Announcement

Collapse
No announcement yet.

AMD P-State EPP Patches Spun An 8th Time For Helping Out Linux Performance & Efficiency

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • mdedetrich
    replied
    Originally posted by reba View Post

    3. They earn extra money when customers have to have their burned computers get repaired because the fans kick in very late at temperatures above 90 °C
    Yeah im gonna call BS on this, I regularly stress all cores and I have monitor that shows temps. This doesn't happen with the newest ARM based macs.

    Leave a comment:


  • reba
    replied
    Originally posted by mdedetrich View Post

    I would say that with how much heat modern CPU's produce, any well built chassis is also designed to handle the passive cooling even if the chassis also happens to have fan/s. Any cooling is desirable and if you look at teardowns of the m1 you won't see anything special regarding "being built for passive cooling" and its not too much difference for Intel/AMD based laptops. The M1 doesn't have alien level engineering to take heat out of the CPU passively, its just a bog standard cold plate design. There are a lot of Intel/AMD laptops that have much better chassis for passive cooling compared to the M1.

    The reason why Apple only turns on the fan in emergency cases is

    1. They can get away with it due to how ridiculously efficient the chips are
    2. They are optimizing for noise rather than top performance, which for a laptop has merit.
    3. They earn extra money when customers have to have their burned computers get repaired because the fans kick in very late at temperatures above 90 °C

    Leave a comment:


  • mdedetrich
    replied
    Originally posted by Anux View Post
    Yes that's what I said. The M1 has a cooling system that is designed to work without a fan, the fan is only for emergency. Passive cooling has much more cooling area and more space between finns. A fan based cooling has many tightly spaced finns and doesn't work well without active cooling.
    I would say that with how much heat modern CPU's produce, any well built chassis is also designed to handle the passive cooling even if the chassis also happens to have fan/s. Any cooling is desirable and if you look at teardowns of the m1 you won't see anything special regarding "being built for passive cooling" and its not too much difference for Intel/AMD based laptops. The M1 doesn't have alien level engineering to take heat out of the CPU passively, its just a bog standard cold plate design. There are a lot of Intel/AMD laptops that have much better chassis for passive cooling compared to the M1.

    The reason why Apple only turns on the fan in emergency cases is

    1. They can get away with it due to how ridiculously efficient the chips are
    2. They are optimizing for noise rather than top performance, which for a laptop has merit.

    Leave a comment:


  • Anux
    replied
    Yes that's what I said. The M1 has a cooling system that is designed to work without a fan, the fan is only for emergency. Passive cooling has much more cooling area and more space between finns. A fan based cooling has many tightly spaced finns and doesn't work well without active cooling.

    Leave a comment:


  • mdedetrich
    replied
    Originally posted by Anux View Post
    Thats not like to like, you're comparing a passiv designed cooling system with an active one and take the largest part from the active cooling away, the M1 would also be slower with a crippled cooling solution.
    I own an M1 and even though the machine has fans, they have only turned on twice in the lifetime of me using the machine and that was when running JDK 8 under Rosetta.

    I can completely turn off the fans and for all I am concerened it would have zero impact, the new macs are configured to only turn on fans on extreme scenarios.

    And do note that I am not a trivial user, I am a programmer and a lot of my time is spent stressing the CPU with either multicore compiling or benchmarking.

    Leave a comment:


  • Anux
    replied
    Originally posted by mdedetrich View Post
    If you want to do a more like to like comparison, disable the fan in your AMD laptop
    Thats not like to like, you're comparing a passiv designed cooling system with an active one and take the largest part from the active cooling away, the M1 would also be slower with a crippled cooling solution.

    Leave a comment:


  • mdedetrich
    replied
    Originally posted by andre30correia View Post

    My amd ryzen 7 5700u makes 9400 points in cinebench r23 even m2 cant beat that, using pstate driver in a cheap lenovo laptop (600 euros)
    This is ignoring the fact that the m2 in the Air is only passively cooled where as almost every laptop with a 5700u/6850u has fan cooling. If you want to do a more like to like comparison, disable the fan in your AMD laptop

    Also arguably in terms of segmentation, the m1 pro is a lot closer to a 5700u rather than a M2
    Last edited by mdedetrich; 21 December 2022, 02:52 AM.

    Leave a comment:


  • andre30correia
    replied
    Originally posted by Mitch View Post
    If your goal is the highest efficiency for the most workloads, it seems like Epp Balance Performance is the best all around choice for efficiency. What's interesting is that Power save is lower PPW, so unless you're working with a bad cooling solution or want minimum fan noise, you should stick with Epp balance performance.

    Obviously other benchmarks may prove the p-state on demand will win more often, but just looking at what they posted here, that's my take.

    My 6850u seems to rival my M1 air's efficiency sometimes (mostly multicore workloads), so I wonder if these patches could push it over that gap for some scenarios
    My amd ryzen 7 5700u makes 9400 points in cinebench r23 even m2 cant beat that, using pstate driver in a cheap lenovo laptop (600 euros)

    Leave a comment:


  • mdedetrich
    replied
    Originally posted by HD7950 View Post

    I found the culprit: CONFIG_X86_INTEL_PSTATE was disabled on my personal kernel but it is needed by the AMD EPP driver now.

    Thanks!
    On a side note, if the pstate is now being generic between Intel and AMD shouldn't the config option be renamed to CONFIG_X86_PSTATE so its less misleading?

    Leave a comment:


  • ptr1337
    replied
    Originally posted by HD7950 View Post

    I found the culprit: CONFIG_X86_INTEL_PSTATE was disabled on my personal kernel but it is needed by the AMD EPP driver now.

    Thanks!
    Maybe report this to Perry Yuan.
    Maybe the KCONFIG needs to be adjusted that it depends on this, if the build otherwise is failing. Happened cause he moved things to the intel_pstate.
    But im not sure.

    Leave a comment:

Working...
X