User:Groceryheist/TACC

From CommunityData
Revision as of 05:35, 31 July 2024 by Groceryheist (talk | contribs) (start drafting tacc notes)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

[TACC tacc.utexas.edu] is the high-performance computing center at the University of Texas. It has several different resources. We are likely to use Stampede 3, which is an HPC resource similar to Hyak. Also, Jetstream 2, an open stack cloud, may be useful for data collection and other tasks that are well suited to a virtual webserver.

Stampede 3 complements Hyak in two significant ways. The first is that it is a time-sharing based system instead of a "condo". This means that we have a budget to spend on compute jobs. A major advantage of this model is the ability to run larger jobs that use many nodes. We can start up a spark cluster large enough to fit a dataset in memory. A second advantage is that Stampede 3 has some nodes with GPUs (Intel Ponte Veccio). These GPUs are designed for deep learning and super fast half and single precision floating point calculations. Stampede 3' disadvantage is that jobs have short maximum wall times (24 hrs or 48hrs, depending on the type of node). Jobs that run longer than that need to implement checkpointing.

The budget is renewable, but we have to do increasing amounts of paperwork and come under increasing scrutiny as it grows. If we do good work and are good stewards of the resource it should has several different types of nodes.