1. I have SETI@Home and World Community Grid as my projects. I have an 8 core CPU and an Nvidia graphics card. Right now, BOINC will use all 8 cores for World Community Grid and the SETI task is using my graphics card and a portion of one core. How do I get a balance to where each project is using 4 cores? Is that even possible?

2. Is it faster for BOINC to process tasks on a Linux-based OS or Windows 10? If a Linux-based OS, is there a preference?

BOINC Manager 7.14.2
Windows 10
I've said the projects above.
I've installed BOINC as a service.

Linux is mostly faster, marginally for CPU, mainly due to less overhead, but about 10% on average for GPU computing.
Also, the way Boinc addresses your resources is optimal. WCG doesn't have GPU computing, so all available computing power goes to that.
Your GPU is anywhere near 2x to 50x faster than your CPU, depending on your hardware, so the majority of the work is done by the GPU, and actually assigned to SETI.
If Seti and WCG had the same points rating, you'd notice that SETI will get you more PPD than WCG, simply because it uses the GPU.

If you really want to limit the CPU, you can set to each project, a maximum amount of resources.
You'd have to go into the directory of the project (differs between Windows and Linux), and edit a file called 'app_config.xml', which is explained here:
https://boinc.berkeley.edu/wiki/Client_configuration (Under project level).
You'll need to know the project names,which can be a pain to set up, as you can't modify the CPU resources without this.
That being said., I personally would recommend you to just leave things like they are.

Linux is mostly faster, marginally for CPU, mainly due to less overhead, but about 10% on average for GPU computing.

Never having had a graphics card capable of crunching I can't really comment on that side of things. I understood that for cpu which OS is faster depends on the project. You may well be right that Linux is faster more often than Windows.

Back in the days when CPDN would send the same tasks out to both Linux and Windows they would run about 2% faster using the Windows code under WINE than natively under Linux. Having defenestrated in 1999 I have never ran BOINC natively with Windows.

1. I have SETI@Home and World Community Grid as my projects. I have an 8 core CPU and an Nvidia graphics card. Right now, BOINC will use all 8 cores for World Community Grid and the SETI task is using my graphics card and a portion of one core. How do I get a balance to where each project is using 4 cores? Is that even possible?

The only way to get this specific control is to run multiple clients. You would then set each client to only use four cores, and attach one to WCG and the other to SETI. It can be done but it's not really worth it just for this. BOINC will balance the two pretty well over time, but think in terms of days or weeks, not hours.

2. Is it faster for BOINC to process tasks on a Linux-based OS or Windows 10? If a Linux-based OS, is there a preference?

BOINC Manager 7.14.2
Windows 10
I've said the projects above.
I've installed BOINC as a service.

I"ve seen little difference in CPU work, but the latest AMD CPUs with higher core counts might show a significant difference. There can be a substantial difference in GPU work due to Windows WDDM which adds significant overhead. SETI is a bit unusual in that there is a special app that only runs on Nivida cards under Linux, it's orders of magnitude faster than anything on Windows. The gain is partly due to Linux advantages and partly due to highly optimized programming.

It's interesting you say you are running GPU work with BOINC installed as a service. My understanding is that hasn't worked since BOINC version 6.

---

Team USA forum
Follow us on Twitter
Help us #crunchforcures!

2. Is it faster for BOINC to process tasks on a Linux-based OS or Windows 10? If a Linux-based OS, is there a preference?

While many projects do show Linux to be marginally better than Windows it is dependent on the application, there are some Linux CPU applications that are worse than the Windows equivalent, and there are some Linux GPU applications that are considerably faster than the Windows ones for the same project.
As with many things in life there is no simple answer - the best thing to do is to visit the forum for the each of the projects you are interested in and see what the "real users" have to say.

Thank you for your information. I will leave things alone then.

It's interesting you say you are running GPU work with BOINC installed as a service. My understanding is that hasn't worked since BOINC version 6.

I can do a screen cap of my Tasks tab to show it if you would like.

How do I get a balance to where each project is using 4 cores? Is that even possible?

It was mentioned about using an app_config.xml for each project but no mention of what should go in there.

You just need to put a <project_max_concurrent>4</project_max_concurrent> statement in the app_config for WCG and it will only crunch 4 tasks at any time.

Leaving the other 4 cpu cores for Seti.

Linux is mostly faster, marginally for CPU, mainly due to less overhead, but about 10% on average for GPU computing.

Never having had a graphics card capable of crunching I can't really comment on that side of things. I understood that for cpu which OS is faster depends on the project. You may well be right that Linux is faster more often than Windows.

Back in the days when CPDN would send the same tasks out to both Linux and Windows they would run about 2% faster using the Windows code under WINE than natively under Linux. Having defenestrated in 1999 I have never ran BOINC natively with Windows.

Yeah, Collatz is probably going to run just as fast on Windows as it is on Linux. it has very little CPU and PCIE bandwidth overhead.
GPUGrid, and PrimeGrid have a lot more overhead. The performance penalty is a lot larger there.

Normally, running Wine under linux, is an emulation layer, that should work slower. Not faster; unless there are some compatibility or driver issues.
Driver issues might be more of an issue with exotic hardware (like Larrabee GPUs, High end, or server hardware, or in some cases AMD GPUs).
Nvidia GPUs run better in Linux for crunching.

I yet have to see the first project that works better on Windows than Linux.
Some projects are better optimize on certain platforms, and might not yet have had optimization for other operating systems.
But the general rule is: Windows is slower than Linux, because Linux has way less background resources, smaller footprint. And certain Linux versions (like Lubuntu, linux lite and a few others) have even lower footprint than most.
If one could only find a way to run Boinc on a DSL-like Linux, or Clear linux... it would perform as well as Linux theoretically can perform.

Linux is faster on CPU the less cores you have. The more cores, the more equal the CPU performance or score becomes.

Projects like Einstein haven't gotten tasks that fully use the most modern GPUs, probably also don't have much PCIE bandwidth issues.

Normally, running Wine under linux, is an emulation layer, that should work slower. Not faster; unless there are some compatibility or driver issues.

I agree that under WINE things should be slower than under native Linux. I suspect in my case it was down to the efficiency of the compiled code rather than a driver issue as I have never had any hardware that is remotely exotic.

You would need to use, at least, the <app_version> template in app_config.xml - see Project-level configuration

Set both the <avg_ncpus> (to schedule the right number of cores) and the --nthreads cmdline (to limit what the app grabs for itself) - using the same value for each. Even so, I'm not certain that Virtual machines (which LHC uses) follow the same rules as other MT applications.