We invite you to shape the future of IBM, including product roadmaps, by submitting ideas that matter to you the most. Here's how it works:
Post your ideas
Post ideas and requests to enhance a product or service. Take a look at ideas others have posted and upvote them if they matter to you,
Post an idea
Upvote ideas that matter most to you
Get feedback from the IBM team to refine your idea
Help IBM prioritize your ideas and requests
The IBM team may need your help to refine the ideas so they may ask for more information or feedback. The product management team will then decide if they can begin working on your idea. If they can start during the next development cycle, they will put the idea on the priority list. Each team at IBM works on a different schedule, where some ideas can be implemented right away, others may be placed on a different schedule.
Receive notification on the decision
Some ideas can be implemented at IBM, while others may not fit within the development plans for the product. In either case, the team will let you know as soon as possible. In some cases, we may be able to find alternatives for ideas which cannot be implemented in a reasonable time.
Extend SessionScheduler being able to manage small parallel application (size up to one node)
Today we have applications/software running in serial mode via session scheduler (runtime per task <<10min) with up to 10000 tasks. These applications/software will migrate to parallel usage soon and we want to keep the SessionScheduler func...
Currently Guaranteed Resource Pools in LSF can fulfill SLAs based on slots, memory, licenses, or packages (slots + memory). With more demand for GPU resources in a cluster, it would be beneficial to add GPU resources to a Guaranteed Resource Pool....
RTM GPU Wall time statistic is limited to "only Finished Exclusive job has GPU wall time reported and it based on DCGM".We would like to see statistic for GPU wall time usage regardless job exist status and GPU mode. Also because GPUs are occupied...
The number of TCP exceeds the max number 2*200 defined in LSF source code
We're getting the following message from master lim.log.doAcceptConn: Can't maintain this big clientMap, Connection from 10.230.236.102:36587 droppedAnd I received the following explanation and advice through case number TS003933691.It looks ...
LSF provides integration with DCGM so that GPU usage data can be collected in lsb.acct log file. However, the latest LSF only supports DCGM v1.7.2, which goes with CUDA10.For customs like us, we used CUDA11, which needs DCGM v2.1.4, and we would l...
TSMC is using MC to balance work load between clusters. When the submission cluster is busy, it forwards some jobs to execution clusters. The execution cluster has it's own workload except the forwarded jobs. To avoid the forwarded jobs occupy mos...
Do not place IBM confidential, company confidential, or personal information into any field.