Details
-
Enabler
-
Should have
-
None
-
None
-
Services
-
-
-
4
-
4
-
2
-
Team_PLANET
-
Sprint 5
-
-
Overdue
Description
As an SDP hardware designer and an SDP software architect, I require insight into the detailed compute profile of SDP workloads.
The pipeline selected should be tailored to the type of investigations that will be done. For example RASCIL may be a good choice for runs that test scalability since that is one of the main things it tries to tackle. For tests on single node WSCLEAN may be a good choice. These two (and most other) astronomy packages are run routinely on HPC-type facilities so packaging should be reasonably straightforward.
There are many things that a framework such as this could be testing, so it is useful to have some end-goals in mind before getting going, e.g., "is scalability of I/O a potential bottleneck ?".
Attachments
Issue Links
- relates to
-
SP-1692 Establish test and benchmarking environment for SDP pipelines
- Done
-
SP-1693 Onboard and test complete set continuum imaging pipelines using simulation data (without DD effects)
- Done
- mentioned in
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...