a portion of the use of the Great Lakes cluster for externally-sponsored research conducted by CoE researchers who are paying to use the Great Lakes cluster on CoE shortcodes. For more information, see:
Medical School Funding
The Medical School will cost-share 90% of the Great Lakes Cluster rates for sponsored research accounts for researchers in these units who are paying to use the Great Lakes Cluster on unit Shortcodes.
College of Literature, Science, & the Arts Funding
The College of Literature, Science, and the Arts provides a cost-share of 80% for researchers in College units paying for their use of the Great Lakes cluster on unit shortcodes. LSA researchers who do not have access to any other account may be eligible to use the accounts provided centrally by LSA. The usage policy and restrictions on these accounts is described in detail on the LSA’s public Great Lakes accounts page. Questions about access or use of these accounts should be sent to arc-support@umich.edu.
School of Public Health Funding
The School of Public Health provides a cost-share of 80% for researchers with a primary appointment in the School paying for their use of the Great Lakes cluster on unit shortcodes.
Getting a Slurm Account
To establish a Slurm account for a class please contact us at arc-support@umich.edu with the following information:
- Students to be put on the account
- Shortcode to be used for billing
- List of individuals to administer the account
Any limits to be placed on the either the users or the account as a whole
Please note: courses are currently free of charge through the end of winter term 2020. All students will need to have a user login to use the account. Use the Login Request form to get an account.
Example jobs and their charges¹
To help illustrate how the job charges work, here are some examples of differently-sized jobs.
Partition | Total CPU Cores Used | Total Memory Used | Total GPUs Used | Cost Per Minute |
---|---|---|---|---|
standard | 1 | 1 GB | N/A | $0.000250 |
standard | 1 | 10 GB | N/A | $0.000358 |
standard | 36 | 5 GB | N/A | $0.009017 |
standard | 1 | 50 GB | N/A | $0.001789 |
largemem | 1 | 180 GB | N/A | $.003321 |
gpu | 1 | 20 GB | 1 | $0.002739 |
¹ The charges above have been rounded for readability.
If you have questions, please send email to arc-support@umich.edu.