| 1 | <?xml version='1.0' encoding='UTF-8'?>
|
|---|
| 2 | <!DOCTYPE topic PUBLIC "-//OASIS//DTD DITA Topic//EN" "topic.dtd">
|
|---|
| 3 | <topic xml:lang="en-us" id="ts_host-freq-boost">
|
|---|
| 4 | <title>Performance Variation with Frequency Boosting</title>
|
|---|
| 5 |
|
|---|
| 6 | <body>
|
|---|
| 7 | <p>
|
|---|
| 8 | Many multicore processors support some form of frequency
|
|---|
| 9 | boosting, which means that if only one core is utilized, it can
|
|---|
| 10 | run possibly 50% faster or even more than the rated CPU
|
|---|
| 11 | frequency. This causes measured performance to vary somewhat as
|
|---|
| 12 | a function of the momentary overall system load. The exact
|
|---|
| 13 | behavior depends strongly on the specific processor model.
|
|---|
| 14 | </p>
|
|---|
| 15 | <p> As a consequence, benchmarking on systems which utilize frequency boosting may produce
|
|---|
| 16 | unstable and nonrepeatable results. This is especially true if benchmark runs are short, of
|
|---|
| 17 | the order of seconds. To obtain stable results, benchmarks must be run over longer periods of
|
|---|
| 18 | time and with a constant system load apart from the VM being tested. </p>
|
|---|
| 19 | </body>
|
|---|
| 20 |
|
|---|
| 21 | </topic>
|
|---|
