TPU's WCG/BOINC Team

[bubbleawsome]

For doing GPU work my client seems to think my computer is always busy and disables it. What counts as "active" for the program?

 

[mstenholm]

For doing GPU work my client seems to think my computer is always busy and disables it. What counts as "active" for the program?

If you want to make sure to do GPU work, aka OPNG, I suggest that you answer Yes to Do work on my graphics card ….
I never tried to answer No since I wanted to be sure to get them. I did at one time do folding and OPNG at the same time by mistake, big hit on the quick return bonus for folding
Just to make it clear NO ONPG jobs have been sent yet.

 

[NastyHabits]

They only sent work out at 12:26 to around 12:50 UTC so if you didn’t actively updated in that period the changes of getting work were slim.

That translates to 5:26 to 5:50 AM for me. I need my beauty sleep. I'll wait for normal operations to resume.

 

[mstenholm]

That translates to 5:26 to 5:50 AM for me. I need my beauty sleep. I'll wait for normal operations to resume.

Make senes. I on the other hand have woken time available and at 12:26 UTC the party began, no OPNG, just APR1 and OPN1 for me. I stopped the MCM1/Linux/AMD combination, points are poor.
Edit: for the first time in weeks all threads on my 3900X and 5950X are busy, I even started my 2700X but only got 11 there. The joy isn’t stopping there, 16 ROSETTA jobs as well. Let’s see when the meter start running with a load of OPNGs as well. This week?

 

[mstenholm]

Same time, fever jobs. Rosetta did issue a bunch 20:40 UTC so now I have a heat problem. These jobs are not held back by memory communication it seems so the CPUs are really working hard. 80 degree C hard on the air cooled ones and 71 C on my 480 mm rad equipped 3900X.

 

[phill]

It's been getting warmer today and tomorrow I think its meant to be even warmer again although not by much with some luck The CPUs and the GPUs been getting toasty today, didn't realise but my 1080 TI in my Threadripper managed to hit 84C, no idea how that happened but after a bit of tweaking and a fan being directed at the back of the card and over the motherboard, it went down to a bit nicer 55C or so..

Monster heat as well. I've even noticed with stock volts and CPU speeds on the 3970X that the temps are registering differently to the actual CPU temps... Its reading 67C there but that's the package temp, so its around 56/57C for the CPU.. I'm going to see if I can get CPU-Z loaded and then see if I can find out what volts its hitting whilst it was crunching... There was only 21 tasks issued to it (better than 1 for sure!!) but I can definitely get that temp down and the watts used... Makes me glad for 1600w PSUs.......

 

[debs3759]

Hopefully be back fully online soon. Got a bunch of MCM and an ARP this evening, so everything that's normally working is testing at last.

Can't say the same for my internet connection though. Builders accidentally cut my cable connection today - cable was only 4 inches under the surface, and they're building me a wheelchair ramp and patio. Gonna be 3 weeks before the cable company fix it, at current estimates. Having to use my neighbour's wi-fi, tethered through a crappy not so smart phone and usb 2. I'm used to a 200 MB connection, getting maybe 30 or 40 with this setup

EDIT:

Just got more MCM, another ARP and an OPN. Feeling hopeful

EDIT2:

And more of everything, plus people are getting OPNG (I don't have a good enough card for them). Maybe we really are close to a roll out

 

[HammerON]

It's been getting warmer today and tomorrow I think its meant to be even warmer again although not by much with some luck The CPUs and the GPUs been getting toasty today, didn't realise but my 1080 TI in my Threadripper managed to hit 84C, no idea how that happened but after a bit of tweaking and a fan being directed at the back of the card and over the motherboard, it went down to a bit nicer 55C or so..

View attachment 257604

Monster heat as well. I've even noticed with stock volts and CPU speeds on the 3970X that the temps are registering differently to the actual CPU temps... Its reading 67C there but that's the package temp, so its around 56/57C for the CPU.. I'm going to see if I can get CPU-Z loaded and then see if I can find out what volts its hitting whilst it was crunching... There was only 21 tasks issued to it (better than 1 for sure!!) but I can definitely get that temp down and the watts used... Makes me glad for 1600w PSUs.......

What are your stock volts? 1.3312?

 

[mstenholm]

Hopefully be back fully online soon. Got a bunch of MCM and an ARP this evening, so everything that's normally working is testing at last.

EDIT2:

And more of everything, plus people are getting OPNG (I don't have a good enough card for them). Maybe we really are close to a roll out

APRs are big, poor phone and you. I noticed that at least 604 OPNGs has been returned. Hmmm

Edit: I have been struggling to download what have been assigned to my machines all day but was clever enough to clean out the last few stuck-in-download and change my setting to NOT get CPU jobs. That way I got 8 OPNGs and the first (got a 8 job/GPU setting) was done in 111 seconds for a claimed 57 points, the rest are still pending download . This no CPU setting is ok when you run ROSETTA on the side otherwise there will be a lot of wasted CPU cycles.

