So given the constraints, I believe the above config would increase the delivered aggregate DP flops, and give you the latest compute capability on every workstation, for development consistency. Looking at both GPUs together, I believe the DP flops can be up to approximately double that number: Titan Z gives more than 1.3TF in aggregate, I believe. Then you have 12 “super” workstations that have the Titan Z and deliver most of the aggregate FP64 flops. Every workstation will have the GTX970, so it has the latest compute capability (cc5.2) and a given code targetting the 970 should run anywhere. This will give you some additional “consistency” across the cluster. You might consider putting a GTX 970 in every workstation, and have 12 of those also include the Titan-Z as an “extra” GPU: You simply want an aggregate of 15TF available along with compute capability in every workstation. “but 400 GFlops FP64 is required per workstation” My previous comment was based on this statement you made: Titan Z delivers more than 1.3TF in aggregate, I believe. I admit a number of these are rather ‘raw’ suggestions Lastly, perhaps you can virtualize your cluster/ turn cuda into a service whilst developing You likely do not need a gpu for/ when developing only when compiling/ testing this also implies some degrees of freedom Remote desktop might be a possibility for individual use cases I suppose it might even permit remote compilation directly or indirectly Rather, this should argue for proper (increased) project management and scaling, etcĪ number of hypothetical options come to mind with regards to your “40 console” requirement Some of the 4 slot motherboards are far more economical than the 8 slot motherboardsĮlse, if you have cash to flash, let me not stand in your wayĪlso, one of the titans has triple width, if i am not mistaken this is something to take note of, as the boards generally cater for/ mind double widthĭevelopment time should be a fraction of deployment/ service time hence this should support the argument for a ‘compacted’ cluster: deploying 40 workstations just because initially 40 workstations were needed for development seems illogical, i would think The cost per slots per motherboard moves disproportionally hence, i suggest shopping around (8 x 1.3 TF = 10.4 TF on a single housing, if used Titan Z)” “I understand that motherboards as of today support 8 GPUs
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |