Skip to content

Latest commit

 

History

History
70 lines (46 loc) · 3.19 KB

2022-08-31-Technology-Refresh.md

File metadata and controls

70 lines (46 loc) · 3.19 KB
title author publish_on type canonical_url image description excerpt card_src card_alt
Technology Refresh
Christina Koch
chtc
news
Thanks to the generous support of the Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation, CHTC has been able to execute a major refresh of hardware. This provided 207 new servers for our systems, representing over 40,000 batch slots of computing capacity.
Thanks to the generous support of the Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation, CHTC has been able to execute a major refresh of hardware. This provided 207 new servers for our systems, representing over 40,000 batch slots of computing capacity.
Image of Servers

Thanks to the generous support of the Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation, CHTC has been able to execute a major refresh of hardware. This provided 207 new servers for our systems, representing over 40,000 batch slots of computing capacity. Most of this hardware arrived over the summer and we have started adding them to CHTC systems.

Continue reading to learn more about the types of servers we are adding and how to access them.

HTC System

On the HTC system, we are adding 167 servers of new capacity, representing 36,352 job slots and 40 high-end GPU cards.

The new servers will be running CentOS Linux 8 – CHTC users should see our website page about how to test your jobs and take advantage of servers running CentOS Stream 8. Details on user actions needed for this change can be found on the OS transition page.

New Server specs

PowerEdge ​R6525
  • 157 servers with 128 cores / 256 job slots using the AMD Epyc 7763 processor
  • 512 GB RAM per server
PowerEdge XE8545
  • 10 servers, each with four A100 SXM4 80GB GPU cards
  • 128 cores per server
  • 512GB RAM per server

HPC Cluster

For the HPC cluster, we are adding 40 servers representing 5,120 cores. These servers have arrived but have not yet been added to the HPC cluster. In most cases, when we add them, they will form a new partition and displace some of our oldest servers, currently in the “univ2” partition.

New server specs:

Dell Poweredge R6525
  • 128 cores using the AMD Epyc 7763 processor
  • 512GB of memory

Users interested in early access to AMD processors before all 40 servers are installed should contact CHTC at [email protected].

We have also obtained hardware and network infrastructure to completely replace the HPC cluster’s underlying file system and infiniband network fabric. We will be sending more updates to the chtc-users mailing list as we schedule specific transition dates for these major cluster components.