Total Tyranny of low utilization servers, part 2
Friday Nov 17, 2006
Intel supporters always want to say woodcrest servers are better at low utilisation. This is redirection and will get you looking at doing the wrong things. Having kinds of techniques to power at lower utilisations, won't in the end save much money.
Previous blog entry showed 400 watt server, this table shows a 300 watt server that saves 20 watts for each 10% reduction in utilisation, compared to 100% utilisation. A more realistic comparison is below. table.
| %Utilisation | 100% | 90% | 80% | 70% | 60% | 50% | 40% | 30% | 20% | 10% | 0% |
| Watts-at-Util | 300 | 280 | 260 | 240 | 220 | 200 | 180 | 160 | 140 | 120 | 120 |
| watts/work | 300 | 311 | 325 | 343 | 367 | 400 | 450 | 533 | 700 | 1200 | infinity |
Here is a more typical example. Lets compare 5 servers running at 10% utilisation (that is 120 watts each or 600 watts for the 5 of them). A single server running at 50% utilisation only uses 200 watts, therefore the 10% case almost require 3 times more power!
Let's say you didn't save any power at 50% utilisation that you burned a whole 300 watts, in this case the 10% case would still require twice the power! still, OUCH.
Bottom line: It is far too easy to be fooled to think you are saving money if power-saving features at low utilisation is your answer.
|
|
|
|
