How can Woodall(LLR)(mt) and 9.01 PPS (LLR) (mt) run at the same time? (10 cores each)
I have 16 cores. 10 are used by PG and the others are spread over 4 tasks from another project and 2 controlling my GPU (combined .36 of a core)

This is physically impossible unless both projects are sharing the same 10 cores?

How can Woodall(LLR)(mt) and 9.01 PPS (LLR) (mt) run at the same time? (10 cores each)
I have 16 cores. 10 are used by PG and the others are spread over 4 tasks from another project and 2 controlling my GPU (combined .36 of a core)

This is physically impossible unless both projects are sharing the same 10 cores?

Actually, to be the bearer of bad news, if you're talking about the Ryzen 2700 on your account, you only have 8 physical cores. The other 8 threads are logical. There are many threads on this forum about this discrepancy and how to get the most out of such a system. Search for "hyperthreading".

Regardless of that, to address the matter at hand, I don't think it is actually doing both at the same time, but rather switching between the two units. This happens a lot when you have mixed long (e.g. Woodall) and short (e.g. PPS) tasks. When you get near the end of a task, Boinc downloads a new one, but when they are vastly different sizes, it starts to switch back and forth because of a) different deadline setups and b) weird logic, until the first finally finishes.

Or, I could be off base. Can you take a screenshot of your Boinc window showing both tasks?

How can Woodall(LLR)(mt) and 9.01 PPS (LLR) (mt) run at the same time? (10 cores each)
I have 16 cores. 10 are used by PG and the others are spread over 4 tasks from another project and 2 controlling my GPU (combined .36 of a core)

This is physically impossible unless both projects are sharing the same 10 cores?

I've seen that on my own pc's before and eventually one of the tasks should say 'waiting to run' as like you say there is not enough resources to run it.

Boinc is letting the tasks run through a time sharing scheme. Although you may have allocated 10 threads to a task Boinc is only allowing a few threads to process your work at a time. I believe this is an artifact of how Boinc implements multi-threading.

If you look at your non-GPU tasks and multiply the run time by 10 (number of threads allocated) you will see a vast difference between runtime and cpu time. These two numbers should be about the same, however when tasks are "running" but not "processing" it will result in what you see is happening to your tasks.

------

Now that LLR2 has been implemented there is far less incentive to run many threads for any given task...


____________
Werinbert is not prime... or PRPnet keeps telling me so.
Badge score: 2x3 + 5x4 + 5x5 + 5x7 + 1x9 + 3x10 = 125