Primegrid Challenge Series 2024 (Updates)

I am *very* much looking forward to this one. All of these RTX400-series cards should tear through GFN tasks at a tremendous clip.
 
[Ion] said:
I am *very* much looking forward to this one. All of these RTX400-series cards should tear through GFN tasks at a tremendous clip.
Click to expand...
What 4000 series cards are you running? Good time to bring out the space heaters. My best card is "only" 3080Ti
 
I'll give it all I can. If I can find something to stick them in, I'll fire up a few 1080 TIs I've not been running b/c their poor efficiency.
 
Looks like Willy has got a BOINCStats team over on PG. It would make a nice show for us to propel his team to greatness during this challenge. Anyone else want to join in on the effort?
 
I most likely won't be joining this challenge as I'm limited on usable hardware to be worth the expense it incurs. Hopefully the temps get better here so that I can justify running all my outdated gear soon.
 
One week until the challenge starts! Hope you're all looking forward to it!
 
Willy's challenge starts mid day tomorrow. 10 Day challenge so plenty of time to drop in for a little or a lot.
 
And we're off!

For Willy!

1733858035196.png
 
I dug out two more 1080 TIs this morning....not going to run them indefinitely, but they ought to do abut 350k PPD each for the challenge, and I'll suffer through the power draw for a week and a half.
 
I am not digging all that deep, for sure. Most of my 30's are idle. For now. Got a couple space heaters going.
 
1733953469169.png


wareyore What board do you have your 7700k in that you can hang five GPUs off of it? I've done that on one of my Xeons, but, well, it's a dual Xeon, so it has a lot of PCIe slots :)
 
Maximus IX Hero. PCIe 1x for the extras. For mining and DC tasks that don't require the bandwidth it works fine.
 
Nice, and it's fun having that many GPUs in a single system too :)
 
Just noticed one of the space heaters was running private gfn , still. Whoops.
 
After 2 days, our team is at 6th spot. We have now 7 crunchers.

I might probably test cpu crunching on GFN-21 on 7950x after my CPDN wah2 region independent run time has reached 100khrs in about 3 days.


1734031453795.png
 
I briefly considered running it on my 5800x but if my math was correct it would probably take at least three or four days to finish a work unit and that wouldn't be worth it. Same reason I didn't fire it up on my 1050Ti. It already takes 7.5-8 hours per work unit on a 6750xt which means the 1050Ti would probably be a minimum of three days per work unit.
 
It isn't ideal but I'm going to try running my old L5639 box that had dual 2080's in it. It is cold enough outside that I can run it for heat. But I have to watch the circuit it is on to not flip a breaker...lol.

Edit: just waiting for the system to pull work.
 
Last edited:
As long as cold temps remain, I will have 2x 2080S and 3x2080.

Edit: and depending on how things go, I "might" be able to fire up my old thread ripper for 2 more 2080's.
 
Priller said:
Ill be joining once my current WU has completed.
Click to expand...
Driver update, Pulling the card VRAM down to stock, lowering clock speed, and even giving it more power still results in a computation error. That's over 15 hours I wont get back.

Guess I'm out of this one. Just doesn't seem to play nice with my 6800XT.
 
Priller said:
Driver update, Pulling the card VRAM down to stock, lowering clock speed, and even giving it more power still results in a computation error. That's over 15 hours I wont get back.

Guess I'm out of this one. Just doesn't seem to play nice with my 6800XT.
Click to expand...
Someone with 6900XT running GFN-22in windows also see the same problem as you did. Not clear if there is a solution to this in this thread: https://www.primegrid.com/forum_thread.php?id=10709#176593

Edit: it seems that he is able to run GFN-21 now https://www.primegrid.com/results.php?hostid=1334599
 
Last edited:
Priller said:
I posted in that thread. It 'may' have been fixed at one point in a driver but it might be broken again.
Click to expand...
The difference between your failed WU and his other GFN is that you got this error message "The operating system cannot run %1." A quick google seems to show that this has to do with printer but I think it is just Microsoft windows OS not knowing how to classify the error. The %1 is used to pass an argument in DOS batch file: https://stackoverflow.com/questions/2309968/what-are-1-and-2-in-batch-files. Maybe something to do with Windows OS issue will be my wild guess. Anyway, let see if anyone replies.
 
Well the L5639 died today...will have to open it up and see if it is the PSU or something else. Not even sure if it will be worth troubleshooting...lol
 
That's a pretty old CPU these days, but the two 2080s you have in there ought to do a pretty good job and be worth running.
 
[Ion] said:
That's a pretty old CPU these days, but the two 2080s you have in there ought to do a pretty good job and be worth running.
Click to expand...
Yup...but if the box isn't worth fixing, they will sit cold. I have some random parts so it just depends on what the problem is.
 
Day 4. Holding steady at 6th spot in team standing. Number of participants increases from 7 → 11 since Day 2. If we can hold on for the next 6 days, that will be out best finishing in 2024 PG challenges.


1734207033959.png
 
Been getting whooped the past couple days unfortunately. Since all of my production comes from GPUs, the slower results of the newer tasks on GPUs has hurt my output pretty bad :/
 
At the beginning I was at around 8 hours per work unit on my 6750xt and now it's at least 13 hours. I've seen considerable increases before but nothing like this, even during a Pentathlon.
 
I stopped the PG work units. After the initial struggle to get them to run, my WU times have gone from 5.5-6 to 7.5-8 hours.

Add the fact the points are exactly the same for both and it's not worth it for me.
 
You must log in or register to reply here.
Back
Top