00:20:04	Jim Edwards:	No mixed cpu/gpu projects?
00:26:10	Siddhartha Ghosh:	ASD for NCAR:
https://www2.cisl.ucar.edu/user-support/asd-derecho-soliciting-ncar-projects

ASD for University Projects:
https://www2.cisl.ucar.edu/user-support/opportunity-university-researchers-derecho-supercomputer
00:27:10	Cecile Hannay:	Could repeat the dates?
00:27:21	Jeremy Sauer:	Could you post those important start-end dates here in the chat please?
00:27:34	Cecile Hannay:	Thanks.
00:27:35	Siddhartha Ghosh:	Just a reminder AI summer school registration deadline is 22nd of July, here is the link: AI summer school:
https://www2.cisl.ucar.edu/tai4es
00:28:28	Jeremy Sauer:	Thanks
00:29:23	bjsmith:	The Daily Bulletin article Mick mentioned: https://arc.ucar.edu/articles/84
00:29:44	Mick Coady:	HPC systems outage dates:
00:30:39	Mick Coady:	Begin the morning of Monday, July 26.  Full restore Monday, Aug 2.
00:30:56	Craig Schwartz:	Is there going to be a call for NSC proposals this fall, and if so, will those awards be for computing on Cheyenne or Derecho?
00:31:05	Mick Coady:	Please see the related announcement in the Daily Bulletin.  Thanks!
00:32:49	Mick Coady:	Good question, Craig.  Let us ask CISL/NCAR management about that and we will get back to everyone with an answer.
01:08:04	Siddhartha Ghosh:	The test system will have 16 cpu nodes (128 cores) 2 GPU nodes (64 cores + 4 A100 GPUs in each and dual SS connection)
01:08:28	Siddhartha Ghosh:	Cpu nodes will have 256 GB and GPU nodes will have 512 GB memory
01:08:45	Jim Edwards:	When do you anticipate that system will be available?
01:09:06	Jeremy Sauer:	Thanks Sidd, how much memory on each A100?
01:09:19	Siddhartha Ghosh:	Production GPU nodes will have 4 SS connection / GPU nodes compared to test system will have 2 SS connection / GPU node
01:09:34	Supreeth Madapur Suresh:	Thank you. Will this be with all the software stack from cray?
01:10:42	Siddhartha Ghosh:	These are lower men A100 I.e. 40GB
01:10:50	Siddhartha Ghosh:	Both for prod and test
01:11:34	Ned Patton:	Any reasons to use or not use FORTRAN’s  “do concurrent()” for automatic GPU acceleration via nvfortran’s “-stdpar” compilation flag?
01:11:37	Siddhartha Ghosh:	Cray PE will not have full OpenACC / OpenMP support by test system life start
01:11:42	Jeremy Sauer:	Great presentation Daniel.  Many thanks!  Looking forward to some useful resources available through your slides.
01:12:05	cmille73:	Really great Daniel. Thanks for the Nsight doc!