Old News - nthu-ioa/cluster GitHub Wiki

  • 2024-02-04 :warning: Downtime for storage expansion starting 6pm Sunday 2024-02-04 until Monday 02/05. See email for details.

  • 2023-10-06: m03 now has 40TB of local SSD scratch.

  • 2023-10-04: Several nodes have intermittent issues (c20, c22); under investigation.

  • 2023-09-04: The configuration of GPUs and RAM on the GPU nodes has changed. See the specification page. If you notice any unexpected behaviour of GPU codes on g03 in particular, please report this to cica_admin.

  • 2023-08-10: new node c23 available.

  • 2023-01-18: Cluster offline for the whole day, see email. :warning:

  • 2023-01-03: c07 out of service (memory fault, waiting for repair)

  • 2023-01-03: 5 new CPU nodes and 1 new memory node available (see email)

  • 2022-12-23: Temporary change in RAM on m01 (now 1.5TB), m02 (now 768GB)

  • 2022-08-17: performance on c17 may be slightly reduced for intensive jobs.

  • [May 28th 2022] Upgrades: new GPU node g04, Lustre filesystem, scratch space on c01-c04.

  • [April 5th 2022] c05 temporarily unavailable.

  • [May 6th 2022] Cluster offline for power supply work in Tsing Hua Lab

  • [October 16th 2021] c05 temporarily unavailable.

  • [September 14th 2021] c08 back in the queue.

  • [September 10th 2021] Scratch space added to c05-c13, m01, m02

  • [September 7th 2021] c08 out of cpu queue due to hardware issue.

  • [June 22nd 2021] New nodes c14-c17 added to CPU queue.

  • [May 20th 2021] g01 node out of service due to GPU hardware issues.

  • [May 13th 2021] Temporary shutdown during rolling blackouts.

  • [May 10th 2021] Maintenance complete, normal service resumed.