Scheduled Maintenance: We are aware of an issue with Google, AOL, and Yahoo services as email providers which are blocking new registrations. We are trying to fix the issue and we have several internal and external support tickets in process to resolve the issue. Please see: viewtopic.php?t=158230

 

 

 

power management and cpu governors :)

Off-Topic discussions about science, technology, and non Debian specific topics.
Message
Author
Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and cpu governors :)

#41 Post by Deb-fan »

P governors, which are capable (not performance) adjust based on workload, so doesn't matter what a given user is doing. In the case of ondemand (my preference) such things are highly tunable. Something a user can control, like voltage applied which have predictable effect on core temps and allow higher freqs also clearly very significant. Not even sure what you're trying to say or get at, shrugs.

To me this, in ways reflects poorly on Intel too. Suggesting a run of chips they know are quite capable of much higher freqs, that they fiddle with this thermal throttling, microcode, pstate driver (applied voltage), whatever to cap some out at lower freqs and charge folks more/less for the same chips. While certainly not illegal, not even really unethical, still kind of shady practice and personally I fully believe all this side-channel business was no oversight or mistake personally. Aka: Think was purposely built in and those in the "know", knew about it for a long time before they got busted. That's a whole nother thing though. Blah blah. In this at the end of the day techies who learn about this undervolting deal can get decent freq gains or other bennie's out of such knowledge and that's pretty cool. :)
Most powerful FREE tech-support tool on the planet * HERE. *

User avatar
wizard10000
Global Moderator
Global Moderator
Posts: 558
Joined: 2019-04-16 23:15
Location: southeastern us
Has thanked: 76 times
Been thanked: 85 times

Re: power management and cpu governors :)

#42 Post by wizard10000 »

pylkko wrote:Yes, but isn't that a "mistake"? This will not be the case in modern processors, right?. I mean, the so called "dynamic voltage", not the static part that the user can sometimes adjust by undervolting.
There are tools out there that can undervolt each p-state with a different value - for Android there's EX Kernel Manager (requires root) :)

Not positive but I *think* intel-undervolt applies the same undervolt setting to each p-state.
we see things not as they are, but as we are.
-- anais nin

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and cpu governors :)

#43 Post by Deb-fan »

Have a tendency to root for the underdog. Hoping to see AMD moving to make the best of this Intel Trainwreck. :)

Bigwig @ Google Inc calls some higher up at AMD, yeah ... We'd like to place a standing order. Seems Intel's been selling us chips with serious built in exploits for 20yrs. Yeah write this down ...

Person @AMD: YOU want HOW many truckloads of chips,. Errr, could you repeat that, this phone must not be working properly. :P
Most powerful FREE tech-support tool on the planet * HERE. *

CwF
Global Moderator
Global Moderator
Posts: 2638
Joined: 2018-06-20 15:16
Location: Colorado
Has thanked: 41 times
Been thanked: 192 times

Re: power management and cpu governors :)

#44 Post by CwF »

wizard10000 wrote:
pylkko wrote:

wizard10000 wrote:

CwF wrote:
...Past that the lowest clock to get it done does save power.



If the number of cycles required to perform a task is fixed, how does a lower clock save power exactly? If clock frequency and power consumption are linear within a given CPU architecture (they are) doesn't running a slower clock just take more time to get the same job done?

Do you have some kind of source for that?
It thought I said that, it's not linear. I ask earlier to stevepusser if max clock reaches max voltage when undervolting, didn't get an answer. I suspect it does. Every clock step is at a voltage step. Every voltage step higher runs slightly less efficient. The voltage is not static. There is a voltage to clock table, undervolting starts of on a lower rung and steps up to increase clock like normal.

User avatar
wizard10000
Global Moderator
Global Moderator
Posts: 558
Joined: 2019-04-16 23:15
Location: southeastern us
Has thanked: 76 times
Been thanked: 85 times

Re: power management and cpu governors :)

#45 Post by wizard10000 »

CwF wrote:The voltage is not static.
Yeah, I learned a lot about p-states from this thread :)
CwF wrote:I ask earlier to stevepusser if max clock reaches max voltage when undervolting, didn't get an answer.


