User Tools

Site Tools


cluster:167

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
cluster:167 [2018/06/27 18:33]
hmeij07 [CPU vs GPU]
cluster:167 [2018/06/28 13:56]
hmeij07
Line 10: Line 10:
   * That period covers 600 hours of time   * That period covers 600 hours of time
   * Assume 99% utilization of cpu core or gpu device   * Assume 99% utilization of cpu core or gpu device
-  * Available time is measured per physical cpu core but by gpu device+  * Available Hours is measured per physical CPU core but by GPU device (exclusivity and persistence modes on)
   * There is no good/bad metric   * There is no good/bad metric
   * Never collated such data before   * Never collated such data before
Line 24: Line 24:
 | Avail Hours | 715,200  |  50:1  |  14,400 | total cpu cores, total gpus | | Avail Hours | 715,200  |  50:1  |  14,400 | total cpu cores, total gpus |
 | Job Hours | 221,136  |  77:1  |  2,872 | cumulative hours of consumed usage | | Job Hours | 221,136  |  77:1  |  2,872 | cumulative hours of consumed usage |
-| Job Hours % | 31  |  6:1  |  5 | as a percentage...weeping... |+| Job Hours % | 31  |  6:1  |  5 | as a percentage |
 | Avail Hours2 | 561,600  |  39:1  |  14,400 | total cpu cores - hp12's 256 cores, total gpus | | Avail Hours2 | 561,600  |  39:1  |  14,400 | total cpu cores - hp12's 256 cores, total gpus |
 | Job Hours % | 39  |  8:1  |  5 | more realistic...hp12 rarely used in June18| | Job Hours % | 39  |  8:1  |  5 | more realistic...hp12 rarely used in June18|
Line 41: Line 41:
 | Memory | 7,408  |  74:1  |  100 | GB | | Memory | 7,408  |  74:1  |  100 | GB |
 | Teraflops | 38  |  1.7:1  |  23 | double precision, floating point, theoretical | | Teraflops | 38  |  1.7:1  |  23 | double precision, floating point, theoretical |
-| Job Count |   |  :  |   | scheduled jobs irregardless of exit status |+| Job Count | 12,798  |  18:  722 | scheduled jobs irregardless of exit status |
 | Avail Hours | 886,848  |  60:1  |  14,880 | total cpu cores, total gpus | | Avail Hours | 886,848  |  60:1  |  14,880 | total cpu cores, total gpus |
-| Job Hours |   |  :  |   | cumulative hours of consumed usage | +| Job Hours |  260,997  |  69:  3,805 | cumulative hours of consumed usage | 
-| Job Hours % |   |  :  |   | as a percentage...weeping... |+| Job Hours % | 30  |  1:  26 | as a percentage |
 | Avail Hours2 | 696,384  |  47:1  |  14,880 | total cpu cores - hp12's 256 cores, total gpus | | Avail Hours2 | 696,384  |  47:1  |  14,880 | total cpu cores - hp12's 256 cores, total gpus |
-| Job Hours % |   |  :  |   | more realistic...hp12 rarely used in June18|+| Job Hours % | 37  |  1.5:  26 | more realistic...hp12 rarely used in June18|
  
-  * Some noise in this data with the inability to match start and end of job (1-5% of records)+  * Some noise in this data with the inability to match start and end of job (~15% of records)
   * The assumption that ''hp12'' was barely used might not be correct   * The assumption that ''hp12'' was barely used might not be correct
-  * + 
 +Based on Jul17 we process about 60-70 times more CPU jobs than GPU jobs, that seems consistent with Jul18. The metric of Job Hours consumed versus Available Hours in %, the picture is probably more like Jul17...30-40% of CPU cycles are consumed and 25% of GPU cycles. We shall wait for Jul18 metrics. 
 + 
  
 **[[cluster:0|Back]]** **[[cluster:0|Back]]**
cluster/167.txt · Last modified: 2018/08/01 14:11 by hmeij07