LHC@home

Understood.

Honestly, by far your best comment for years throughout all BOINC forums.
Short, topic focused and polite.
More of those and you will persuade more volunteers to help you.

Understood.

Honestly, by far your best comment for years throughout all BOINC forums.
Short, topic focused and polite.
More of those and you will persuade more volunteers to help you.

I do not take in new knowledge unless it has an explanation. Half of what you've told me doesn't have an explanation, just "that's the way it is". Doesn't work for me.

Are my Atlases ok? On defaults, they only use 2 cores out of the 8 Boinc gives them. I see yours do too. I told Boinc to only give it 2 (in nthreads and avg_ncpus), and it still uses 2, although that does mean I use 10GB for 2 cores, so until I upgrade the 24 core machines from 64GB to 128GB, that's a bit limiting.

Have you Hyperthreading on or off?
For me only OFF!
When you have a app_config in the Project-Folder for example Atlas and Theory:
<app_config>
<app>
<name>Theory</name>
<max_concurrent>45</max_concurrent>
</app>
<app>
<name>ATLAS</name>
<max_concurrent>6</max_concurrent>
</app>
<app_version>
<app_name>ATLAS</app_name>
<avg_ncpus>10</avg_ncpus>
<plan_class>vbox64_mt_mcore_atlas</plan_class>
<cmdline>--memory_size_mb 10250</cmdline>
</app_version>
</app_config>

This is used with a AMD Ryzen Threadripper PRO 3995WX 64-Cores. (128 GByte RAM)

Have you Hyperthreading on or off?
For me only OFF!

On. Might aswell use the full power available.

Yes, connect your Squid to that switch, but do it!
And be patient and careful when you set it up.
Doing it in a hurry will fail.

Easier than Apache? Really? I think I got there eventually. Took me a couple of hours.

This seems to show it cached something. The graphs are 40 minutes across, the right half of them is after Squid was set up. Note garage downloads is more than internet downloads, so it must have downloaded from my house machine. I'll keep an eye on that and hope it does more in the future.

Is 40GB a sensible cache? I doubled the default 20GB since I have 300GB free on that drive.



Will other projects make good use of Squid too? For example a new dataset for Rosetta is downloaded by one computer, the others will just get that from Squid?

Ah, that works nicely! :-)
Downloading 10 times more data to the garage than is coming off the internet.

Have you Hyperthreading on or off?
For me only OFF!

On. Might aswell use the full power available.

HT is 1/3 more power as without!

Have you Hyperthreading on or off? For me only OFF!

On. Might aswell use the full power available.

HT is 1/3 more power as without!

So why do you have it off? Why would you not want the extra power?

And in my experience it's 1/2 more.

Only reason to turn it off is if you're more interested in per core speed than overall speed.

Atlas running better without HT,
had testing both at the beginning with this Processor.
Have two, but from different manufactories.

Atlas running better without HT,
had testing both at the beginning with this Processor.
Have two, but from different manufactories.

I've tried many projects with HT on and off and it pretty much averages to 50% faster with HT. HT wouldn't have been invented if it was no use.