[Bug 215135] proposed cpufreq driver amd-pstate regresses wrt acpi-cpufreq on some AMD EPYC Zen3

[bugzilla-daemon@bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=215135

Joe (jinzhou.su@amd.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jinzhou.su@amd.com

--- Comment #2 from Joe (jinzhou.su@amd.com) ---
Hello Giovanni,

Thanks for testing AMD Pstate driver. Finally we have set up the same device on
our local, here is the lscpu info:

Architecture:        x86_64
CPU op-mode(s):      32-bit, 64-bit
Byte Order:          Little Endian
CPU(s):              256
On-line CPU(s) list: 0-255
Thread(s) per core:  2
Core(s) per socket:  64
Socket(s):           2
NUMA node(s):        2
Vendor ID:           AuthenticAMD
CPU family:          25
Model:               1

Here is the tbench test result.

acpi-cpufreq-ondemand  17628  MB/sec  ( 1.00)
acpi-cpufreq-perfgov   25317  MB/sec  ( 1.43)
acpi-cpufreq-sugov     21369   MB/sec ( 1.21)

amd-pstate-ondemand    17913  MB/sec  ( 1.02)
amd-pstate-perfgov     25865  MB/sec  ( 1.47)
amd-pstate-sugov       22359  MB/sec  ( 1.26)

From our site of view, amd-pstate performs slight better acpi driver in
Tbench4. The test based on 50 times test in each policy.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.