x86: intel_epb: Set Alder Lake N and Raptor Lake P normal EPB
PATCH 1/1 x86: intel_epb: Set Alder Lake N and Raptor Lake P normal EPB
Intel processors support additional software hint called EPB ("Energy Performance Bias") to guide the hardware heuristic of power management features to favor increasing dynamic performance or conserve energy consumption.
Since this EPB hint is processor specific, the same value of hint can result in different behavior across generations of processors.
commit 4ecc933b7d1f ("x86: intel_epb: Allow model specific normal EPB value")' introduced capability to update the default power up EPB based on the CPU model and updated the default EPB to 7 for Alder Lake mobile CPUs.
The same change is required for other Alder Lake-N and Raptor Lake-P mobile CPUs as the current default of 6 results in higher uncore power consumption. This increase in power is related to memory clock frequency setting based on the EPB value.
Depending on the EPB the minimum memory frequency is set by the firmware. At EPB = 7, the minimum memory frequency is 1/4th compared to EPB = 6. This results in significant power saving for idle and semi-idle workload on a Chrome platform.
For example Change in power and performance from EPB change from 6 to 7 on Alder Lake-N:
Workload Performance diff (%) power diff
----------------------------------------------------
VP9 FHD30 0 (FPS) -218 mw
Google meet 0 (FPS) -385 mw
This 200+ mw power saving is very significant for mobile platform for battery life and thermal reasons.
But as the workload demands more memory bandwidth, the memory frequency will be increased very fast. There is no power savings for such busy workloads.
For example:
Workload Performance diff (%) from EPB 6 to 7
-------------------------------------------------------
Speedometer 2.0 -0.8
WebGL Aquarium 10K
Fish -0.5
Unity 3D 2018 0.2
WebXPRT3 -0.5
There are run to run variations for performance scores for such busy workloads. So the difference is not significant.
Add a new define ENERGY_PERF_BIAS_NORMAL_POWERSAVE for EPB 7 and use it for Alder Lake-N and Raptor Lake-P mobile CPUs.
This modification is done originally by Jeremy Compostella <[email protected]>.
--- arch/x86/include/asm/msr-index.h | 1 + arch/x86/kernel/cpu/intel_epb.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletion(-)
arch/x86/include/asm/msr-index.h +1 -0
@@ -796,6 +796,7 @@#define ENERGY_PERF_BIAS_PERFORMANCE 0#define ENERGY_PERF_BIAS_BALANCE_PERFORMANCE 4#define ENERGY_PERF_BIAS_NORMAL 6 +#define ENERGY_PERF_BIAS_NORMAL_POWERSAVE 7#define ENERGY_PERF_BIAS_BALANCE_POWERSAVE 8#define ENERGY_PERF_BIAS_POWERSAVE 15
arch/x86/kernel/cpu/intel_epb.c +6 -1
@@ -204,7 +204,12 @@ static int intel_epb_offline(unsigned int cpu) }static const struct x86_cpu_id intel_epb_normal[] = { -X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L ,7 ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_N , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (RAPTORLAKE_P , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), {} };
Signature · 1 line
2.31.1
So, the tl;dr on this is:
The EPB setting is fundamentally a classic power vs. performance knob. This change helps power a bunch (200+mw) while having a negligible performance impact (<1% if that), so it's a great trade-off.
Right?
Quoted 10 lines · Dave Hansen
So, the tl;dr on this is:
The EPB setting is fundamentally a classic power vs. performance
knob. This change helps power a bunch (200+mw) while having a
negligible performance impact (<1% if that), so it's a great trade-off.Right?
Yes.
We already using EPB 7 on other AlderLake Mobiles.
Thanks,
Srinivas
Quoted 52 lines · Srinivas Pandruvada
Intel processors support additional software hint called EPB ("Energy Performance Bias") to guide the hardware heuristic of power management features to favor increasing dynamic performance or conserve energy consumption.
Since this EPB hint is processor specific, the same value of hint can result in different behavior across generations of processors.
commit
4ecc933b7d1f("x86:intel_epb: Allow model specific normal EPB value")' introduced capability to update the default power up EPB based on the CPU model and updated the default EPB to 7 for Alder Lake mobile CPUs.The same change is required for other Alder Lake-N and Raptor Lake-P mobile CPUs as the current default of 6 results in higher uncore power consumption. This increase in power is related to memory clock frequency setting based on the EPB value.
Depending on the EPB the minimum memory frequency is set by the firmware. At EPB = 7, the minimum memory frequency is 1/4th compared to EPB = 6. This results in significant power saving for idle and semi-idle workload on a Chrome platform.
For example Change in power and performance from EPB change from 6 to 7 on Alder Lake-N:
Workload Performance diff (%) power diff
----------------------------------------------------VP9 FHD30 0 (FPS) -218 mw Google meet 0 (FPS) -385 mwThis 200+ mw power saving is very significant for mobile platform for battery life and thermal reasons.
But as the workload demands more memory bandwidth, the memory frequency will be increased very fast. There is no power savings for such busy workloads.
