]> git.ipfire.org Git - thirdparty/linux.git/commitdiff
cpufreq: intel_pstate: Update Balance performance EPP for Emerald Rapids
authorPedro Henrique Kopper <pedro.kopper@canonical.com>
Thu, 1 Aug 2024 16:41:50 +0000 (13:41 -0300)
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>
Fri, 2 Aug 2024 12:40:13 +0000 (14:40 +0200)
On Intel Emerald Rapids machines, we ship the Energy Performance Preference
(EPP) default for balance_performance as 128. However, during an internal
investigation together with Intel, we have determined that 32 is a more
suitable value. This leads to significant improvements in both performance
and energy:

POV-Ray: 32% faster | 12% less energy
OpenSSL: 12% faster | energy within 1%
Build Linux Kernel: 29% faster | 18% less energy

Therefore, we should move the default EPP for balance_performance to 32.
This is in line with what has already been done for Sapphire Rapids.

Signed-off-by: Pedro Henrique Kopper <pedro.kopper@canonical.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/Zqu6zjVMoiXwROBI@capivara
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
drivers/cpufreq/intel_pstate.c

index 392a8000b238b52fce006f9c325439be930dc147..c0278d023cfce595b7cd9f8bb72df60477c8792c 100644 (file)
@@ -3405,6 +3405,7 @@ static const struct x86_cpu_id intel_epp_default[] = {
         */
        X86_MATCH_VFM(INTEL_ALDERLAKE_L, HWP_SET_DEF_BALANCE_PERF_EPP(102)),
        X86_MATCH_VFM(INTEL_SAPPHIRERAPIDS_X, HWP_SET_DEF_BALANCE_PERF_EPP(32)),
+       X86_MATCH_VFM(INTEL_EMERALDRAPIDS_X, HWP_SET_DEF_BALANCE_PERF_EPP(32)),
        X86_MATCH_VFM(INTEL_METEORLAKE_L, HWP_SET_EPP_VALUES(HWP_EPP_POWERSAVE,
                      179, 64, 16)),
        X86_MATCH_VFM(INTEL_ARROWLAKE, HWP_SET_EPP_VALUES(HWP_EPP_POWERSAVE,