github-jwinterm/xmr-stak
1
0
Fork 0
mirror of https://github.com/jwinterm/xmr-stak.git synced 2026-10-28 06:28:38 +01:00
Code Issues Packages Projects Releases Wiki Activity

Update documentation for low_power_mode.

Browse Source
This commit is contained in:
Grzegorz Hasse
2017-12-06 16:37:52 -08:00
parent 74d4b2d1c3
commit aa6a6040e3
2 changed files with 18 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
doc/tuning.md
+14
View File
@@ -9,6 +9,8 @@
* [Add more GPUs](#add-more-gpus)
* [Increase Memory Pool](#increase-memory-pool)
* [Scratchpad Indexing](#scratchpad-indexing)
* [CPU Backend](#cpu-backend)
* [Choose Value for `low_power_mode`](#choose-value-for-low_power_mode)
## NVIDIA Backend
@@ -87,3 +89,15 @@ export GPU_SINGLE_ALLOC_PERCENT=99
The layout of the hash scratchpad memory can be changed for each GPU with the option `strided_index` in `amd.txt`.
Try to change the value from the default `true` to `false`.
## CPU Backend
By default the CPU backend can be tuned in the config file `cpu.txt`
### Choose Value for `low_power_mode`
The optimal value for `low_power_mode` depends on the cache size of your CPU, and the number of threads.
The `low_power_mode` can be set to a number between `1` to `5`. When set to a value `N` greater than `1`, this mode increases the single thread performance by `N` times, but also requires at least `2*N` MB of cache per thread. It can also be set to `false` or `true`. The value `false` is equivalent to `1`, and `true` is equivalent to `2`.
This setting is particularly useful for CPUs with very large cache. For example the Intel Crystal Well Processors are equipped with 128MB L4 cache, enough to run 8 threads at an optimal `low_power_mode` value of `5`.
xmrstak/backend/cpu/config.tpl
+4 -2
View File
@@ -1,9 +1,11 @@
R"===(
/*
* Thread configuration for each thread. Make sure it matches the number above.
* low_power_mode - This mode will double the cache usage, and double the single thread performance. It will
* low_power_mode - This can either be a boolean (true or false), or a number between 1 to 5. When set to true,
this mode will double the cache usage, and double the single thread performance. It will
* consume much less power (as less cores are working), but will max out at around 80-85% of
* the maximum performance.
* the maximum performance. When set to a number N greater than 1, this mode will increase the
* cache usage and single thread performance by N times.
*
* no_prefetch - Some sytems can gain up to extra 5% here, but sometimes it will have no difference or make
* things slower.