Finally my wingman completed the first batch - result 800-1170 points for around 2 minutes work. I forgot how well they are rewarded.

Anyone got some OPNGs?

 

[80-watt Hamster]

Finally got some MCM1. Yay!


But BOINC seems to be crashing my system. Boo!

 

[bubbleawsome]

Jeez, I'm not sure why, but my system got a ton of MCM1 and OPN1 units that really slammed the CPU. I checked it this morning and it was sitting at 95c, lord knows how long it'd been there. Weird behavior I haven't seen before from any BOINC project.

 

[stinger608]

Wow, I finally got a ton of work units as well. Woohoo.

 

[80-watt Hamster]

Jeez, I'm not sure why, but my system got a ton of MCM1 and OPN1 units that really slammed the CPU. I checked it this morning and it was sitting at 95c, lord knows how long it'd been there. Weird behavior I haven't seen before from any BOINC project.

Yeah, mine was at 90C. Set PL1/2 to 150W, and got it down to 85. Had set BOINC to snooze since I suspected it was causing my clock watchdog errors, but it keeps resuming itself. Been running for a bit now with no further issues; not sure what's going on.

 

[NastyHabits]

It looks like things are back to normal for me. Running smoothly for over 24 hours with little interference from me.

 

[bubbleawsome]

Yeah, mine was at 90C. Set PL1/2 to 150W, and got it down to 85. Had set BOINC to snooze since I suspected it was causing my clock watchdog errors, but it keeps resuming itself. Been running for a bit now with no further issues; not sure what's going on.

I went and re-enabled my old Hydra profiles and got it down to 70c now. Before it was running 1.375v and that had the PL pushed past 200w, now it seems to be hovering around 1.12v and 140w. I'm guessing I have something configured wrong if a full-core load is boosting to that kind of voltage/power. Still, the work units I'd been getting until now didn't do it. Maybe there just weren't enough of them?

 

[HammerON]

Decided to run WCG again last night. This is what I have right now:

 

[phill]

What are your stock volts? 1.3312?

Its whatever the board decides for it mate, I'll have to look and run a CPU-Z to check/see I have noticed that the motherboard CPU reported temp compared to Core Temp is sometimes up to 10 to 12C higher as its picking up the package temps and not the actual CPU temp which was a little worrying to start with but considering the threads and power this thing is eating through, I'm not at all surprised its a little warmer than I'd like it to be...

I think I've had maybe about 1000 work units downloaded over the few machines I have running at the moment? Was hoping to get the 5950X up and running but I had been trying to swap out my mates monitor he lent me and I've not had chance to put a replacement back on yet... Hopefully soon!

 

[NastyHabits]

It is now 3:00 PM PDT and I have not gotten any new work for at least 2 hours.

 

[HammerON]

Its whatever the board decides for it mate, I'll have to look and run a CPU-Z to check/see I have noticed that the motherboard CPU reported temp compared to Core Temp is sometimes up to 10 to 12C higher as its picking up the package temps and not the actual CPU temp which was a little worrying to start with but considering the threads and power this thing is eating through, I'm not at all surprised its a little warmer than I'd like it to be...

I think I've had maybe about 1000 work units downloaded over the few machines I have running at the moment? Was hoping to get the 5950X up and running but I had been trying to swap out my mates monitor he lent me and I've not had chance to put a replacement back on yet... Hopefully soon!

I set my 3960X to 3.8 GHz and left Vcore on auto and it usually stays around 1.14. I find that keeps my temps lower.

It is now 3:00 PM PDT and I have not gotten any new work for at least 2 hours.

I have been dry today as well

 

[80-watt Hamster]

I set my 3960X to 3.8 GHz and left Vcore on auto and it usually stays around 1.14. I find that keeps my temps lower.


I have been dry today as well

The only logical conclusion is that we're just too damned efficient.

 

[NastyHabits]

I just noticed that I'm getting work units again.

 

[stinger608]

I actually got 8 GPU work units today!

Woohoo, it just downloaded 6 more.

 

[HammerON]

I know it is a known issue, but this download error shit is just annoying. I cannot believe this whole move has been so difficult and full of problems. It makes me wonder if these folks have their shit together or not. I have been spamming the "Retry now" link to no avail. I have only been crunching 4 or 5 WU's at a time. Sorry - just have to vent some frustration.

 

[80-watt Hamster]

I know it is a known issue, but this download error shit is just annoying. I cannot believe this whole move has been so difficult and full of problems. It makes me wonder if these folks have their shit together or not. I have been spamming the "Retry now" link to no avail. I have only been crunching 4 or 5 WU's at a time. Sorry - just have to vent some frustration.
View attachment 259533

Oh good, so it's not just me.

I mean, not good per se, but it's nice to know the problem's not on my end.

 

[Chicken Patty]

I see you peeps are still going strong heh?