PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
11) Message boards : Problems and Help : BOINC CPU Settings and GPU tasks (Message 149508)
Posted 432 days ago by Profile Rafael
Hi.
Currently BOINC is allowed to use 6 out of 16 threads on my Ryzen 7 5800X, which means i run SOB with 6 cores. I also run a GPU task and it also needs a cpu thread?

The question is, should i let BOINC use 7 out of 16 threads so the GPU can get a core as well for better performance? Or does the GPU app do this automatically?

The thing i want to avoid is the GPU app sharing 1 of the 6 threads with SOB slowing it down.

GPU core sharing is hardly a problem, especially since you have an SMT capable CPU. Use all 8 cores for SoB, it will hardly make a difference for the GPU, but will surely help on the SoB end.
12) Message boards : Number crunching : a variety of CPU generations, overwelmed by all the subprojects in a good way (Message 149265)
Posted 441 days ago by Profile Rafael
Does the 3600x suffer from this same issue regarding CCX units?

Has 2 CCX as well. AFAIK the 3300x is the only non apu Ryzen 3xxx that has a single CCX.


How does Linux fair in this regard, majority of my hosts are running Ubuntu 20.04.

Apparently Linux does a lot better than Windows in this regard and there's little need to deal with affinity.

I do have a 1800x running Windows. Does this still apply on older Ryzens?

Yes.

I am assuming re: LLR and LLR2 instruction set support remains the same e.g. avx?

I think LLR2 can't run on 32bit? Not sure about that.

So to be clear - is hyperthreading (Boinc at 100% CPU on hyperthreaded processors like Ryzen 3700x) useful for non LLR projects, or should I leave it disabled?

It's generally useful for non LLR projects and thus should be enabled most of the time.
13) Message boards : Number crunching : a variety of CPU generations, overwelmed by all the subprojects in a good way (Message 149252)
Posted 441 days ago by Profile Rafael
As far as threads are concerned, what is more benifitial?
Let's give a Ryzen 7 3700x, for instance. Would running 4x 2 threads give more throughput than 2x 4 CPUs? Ultimately - what should my end goal be, or does that also depend heavily on subproject?

It depends heavily on the subproject, or rather the size of the FFT used. The rule is to try and use as few threads per task as possible while still keeping everything within CPU cache. In your case, you have 32MB of cache, so something like Sob that's going at 3200k*8 =~25MB of cache is probably better left as 1x8threads rather than 2x4t, whereas something smaller like ESP at 1600k*8 = 12.8MB would be better at 2x4t since it all fits.

With that said, this is just an educated guess and the best option is for you to test yourself. You might also be fine with suboptimal settings in the name of doing individual tasks faster, or perhaps leave a core free for the GPU or your daily driver needs. It's all up to you.


If I remember correctly, SGS and are all single threaded anyway and only the LLR projects are multithreaded

SGS and PPSE use LLR, and while not very efficient, you can multithread it if you want to. You might be mistaking it for LLR2, which is a similar algorithm used for most other CPU projects and is generally a bit slower on it's own, but significantly cuts down on double checking work (along with some other goodies).

so will lack of hyperthreading in the single threaded projects be hurtful?

HyperThreading (or SMT in AMD's case) is generally not recommended for LLR/2 tasks. It is recommended that you either have BOINC use only 50% of the CPUs or turn it off in BIOS.
14) Message boards : Number crunching : Hash rate for RTX nerf is happening, will it affect crunching? (Message 148862)
Posted 458 days ago by Profile Rafael
I don't think so as it's saying the driver will specifically detect ethereum (why not others?) calculations and slow them down.

Theoretically yes...? But I don't trust Nvidia's drivers to know what their doing if my fun days with the Gtx 970 are anything to go by.
15) Message boards : Number crunching : Badges III (Message 148642)
Posted 464 days ago by Profile Rafael


So, uh. Turing is fast. Like, really fast. Not bad for a "cheap" 1650.
16) Message boards : Number crunching : Nvidia factory overclock and Genefer (Message 147995)
Posted 479 days ago by Profile Rafael
There seem to be very loud warnings in the preferences page about factory overclocked Nvidias and genefer. I'm going to be buying a new fast Nvidia soon (I currently only have AMDs so I've never seen this problem). What exactly does it do? Will I see computation errors so I know it's broken? Or will they complete and not validate? AFAIK, most cards are factory overclocked. Do I really have to somehow undo this, or can I test if it's ok somehow?

