Looks like new ones are being sent

Tom….


Now that Nbody is back to a reasonable number is it possible to change the credit per WU, on my I7 processor a WU takes roughly 1,000 CPU seconds and 240 seconds elapsed using 6 CPU’s. The credit seems to be aligned to elapsed time not CPU time. Compared to a Separation WU it should get about 70 credits not 26 or 27. The change may encourage more people to process them and maybe stop a large backlog growing in the future.

Just a thought.

What would encourage people like me to do them is if I could get Nbody for my CPUs and Seperation for my GPUs. At the moment I have to do zero Nbodies so I can use the GPUs for seperation.

What would encourage people like me to do them is if I could get Nbody for my CPUs and Seperation for my GPUs. At the moment I have to do zero Nbodies so I can use the GPUs for seperation.

You CAN run n body only on CPUs if you set it up properly using boinc tasks 1.78. reference https://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=4902 I am doing that now. I must confess I tried to do that on my own. No joy. But Wailing Angus Beef helped me set it all up, and with help like that, I joined The Scottish Boinc Team. Send him a PM, he will likely help. And then you can join The Scottish BoincTeam as well! Oh... wait...

What would be nice, would be for some software gurus(and they are thick as thieves in the DC community) to get together and craft a shell to go around boinc tasks 1.78, ala what I suggested here: https://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=4896&postid=73594
I was intending this for each computer.

I just want things nice and simple, it's hard enough keeping the hardware running. Multiple instances etc, no....

I just want things nice and simple, it's hard enough keeping the hardware running. Multiple instances etc, no....

I agree adding the choices we'd like would make things soooo much simpler for us crunchers

All 6 of my nodes are quiet, any new bit's going out, or are we in a lul?

All 6 of my nodes are quiet, any new bit's going out, or are we in a lul?

plenty of work here, across all types

All 6 of my nodes are quiet, any new bit's going out, or are we in a lul?

Your pc's are hidden so it's really hard to tell why, but are your pc's asking for tasks or are they busy doing something else that has already filled your cache?

All 6 of my nodes are quiet, any new bit's going out, or are we in a lul?

There are plenty of tasks ready to be called and are being sent out since several days), as one can more or less see on the Server Status Page.
As far as I can tell, things are fine - at least for N-body ...

I ended up doing

docker run --rm --network boinc boinc/client boinccmd_swarm --get_tasks --force

and that looks to have brought down new tasks. Thanks for the replies.

I have also unhid my nodes, feel free to take a peak.

I ended up doing

docker run --rm --network boinc boinc/client boinccmd_swarm --get_tasks --force

and that looks to have brought down new tasks. Thanks for the replies.

I have also unhid my nodes, feel free to take a peak.

So you took a 12core cpu and split it into 6 different pc's? What advantage does that give you? That seems like alot of work for just 2 cpu cores, I can see it if you had 128 cpu cores and split them into smaller bundles of cores.

I ended up doing

docker run --rm --network boinc boinc/client boinccmd_swarm --get_tasks --force

and that looks to have brought down new tasks. Thanks for the replies.

I have also unhid my nodes, feel free to take a peak.

So you took a 12core cpu and split it into 6 different pc's? What advantage does that give you? That seems like alot of work for just 2 cpu cores, I can see it if you had 128 cpu cores and split them into smaller bundles of cores.

That CPU only has 4 threads, and doesn't allow dual CPU. So either he has at least 3 computers with the same chip, or they're sharing threads which is pointless.

For some reason http://valid.x86.fr/bench/rhwid8/12 says that Intel Xeon E312xx (Sandy Bridge) has 12 threads

For some reason http://valid.x86.fr/bench/rhwid8/12 says that Intel Xeon E312xx (Sandy Bridge) has 12 threads

And Intel themselves say 4.
https://ark.intel.com/content/www/us/en/ark/products/series/59137/intel-xeon-processor-e3-family.html

As does https://openbenchmarking.org/vs/Processor/Intel+Xeon+E312xx

For some reason http://valid.x86.fr/bench/rhwid8/12 says that Intel Xeon E312xx (Sandy Bridge) has 12 threads

And Intel themselves say 4.
https://ark.intel.com/content/www/us/en/ark/products/series/59137/intel-xeon-processor-e3-family.html

As does https://openbenchmarking.org/vs/Processor/Intel+Xeon+E312xx

Hmmm, threads or cores ?

Some cpus have hyperthreading and some don't. https://ark.intel.com/content/www/us/en/ark/products/52278/intel-xeon-processor-e31280-8m-cache-3-50-ghz.html
https://ark.intel.com/content/www/us/en/ark/products/52269/intel-xeon-processor-e31220-8m-cache-3-10-ghz.html

For some reason http://valid.x86.fr/bench/rhwid8/12 says that Intel Xeon E312xx (Sandy Bridge) has 12 threads

And Intel themselves say 4.
https://ark.intel.com/content/www/us/en/ark/products/series/59137/intel-xeon-processor-e3-family.html

As does https://openbenchmarking.org/vs/Processor/Intel+Xeon+E312xx

Hmmm, threads or cores ?

The 2nd link clearly states both are 4.

Some cpus have hyperthreading and some don't. https://ark.intel.com/content/www/us/en/ark/products/52278/intel-xeon-processor-e31280-8m-cache-3-50-ghz.html
https://ark.intel.com/content/www/us/en/ark/products/52269/intel-xeon-processor-e31220-8m-cache-3-10-ghz.html

But is there one with 12 threads? And why are we seeing those x's? Presumably Boinc is just reporting a range, which is annoying and useless.

Hmmm, threads or cores ?

From the already given Intel overview just click on the links to the detailed specs.
Most CPUs have 4 cores/8 threads and hyper-threading. Only a few don't.

Hint:
https://ark.intel.com/content/www/us/en/ark/products/52271/intel-xeon-processor-e31230-8m-cache-3-20-ghz.html