BOINC Pentathlon 2021

[Gorgon]

1 hour ago, Windows7ge said:

I'm still trying to make sense of how I'm doing. SiDock is really slow to ramp up stats. I'm starting to think I won't see accurate numbers till we're half way through the event.

Yeah, the dual quorum doesn't help either.

 

My estimate is 180,000 PPD but I'm only at 41,000 according to their stats page. In theory, if their using Recent Average Credit (RAC) it should be a running average and settle after 3.5 days of consistent production.

 

Well I almost have that three 1/2 days in now and though my total points agree with the estimate their average is way off base.

 

BOINCstats appears to be a more accurate average

[GOTSpectrum]

I personally just don't look and hope for the best lmfao 

[GOTSpectrum]

I had Foldatron offline for about an hour today, for some reason the Nvidia driver shit the bed and it was not happy. Something caused BOINC to die too on a restard and needed a reinstall. But everything seems fine now... we shall see

[Windows7ge]

Yeah, when I have a chance today I'm moving the 1950X to the Javlin throw. It hasn't completed any SiDock work at all.

 

No checkpoints is really dumb if you can't even turn the computer off at the end of the day. Not everyone can run everything 24/7.

 

7 minutes ago, Gorgon said:

Yeah, the dual quorum doesn't help either.

 

My estimate is 180,000 PPD but I'm only at 41,000 according to their stats page. In theory, if their using Recent Average Credit (RAC) it should be a running average and settle after 3.5 days of consistent production.

 

Well I almost have that three 1/2 days in now and though my total points agree with the estimate their average is way off base.

I didn't really look at total credit. Total credit suggests I aught to be somewhere around 80,000~85,000 PPD but I'm not even at the 48hr mark. A far cry from 3.5 days.

[Red :)]

Welp, I got a bit slowed down earlier today thanks to a planned power outage.

[leadeater]

3 hours ago, Shlouski said:

@leadeater

 

Hiding your computers awww

I had to, normally I do not but the RHEL servers hostnames have my work's domain name on them (FQDN) showing and I'd like to not tell the world where I work lol

[Windows7ge]

1 hour ago, leadeater said:

I had to, normally I do not but the RHEL servers hostnames have my work's domain name on them (FQDN) showing and I'd like to not tell the world where I work lol

Because you know I'd show up asking to "borrow" some of the excess equipment?

[Gorgon]

35 minutes ago, Windows7ge said:

Because you know I'd show up asking to "borrow" some of the excess equipment?

Good thing they wouldn’t let you into the country then 

[GOTSpectrum]

24 hours until the start of primegrid and the next javelin throw, make sure you are getting your bunkering in guys!

[Favebook]

6 hours ago, GOTSpectrum said:

Yeah, primegrid is the GPU task, you may need to edit your cc_config file to let it use multiple GPUs though.

 

Looks like I can't add WCG for some reason so guess I won't be helping with the jav throw

By default, on one of my systems, PrimeGrid uses 2 out of 2 GPUs but does not utilize them fully. Do you know where I can find a guide on editing cc_config files to use them fully and efficiently? I just spent about 30min searching for it but couldn't find it. I remember setting up my app_config file for WCG but I do not remember anything about PrimeGrid.

[Windows7ge]

1 hour ago, Gorgon said:

Good thing they wouldn’t let you into the country then 

I have connections, and people.

 

[Baha]

Hmm. For some reason I don't seem to be able to pull down any more tasks for PrimeGrid - I have around 100 tasks bunkered pending upload, and can pull down work tasks for other projects (so it isn't a disk space/network issue), but not PrimeGrid.

Is anyone else having issues pulling new tasks for PrimeGrid at the moment?

Edit: On closer inspection, it looks like I was mistaken about being able to pull down other tasks - seems like you might just have to clear out your upload queue before you're allowed to pull down new tasks.

[Gorgon]

1 hour ago, Favebook said:

By default, on one of my systems, PrimeGrid uses 2 out of 2 GPUs but does not utilize them fully. Do you know where I can find a guide on editing cc_config files to use them fully and efficiently? I just spent about 30min searching for it but couldn't find it. I remember setting up my app_config file for WCG but I do not remember anything about PrimeGrid.

Here from the manual 

 

You’ll have to season to taste what portion of tasks per CPU and GPU to use.

[rkv_2401]

@leadeater here are some of the runtimes I've seen for finished WCG MIP tasks - ranges from 36 mins to 2 hours 25 mins, with the average around 1:15- 1:45 on a Ryzen 3600 @ 4.1Ghz. Also, there's a task that's been running for an hour and 10 mins, but is only 30% complete. There does seem to be a lot of variation in runtimes.

[Windows7ge]

Bunkering WCG on the 1950X. I assume it doesn't matter which one I release it on if the lowest score of three is the one they count?

[leadeater]

21 minutes ago, Windows7ge said:

Bunkering WCG on the 1950X. I assume it doesn't matter which one I release it on if the lowest score of three is the one they count?

Correct, saying that this upcoming throw will have to be out "best one" as in try the most on then we just need to get as close to that as possible 2 more times. But if we think we did badly we can discard this one as there are 4 left and find a way to try harder.

[Shlouski]

7 hours ago, Favebook said:

By default, on one of my systems, PrimeGrid uses 2 out of 2 GPUs but does not utilize them fully. Do you know where I can find a guide on editing cc_config files to use them fully and efficiently?

I was worried that this might be a thing, hopefully it's possible to run multiple tasks simultaneously. 

[Arrogath]

I’m bad at managing things like bunkering, especially since I’m away from the pc and busy with work and other projects so instead you just get my contributions as a baseline increase. 3950x and 5800x are on the marathon and javelin throw running at power ratings that won’t burn the house down.

[GOTSpectrum]

1 minute ago, Arrogath said:

I’m bad at managing things like bunkering, especially since I’m away from the pc and busy with work and other projects so instead you just get my contributions as a baseline increase. 3950x and 5800x are on the marathon and javelin throw running at power ratings that won’t burn the house down.

You are spectacular

 

Massively appreciated 

[Shlouski]

Ok, I have joined primegrid and linustechtips team.

 

Have to go out for a bit, but I will get to setting every up when I get back.

[GOTSpectrum]

7 minutes ago, Shlouski said:

Ok, I have joined primegrid and linustechtips team.

 

Have to go out for a bit, but I will get to setting every up when I get back.

If you have multiple GPUs in the same system make sure that you see the same number of primegrid WUs being crunched at the same time

[GOTSpectrum]

Cross country, Einstein@home runs on the following platforms, 

 

AMD GPUs, Intel GPUs(yes that means iGPUs), Nvidia GPUs, Linux on ARM(Respberry Pis and similar devices), Android, Linux x86, Mac and windows. 

 

So get your Intel GPUs running and make use of all that silicon. 

 

We are a few days out for this discipline, but how are we thinking about going with this one? As it will overlap the end of the primegrid discipline. 

[Shlouski]

So far primegrid is ignoring my gpu's, messing around in Boinc managers settings hasn't help thus far.

[GOTSpectrum]

1 minute ago, Shlouski said:

So far primegrid is ignoring my gpu's, messing around in Boinc managers settings hasn't help thus far.

Have you set your preferences on the primegrid website, you need to tell it what to do. 

[Shlouski]

3 minutes ago, GOTSpectrum said:

Have you set your preferences on the primegrid website, you need to tell it what to do. 

 

Ah ok, didn't know that.

 

On it thanks.