- About
- Resources
- PROJECTS
- Training
- NEWS
- User Support
- User Support
- Accounts, Allocations, Data Storage, Documentation
- Create an Account
- PI request to add a user
- Request over 10,000 Service Units
- Allocation Proposal Guidelines
- Data Storage Guidelines
- Hardware Investing Policy
- Resource Documentation
- Installed Software
- Introduction to Linux
- Advanced HTC Support
- People
- CONTACT
Service Units
One service unit (SU) is approximately equal to 1 core hour of computing. The charge is calculated based on 2 factors:
- Number of cores requested
- RAM requested
We take the maximum charge of the 2 factors. If one requests the default RAM (Requested Cores * Total RAM / Total Cores) you are charged the exact same amount for cores and RAM. Additionally, each of the partitions (see Clusters and Partitions) have a scaling factor. The scaling factor compensates for the cost of the hardware itself. Some examples (using a scale factor of 1):
- 12 core job, default RAM, 2 hours: 24 SUs
- 1 core job, all RAM on 12 core machine, 4 hours: 48 SUs
Below, I will break down H2P in clusters (numbered) and partitions (lettered) and provide the scaling factors.
- smp [default] (Shared Multiprocessing): Single node (OpenMP) and single core jobs
- smp [default]: 0.8
- smp-high: 1
- gpu (Graphics Processing Units): Single- or multi-GPU processing
- gtx1080 [default]: 1 per GPU card
- titanx: 3 per GPU card
- titan: 0.333 per GPU card
- k40: 6 per GPU card
- mpi (Message Passing Interface): Multi-Node Jobs
- opa [default]: 1
- legacy: 0.333
- ib: 1
You can simply look at this information by running scontrol -M smp,mpi,gpu show partition.