The Ubuntu forums have your answer:

Setting to High Performance

There's two very good solutions given there:

* One using commands (as root or "sudo") on the commandline to directly tweak the running kernel;
* Or better, do the install of "indicator-cpufreq" to tweak however you like from the desktop notifications panel.

Yes I had already see that 3 year old post, when I tried the command it says "performance powersave not found."

When I installed the "incicator-cpufreq" it crashed within a minute.

I do not think anything I have found applies to Ubuntu 10.04.1 LTS

It really isn't that important, I wasn't expecting much from Linux, it is just a stopgap till the new year when I will have the time...

sudo apt-get install indicator-cpufreq

Ah - the tortoise rides again. ;-)

Would be interesting to do some profiling under Linux vs Windows to see from where exactly such noticeable slow down comes...

Yes the

watch -n1 "cat /proc/cpuinfo | grep \"^[c]pu MHz\""

Shows three CPU cores at around 2.5 Ghz which is correct and one CPU core fluctuating as I would expect.

... If Ubuntu slows down my CPU well until I have the time...

* One using commands (as root or "sudo") on the commandline to directly tweak the running kernel;
* Or better, do the install of "indicator-cpufreq" to tweak however you like from the desktop notifications panel.

Note that s@h tries to be kindly and runs at minimum disturbance which also means by default for most Linux distros, running at minimum power...

(Reminds me, we really need an option in Boinc for the user to easily be able to set the run priority of Boinc...)


Happy cool crunchin',
Martin

  │ CONFIG_NUMA:
  │
  │ Enable NUMA (Non Uniform Memory Access) support. 
  │
  │ The kernel will try to allocate memory used by a CPU on the
  │ local memory controller of the CPU and add some more
  │ NUMA awareness to the kernel.
  │
  │ For 64-bit this is recommended if the system is Intel Core i7
  │ (or later), AMD Opteron, or EM64T NUMA.

  │ Symbol: NUMA [=n]
  │ Type  : bool
  │ Prompt: Numa Memory Allocation and Scheduler Support
  │   Location:
  │     -> Processor type and features
  │   Defined at arch/x86/Kconfig:1515
  │   Depends on: SMP [=y] && (X86_64 [=y] || X86_32 [=n] && HIGHMEM64G [=n] && X86_BIGSMP [=n])

  │ CONFIG_KSM:
  │
  │ Enable Kernel Samepage Merging: KSM periodically scans those areas
  │ of an application's address space that an app has advised may be
  │ mergeable.  When it finds pages of identical content, it replaces
  │ the many instances by a single page with that content, so
  │ saving memory until one or another app needs to modify the content. 
  │ Recommended for use with KVM, or with other duplicative applications.
  │ See Documentation/vm/ksm.rst for more information: KSM is inactive
  │ until a program has madvised that an area is MADV_MERGEABLE, and
  │ root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
  │
  │ Symbol: KSM [=n]
  │ Type  : bool
  │ Prompt: Enable KSM for page merging
  │   Location:
  │     -> Memory Management options
  │   Defined at mm/Kconfig:297
  │   Depends on: MMU [=y]

... There's no immediate alternative that's as good and has universal driver/hardware/software support...

... See that it is an AMD APU. Are you trying to crunch on the builtin graphics core along with processing on the cpu? That never works well. ...

So the special app produces just as valid "science" as any stock application.

I am still looking for the "best" sweet spot but it appears to probably be near 30 cores.

Unless the NUMA scheduling works to give some advantage, then you might see the best results are with disabling the SMT to run natively one thread per core with minimum overhead/latency...?

I'm guessing that running non-SMT will also make best use of all of the CPU cache for a speedup also.


Good sleuthing!

Happy cool fast crunchin',
Martin

For my bios, NUMA appears to be turned on by default. I think I tried a short run with UMA on CL16 ram that seemed to show it was slower. But I also think I was running 40 cores.

So another experiment could be SMT turned off (so now max is 32 threads/cores) and UMA turned on. On the grounds of theory, I would always assume UMA would be slower. If the test load is still 26 threads, then a whole lot more of the processing is being done on non-direct memory access cpus which should slow things down. When you add UMA which means even the direct access cpus are going to be accessing memory in a slower way, it "oughta" be slower.

Sigh, yet another test :)

I am still looking for the "best" sweet spot but it appears to probably be near 30 cores.

... Windows 10 can carry on slurping even when you're sure you yelled STOP!

All your activity are belong to us

... In what is likely to be more cock-up than conspiracy, it appears that Microsoft is continuing to collect data on recent user activities even when the user has explicitly said NO...

... Deliberate slurpage, or a case of poor QA and one team not talking to the other aside, it isn't a great look for Microsoft...

... Microsoft got in touch to insist it is committed to privacy and transparency, but admitted there is indeed a bit of naming problem ... Marisa Rogers, Privacy Officer at the software giant, told us: "We are working to address this naming issue in a future update." The slurpage collection is of course for your benefit and Rogers added that users have "controls to manage your data."...

...

1. Under Settings->Privacy->Activity history: ensure the setting "Let Windows sync my activities from this PC to the cloud" is not checked

2. Under Settings->Privacy->Diagnostics & feedback: ensure Diagnostic data is set to Basic

Never attribute to Malice that which can be explained by Incompetence and Laziness.

We have one Win 10 system. The rest are Win 7.

All I can say about Win 10 is ___ and ___ and ___ and ___.

Well, just fill-in the curses.

This article puts some numbers on the reason for the massive ballooning in processing time when I run 56-60 cores of Seti processing.

https://www.pcworld.com/article/3298859/components-processors/how-memory-bandwidth-is-killing-amds-32-core-threadripper-performance.html

The available memory bandwidth goes from about 5 GB for 8/16 cores to around 2 GB for 32 cores.

This makes it sound like the "sweet spot" is going to be 26 threads with SMT running. ...

Latest news, read all about it!
https://www.howtogeek.com/fyi/windows-10-sends-your-activity-history-to-microsoft-even-if-you-tell-it-not-to/

Windows 10 collects an “Activity History” of applications you launch on your PC and sends it to Microsoft. Even if you disable or clear this, Microsoft’s Privacy Dashboard still shows an “Activity History” of applications you’ve launched on your PCs.

* Malevolence?
* Incompetence?
* Dumb uncontrolled uncoordinated shoddiness?
* Incoherent unmaintainability?
* Or something else??

What can you believe?!


IT is what we make it...
Martin