Nowadays it's quite rare to see problems since clock control was taken away from people and left up to GPU Boost, so you probably don't need to worry.
17) Message boards : Number crunching : Tour de Primes 2021 (Message 147868)
Posted 483 days ago by Profile Rafael
Ok, so for setting up my dell server (24-core, HT enabled) for LLRI should either turn off HT in the BIOS or tell PG to only use 24 cores? I ask since in preferences it is suggested to enable HT which it already was, and set my core count to 48. As such the system is only running about 45% loaded. While the answer would seem self-evident, I've been working on this project for all of a couple days and am not sure how to optimize my CPU crunching.

If the system is solely dedicated to running PG and doing nothing else (say, being used as a daily driver), then turning HT in BIOS is generally better for LLR performance. And if so, set usage to 100%. Alternatively, if you cannot disable HT for whatever reason, then set usage to 50% to mitigate the effects of leaving it on.

Outside of that, the usual applies: try to keep all your work inside L3 cache to boost performance, and assuming you're talking about that E5-2697 v2, remember that it doesn't have AVX2 and thus will be way lesse efficient than more modern machines.
18) Message boards : Number crunching : New warnings when LLR2 detects Gerbicz errors (Message 147520)
Posted 496 days ago by Profile Rafael
Firstly no OC no boost bios all default values I didn't change any values last few weeks. MB MSI mag 570x latest bios installed. CPU Ryzen 2700x with water cooling. avg temp 60-65 Celsius. Rig has 3 intake 1 out additional fans. No additional OC app installed like MSI dragon only for temps CPUID HW Monitor installed. I will try to see what's going on.

If all is default, then you do have boost enabled, as it's on by default. Go into your BIOS and try disabling it.

Also, what is your RAM and what settings is it actually running at?
19) Message boards : Number crunching : New warnings when LLR2 detects Gerbicz errors (Message 147518)
Posted 496 days ago by Profile Rafael
Hi I get 2 error message at task results which is "Error while computing WARNING!"
When I hover mouse over warning shows " Errors occurred and were corrected during this calculation. Your computer is not operating correctly. This is a hardware problem which you should fix." What is that mean ? I am not change any option or hardware and also 1 day ago completed TRP tasks. Any advice appreciated.

http://www.primegrid.com/result.php?resultid=1169128644
http://www.primegrid.com/result.php?resultid=1167613948


Simply put, your computer is unstable and making mistakes while performing calculations.

As to why, we need a bit more info on your system. If you've done any overclocking, back it off a bit, your OC is unstable. If not, but you do have boost enabled, Ryzen CPUs are known to push the silicon a little bit too far out of the box and require manual intervention, so either disable it or control the clocks manually to prevent it from going ham.
20) Message boards : General discussion : Advice on buying used computers (Message 147333)
Posted 505 days ago by Profile Rafael
I was thinking to get a couple of used computers for use with Primegrid; I have a list of cpus that support AVX-512. But on ebay and craigslist, its hard to find them.

Any advice for what the "sweet spot" is, for used computers?

Laptop or desktop?
i3, i5, or i7?
What video cards to seek or avoid (desktop)

I am looking for the most computing power per dollar, how to find that? The CPU in my laptop is an i7-1065G7, and I am also searching on ebay for used laptops that work but have broken screens; can use an external monitor.

Intel has so many cpus, I dont see why there are so many, its ridiculous!

Any advice is appreciated.

Kevin


Laptops aren't a very good choice; they are expensive for the hardware due to size constraints, and the performance ain't good either because of cooling. Unless they are particularly cheap due to defects like a broken screen or wifi or whatnot, I wouldn't go for them.

For the hardware, getting AVX512 machines is probably a trap. The CPUs are expensive, the motherboards even more so, and the cooling is a nightmare. You'd be better off of the money going with FMA3 capable machines, aka Haswell and older or 3rd gen Ryzen and up (1st and 2nd have a bad implementation). If I was to bet, sweetspot would be 8th/9th gen i5 due to 6 cores per machine, or possibly 9th gen i7 for 8c. Xeons could offer more cores per machine, but the motherboards are probably a lot more expensive than some cheapo H310s.


For GPUs, the more modern, the better efficiency. Nvidia Gtx 16xx are a nice balance of cost (due to not having RTX) and power eficiency until the new cheap 3xxx drop.


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2022 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.38, 0.46, 0.51
Generated 24 May 2022 | 5:41:29 UTC