Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > alt.comp.os.windows-10 > #184245

Re: Different levels of sleep.

From Paul <nospam@needed.invalid>
Newsgroups alt.comp.os.windows-10, alt.comp.os.windows-11
Subject Re: Different levels of sleep.
Date 2025-05-02 06:49 -0400
Organization A noiseless patient Spider
Message-ID <vv280k$upmt$1@dont-email.me> (permalink)
References (1 earlier) <vuv4ai$2337s$1@dont-email.me> <20250501083350.f9dcf05496619c9fa3da2f4d@127.0.0.1> <m7gn7qF57lhU1@mid.individual.net> <MPG.427d52fec8b442ef9903f6@news.individual.net> <j0e71k5tuofmnd1cfnvus800924d86mg4d@4ax.com>

Cross-posted to 2 groups.

Show all headers | View raw


On Thu, 5/1/2025 2:28 PM, micky wrote:

> Anyhow, according to several pages, including the one you list below: 
> Working 	S0 	The system is fully usable. Hardware components
> that are not in use can save power by entering a lower power state.
> 
> Googles AI, which I don't trust at all, says "S0 (Active):
> This is the standard working state where the computer is fully
> operational"  which is pretty close to what the pages I do trust say. So
> maybe you have S0 but S0 on win10 is not lower power idle?????   Or
> maybe lower power idle is not as idle as it sounds? 

S0 is not a sleep state. That's the "operating" state.

The rest of it, mixes two schemes together, to no particular purpose.

The purpose of states, is to draw state diagrams, and as
a classification mechanism for behaviors. You will notice
there is "air space" between the textual words. This is
meant to imply "something poorly designed is missing", OK ?
We simply expunge the parts we don't understand, to help
people deal with the parts that might have meaning to them.

    Operating        Sleep Hibernate SoftOff  (MechanicalOff)
                                      +5VSB     switched off

The newest idea, doesn't like the states, so it crushes some
together. It can do this, because there are other states
that can be used. C0...C6,C12. Where C12 would be a state where
a processor core has lost all power. To keep a CPU running at
top clock rate, and not taking a break from that top clock
rate, you can turn off the C state handling at the BIOS level.

Even the instrumentation has taken a nose dive. But an
understandable one. If I run CPUZ and check the clock rate,
the clock rate might measure 3.4GHz or 4GHz or so. The
act of measuring the thing, like Heisenburg Uncertainty,
pollutes the reading. CPUZ uses a straightforward measurement
method that more of the CPUs could support. AMD Ryzen Master
uses some instrumentation in the CPU (there are Performance Counters
for all sorts of things, for which the info is thin). The clock
measured there, is likely being measured by a logic block and
does not require spinning up a CPU core to do the math. Then,
the two measurements are at odds with one another.

When equipment has external power consumption (you pull the
battery out of the laptop, forcing it to be honest about
its power usage), even when the equipment lies, the power
consumption does not lie. I can tell when Task Manager is
hiding things. I can watch the LEDs on my IPV4 network,
for evidence of skullduggery. The 1800+ pages of ACPI specs
add nothing to the color commentary any more, due to a
confused set of purposes.

*******

Now, let's check mine. I have removed the garbage from the trace.

PS> powercfg /availablesleepstates
The following sleep states are available on this system:
    Standby (S3)                           <=== That's a deep sleep state with a fast recovery (2-3 seconds)

The following sleep states are not available on this system:

    Hibernate                              <=== This is bashful S4 state
        Hibernation has not been enabled.
PS>

It seems my machine can sleep, but not hibernate. It might
have soft off and it definitely has mechanical off (there is
a mains switch on the back or I can pull the cord out of the wall).

When it is in S0 state, the cores are dancing a jig. If you
look in AMD Ryzen Master, one core is running at 500MHz or so,
the other cores are "C6 or lower". C12 is the lowest state
for the CPU. On the 5700G, this is 32 watts (but some of that power
is for the NVidia card). On the 5600G, this is 22 watts (uses iGPU).

   500   Sleep   Sleep   Sleep   Sleep   Sleep   Sleep   Sleep     # My 5700G right now (Ryzen Master readout)

    C0    <=C6    <=C6    <=C6    <=C6    <=C6    <=C6    <=C6     # Imaginary core states

  [                                                            ]   # Machine is in S0, while CPU cores dance a jig

What was curious about one machine here, is the power consumption
changed with time. When initially built, I got one power figure
at idle. Then one day, the power was cut in half at idle. The
other machines missed this transition and have been consistent
since being built. The machines were built, while debugging
my daily driver (which had some really weird crashes I needed
to understand). The issue is fixed... but leaving me with
an excess of test equipment. That is why I have lots of
data points to offer now.

    Paul

Back to alt.comp.os.windows-10 | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

Different levels of sleep.  micky <NONONOmisc07@fmguy.com> - 2025-05-01 01:02 -0400
  Re: Different levels of sleep. "R.Wieser" <address@is.invalid> - 2025-05-01 08:28 +0200
    Re: Different levels of sleep. "Kerr-Mudd, John" <admin@127.0.0.1> - 2025-05-01 08:33 +0100
      Re: Different levels of sleep. Andy Burns <usenet@andyburns.uk> - 2025-05-01 08:46 +0100
        Re: Different levels of sleep. Stan Brown <the_stan_brown@fastmail.fm> - 2025-05-01 10:45 -0700
          Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 14:28 -0400
            Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 14:40 -0400
            Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-02 01:48 -0400
              Re: Different levels of sleep. Paul <nospam@needed.invalid> - 2025-05-02 14:08 -0400
            Re: Different levels of sleep. Paul <nospam@needed.invalid> - 2025-05-02 06:49 -0400
          Re: Different levels of sleep. Daniel70 <daniel47@eternal-september.org> - 2025-05-02 19:36 +1000
      Re: Different levels of sleep. "R.Wieser" <address@is.invalid> - 2025-05-01 10:51 +0200
        Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 13:39 -0400
    Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 04:59 -0400
      Re: Different levels of sleep. MikeS <mikes@is.invalid> - 2025-05-01 16:30 +0100
        Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 14:33 -0400
      Re: Different levels of sleep. "R.Wieser" <address@is.invalid> - 2025-05-01 17:35 +0200
        Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 14:31 -0400
      Re: Different levels of sleep. Newyana2 <newyana@invalid.nospam> - 2025-05-01 11:51 -0400
        Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-01 14:03 -0400
          Re: Different levels of sleep. Newyana2 <newyana@invalid.nospam> - 2025-05-01 15:48 -0400
            Re: Different levels of sleep. Paul <nospam@needed.invalid> - 2025-05-02 06:58 -0400
              Re: Different levels of sleep. Newyana2 <newyana@invalid.nospam> - 2025-05-02 08:15 -0400
                Re: Different levels of sleep. Paul <nospam@needed.invalid> - 2025-05-02 12:01 -0400
          Re: Different levels of sleep. micky <NONONOmisc07@fmguy.com> - 2025-05-02 01:36 -0400
  Re: Different levels of sleep.  micky <NONONOmisc07@fmguy.com> - 2025-05-21 15:21 -0400
    Re: Different levels of sleep. Sleep Expert <invalid@invalid.invalid> - 2025-05-22 00:01 +0000

csiph-web