I ran a stress test against a single core using a -80mv undervolt - processor is an i7-4800MQ with a base clock of 2.7GHz and a turbo frequency of 3.7GHz, during the stress test the single core ran at 3.7GHz and the other seven cores ran at 3.5GHz. CPU temp remained steady at 84C (better than before, I can hit the thermal limiter with a single core at stock voltage).

edit: I had to back it off a tiny bit more, Deb-fan; -75.2mv now and my idle temp has gone up to 37C but that's a lot better than the 43C idle I had before undervolting :)
we see things not as they are, but as we are.
-- anais nin

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and cpu governors :)

#46 Post by Deb-fan »

That's still fairly bad arse Wiz10k and those temps are still well within tolerance. :) Stands to reason a beast of a heatsink or cooling system may complement this undervolting (thermal throttling) situation too. Not sure what's practical in that for a laptop outside of what you're doing already. Have never really checked what methods of stepping up heat dissipation on laptops are being used. People with 6 fans and a liquid cooled system on a desktop pc. Such has always seemed excessive to me, well out of my league. Things throughout this thread taught or clarified a lot of cpu-core related things for me too. Thanks guys ...

Intel is apparently behind this Clear OS distro and all the patchery they're applying to the Linux kernel is apparently causing real improvements in performance. Compared to others ... Considering Intel's been, still is the premiere cpu manufacturing people in the world, pretty much since there's been cpu(s)am sure they've got people with knowledge unequaled on the subject, other technology too. Still AMD's people must not be slouches either, they've been holding their own and staying competitive with a much larger entity forever themselves and Google Inc same .... Last time I'd bothered checking, was making in excess to 20 billion per/Yr in revenues, no doubt that's increased since then as they're always expanding and absorbing new things. Those kind of resources have to buy some of the top tech talent in the world and even with that, still took quite some time before these "flaws" in Intel's design architecture came to light.

Again I think came to light means, well fellas, didn't think it'd ever happen but we finally got busted, was a good run while it lasted though ... 10-15yrs, damn those pesky software engineers at Google! No worries, in this type of thing the waters so murky/muddy by this point, nobody will ever be able to point a finger, it was a "design flaw" afterall. Oops ... sowwie. Meanwhile xyz-govt/Corp has been spying it's booty off or someone's been siphoning off who knows how many gazillions of bucks on the worlds stock markets or whatever else that whole time. :D

Truthfully all this stuff makes me inherently less trustful of Intel's gear. Though because I prefer doing so (buying older hardware), next system will likely end up being Intel based too. Am not going to raise my nose in the air if come across a 3rd/4th gen i5 or i7 at a good price etc. Will snatch it up and begin tweaking hades out of the system per usual. Every now and then, less so lately, I still try to research the Intel vs AMD situation and see who has what, what's the best thing going on in the chip market. Did so last night ...

Ah this chip from Intel is the best choice in performance ... Has this, has that, these things ... pricetag of only $2,999. :P Could put together 10 super-lightspeed space beast systems just for the cost of that chip. Don't believe Intel overly cares about this undervolt affair, still only ever going to be a minority who bothers with it. The base premise just sounds odd and majority of people will stay away. Fiddling around with the current in an electronic component. Yep you just set your mobo so that the chip get's less voltage than the manufacturer sets it up for and whala ... It runs xyz faster because xyz does xyz and the xyz thermal thingy's. Avg person(pc user), errrr yeah, that's great, please step away from my computer now. Yeah ... sounds awesome, stop pushing buttons on my $750 buck computer. YEP ... GREAT, stop touching that!

Used to do that to family/friends, come over and absent mindedly be sitting there tweaking settings on their pc's, disabling default services running on windows, that are known security issues etc. Had to curb the impulse to tweak this junk and strictly confine that to stuff they couldn't ever possibly use etc. :P
Most powerful FREE tech-support tool on the planet * HERE. *

User avatar
stevepusser
Posts: 12930
Joined: 2009-10-06 05:53
Has thanked: 41 times
Been thanked: 71 times

Re: power management and cpu governors :)

#47 Post by stevepusser »

Here's a little screencast gif (made with Peek) of part of my XFCE taskbar in MX 19 right now as all twelve threads are at 100% use compiling the Pale Moon browser update packages.

Image

Left to right are
  • the network down-up (streaming some music) in the hardware-monitor plugin
  • wmhdplop in the wmdock plugin showing hard drive use in a colorful way
  • wmtemp in the wmdock plugin (top is CPU temp in F, don't really know what the SYS is, but the readings are always close)
  • RAM use (top) and swap use bottom (showing a leftover hibernation file from the last time I hibernated the machine), in the hardware-monitor plugin
  • CPU speed
  • CPU use--twelve narrow bars are per core, and the thicker one is overall
The speed is 3.90 GHz for about the first 15 seconds when all are 100%, but then the BIOS drops it down to 3.60. Without undervolting, those values are 3.60 and 3.20. When I first got the Optimus laptop, I couldn't get Bumblebee working, so the Nvidia discrete GPU was always turned on and eating power, so the maximum speed under sustained stress was even lower: 2.80 GHz, which was very disappointing. Later on I got Bumblebee and undervolting working, so the Nvidia GPU is off unless I want it running, and speeds are good.