For example:
Workload Performance diff (%) from EPB 6 to 7
-------------------------------------------------------
Speedometer 2.0 -0.8
WebGL Aquarium 10KFish -0.5 Unity 3D 2018 0.2 WebXPRT3 -0.5There are run to run variations for performance scores for such busy workloads. So the difference is not significant.
Quoted 7 lines
Add a new define
ENERGY_PERF_BIAS_NORMAL_POWERSAVEfor EPB 7 and use it for Alder Lake-N and Raptor Lake-P mobile CPUs.This modification is done originally by Jeremy Compostella <[email protected]>.
Signed-off-by Srinivas Pandruvada <[email protected]>
Quoted 38 lines
--- arch/x86/include/asm/msr-index.h | 1 + arch/x86/kernel/cpu/intel_epb.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletion(-)arch/x86/include/asm/msr-index.h +1
-0@@ -796,6 +796,7 @@#define ENERGY_PERF_BIAS_PERFORMANCE 0#define ENERGY_PERF_BIAS_BALANCE_PERFORMANCE 4#define ENERGY_PERF_BIAS_NORMAL 6 +#define ENERGY_PERF_BIAS_NORMAL_POWERSAVE 7#define ENERGY_PERF_BIAS_BALANCE_POWERSAVE 8#define ENERGY_PERF_BIAS_POWERSAVE 15arch/x86/kernel/cpu/intel_epb.c +6
-2@@ -204,7 +204,12 @@ static int intel_epb_offline(unsigned int cpu) }static const struct x86_cpu_id intel_epb_normal[] = { -X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L ,7 ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_N , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (RAPTORLAKE_P , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), {} }; --2.31.1
Quoted 58 lines · Rafael J. Wysocki
Quoted 55 lines · Srinivas Pandruvada
Intel processors support additional software hint called EPB ("Energy Performance Bias") to guide the hardware heuristic of power management features to favor increasing dynamic performance or conserve energy consumption.
Since this EPB hint is processor specific, the same value of hint can result in different behavior across generations of processors.
commit
4ecc933b7d1f("x86:intel_epb: Allow model specific normal EPB value")' introduced capability to update the default power up EPB based on the CPU model and updated the default EPB to 7 for Alder Lake mobile CPUs.The same change is required for other Alder Lake-N and Raptor Lake-P mobile CPUs as the current default of 6 results in higher uncore power consumption. This increase in power is related to memory clock frequency setting based on the EPB value.
Depending on the EPB the minimum memory frequency is set by the firmware. At EPB = 7, the minimum memory frequency is 1/4th compared to EPB = 6. This results in significant power saving for idle and semi-idle workload on a Chrome platform.
For example Change in power and performance from EPB change from 6 to 7 on Alder Lake-N:
Workload Performance diff (%) power diff
----------------------------------------------------VP9 FHD30 0 (FPS) -218 mw Google meet 0 (FPS) -385 mwThis 200+ mw power saving is very significant for mobile platform for battery life and thermal reasons.
But as the workload demands more memory bandwidth, the memory frequency will be increased very fast. There is no power savings for such busy workloads.
For example:
Workload Performance diff (%) from EPB 6 to 7
-------------------------------------------------------
Speedometer 2.0 -0.8
WebGL Aquarium 10KFish -0.5 Unity 3D 2018 0.2 WebXPRT3 -0.5There are run to run variations for performance scores for such busy workloads. So the difference is not significant.
Add a new define
ENERGY_PERF_BIAS_NORMAL_POWERSAVEfor EPB 7 and use it for Alder Lake-N and Raptor Lake-P mobile CPUs.
Quoted 6 lines
Quoted 4 lines
This modification is done originally by Jeremy Compostella <[email protected]>.
Signed-off-by Srinivas Pandruvada <[email protected]>Reviewed-by Rafael J. Wysocki <[email protected]>
I'm wondering if there are any plans to pick up this one into the x86 tree? If not, I can take care of it.
Quoted 38 lines
Quoted 38 lines
--- arch/x86/include/asm/msr-index.h | 1 + arch/x86/kernel/cpu/intel_epb.c | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletion(-)arch/x86/include/asm/msr-index.h +1
-0@@ -796,6 +796,7 @@#define ENERGY_PERF_BIAS_PERFORMANCE 0#define ENERGY_PERF_BIAS_BALANCE_PERFORMANCE 4#define ENERGY_PERF_BIAS_NORMAL 6 +#define ENERGY_PERF_BIAS_NORMAL_POWERSAVE 7#define ENERGY_PERF_BIAS_BALANCE_POWERSAVE 8#define ENERGY_PERF_BIAS_POWERSAVE 15arch/x86/kernel/cpu/intel_epb.c +6
-2@@ -204,7 +204,12 @@ static int intel_epb_offline(unsigned int cpu) }static const struct x86_cpu_id intel_epb_normal[] = { -X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L ,7 ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_L , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (ALDERLAKE_N , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), +X86_MATCH_INTEL_FAM6_MODEL (RAPTORLAKE_P , +ENERGY_PERF_BIAS_NORMAL_POWERSAVE ), {} }; --2.31.1
Quoted 2 lines · Rafael J. Wysocki
I'm wondering if there are any plans to pick up this one into the x86 tree? If not, I can take care of it.
I'll bump it up to the front of the queue. :)
Thanks for the reminder.