The MSI also has a undocumented way to unlock the BIOS settings to access all kinds of advanced features, including changing a voltage "slope" setting so it can run at 3.90 GHz all day at 100%, but I didn't like the 93-94 C CPU temperatures I was getting, so reverted that.
MX Linux packager and developer

User avatar
wizard10000
Global Moderator
Global Moderator
Posts: 558
Joined: 2019-04-16 23:15
Location: southeastern us
Has thanked: 76 times
Been thanked: 85 times

Re: power management and cpu governors :)

#48 Post by wizard10000 »

stevepusser wrote:I couldn't get Bumblebee working, so the Nvidia discrete GPU was always turned on and eating power...
Having some difficulty with it myself - I can boot with Intel graphics but for some reason I lose my laptop keyboard when I do and this doesn't happen with nouveau or proprietary Nvidia drivers. I also lose compositing but am less worried about that :)
we see things not as they are, but as we are.
-- anais nin

User avatar
stevepusser
Posts: 12930
Joined: 2009-10-06 05:53
Has thanked: 41 times
Been thanked: 71 times

Re: power management and cpu governors :)

#49 Post by stevepusser »

wizard10000 wrote:
stevepusser wrote:I couldn't get Bumblebee working, so the Nvidia discrete GPU was always turned on and eating power...
Having some difficulty with it myself - I can boot with Intel graphics but for some reason I lose my laptop keyboard when I do and this doesn't happen with nouveau or proprietary Nvidia drivers. I also lose compositing but am less worried about that :)
I'm using the filthy tainted "bumblebee-nvidia" with this laptop. We didn't even try to get bumblebee-nouveau working with the MX Nvidia installer (please don't tell RMS!)
MX Linux packager and developer

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and 'cpu governors :)

#50 Post by Deb-fan »

^ Lmao :) ... Yep, like you don't clearly already know, mid 90's temps would make any techie nervous. Output of sensors having long been informing me 100deg is "critical" would be uncomfortable ever seeing it anywhere close for long. Much less for a sustained period, though you know what's possible, more so than most with application/experience. Would be fiddling and twitching if temps hit those ranges. Also been stringently evading proprietary graphics drivers, will cross that bridge when I have to. :)

Funny post tweaking someone else's pc call back:

Errrrrrr yeah seems to work much better/faster in general, just can't print anything.

Me: Damn you, stupid spooler service!!! :P Why wouldn't it auto launch when needed!? Stupid thing! Oops, over tweaking, will fixy. My fault. :)
Most powerful FREE tech-support tool on the planet * HERE. *

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and cpu governors :)

#51 Post by Deb-fan »

@Steve ... an alternative to making swappiness setting less of a concern on all systems, providing option to install pre-tuned web browsers, include Noscript or equivalent, tweak some things out-of-box. Many of the Firefox tweaks don't effect performance, even improve it, only it's resource useage isn't tarded and Chrome/ium can likewise be tuned. Been advocating this since #!(Crunchbang) dys and no distro devs-maintainers seem interested. For many even adjusting swappiness looks like it'll take an act of Congress to have them even consider it.

Bottomline ... modern browsers are ridiculous in terms of system overhead nowadays, w/o proper tuning. Yes this is on topic, untuned browsers smash people's cpu's now too. Could be even more dreadful depending on power governor. Chrome out-of-box was so bad almost vowed to never launch, install or use the thing again. Then investigated all of what users can do to tune it and had a change of heart. Still prefer Firefox o course.
Most powerful FREE tech-support tool on the planet * HERE. *

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: power management and cpu governors :)

#52 Post by Deb-fan »

HAPPY BIRTHDAY wizard10000!!! Sheesh just thought that was your username, not how many years old ya are. :P No worries, you are getting wiser and stronger in the source. Like star wars, may the source be with you.
Most powerful FREE tech-support tool on the planet * HERE. *

User avatar
wizard10000
Global Moderator
Global Moderator
Posts: 558
Joined: 2019-04-16 23:15
Location: southeastern us
Has thanked: 76 times
Been thanked: 85 times

Re: power management and cpu governors :)

#53 Post by wizard10000 »

Many thanks for the birthday wishes :)

Yesterday was the 23rd anniversary of my 40th birthday. My 40th was the only time I've been surprised by a birthday party and I had so much fun I decided to stop counting there :mrgreen:

On my 40th the spousal unit told me we were going to have a quiet birthday dinner out but when we got to the restaurant she'd reserved a room and all my friends were there. She'd asked all my friends to bring me a toy that cost no more than $3 and we had a blast - I still have a few of those toys more than 20 years later :)
we see things not as they are, but as we are.
-- anais nin

Post Reply