Thanks! I don't blame you there, I picked up a big ol lancool case and a 750w platinum psu, I'm moving my 2080s to the garage once winter is over, it's cooled but theres no return so in theory it won't hurt the electric bill as much since it'll be isolated from the t-stat heh, we shall see. You're doing some serious damage, you might catch me!
That's a great idea to move a heatsource out of the house, I wish I had anywhere to put those systems besides my bedroom or my office. My apartment is one level with a flat roof, so even without the systems running it's pretty warm.
I didn't have that luxury last year, so it's a nice improvement. Thats a big swing, are you folding on cpus? Every cpu I own is also folding, I think they tend to have much smaller faster units. It took me 6 years to get my first billion, at my present rate I'll be rolling over the odometer every 3 months lol.
I was folding on every CPU a few months ago, I think it was before the Cuda update. The 3950x was pumping out a unit every 30 minutes, pretty nuts!
Thats news to me, I hadn't heard about cpu folding hurting the cuda gpu wus. I've always folded cpu, so I suppose I probably wouldn't have noticed a difference, but my cards seem to be getting whats expected. 3-3.5 on my 2080s with i5's set to 5 cores, 1.2ish average on my 1660 fleet with smp 14 on 2700, and 10 on 5600x and 1600. I'll have to pay attention once I get the dedicated 2080 rig running as I don't intend to run cpu on it. I've not had good luck attempting to run AMD machines with multiple folding cards, while also cpu folding.I was folding on every CPU a few months ago, I think it was before the Cuda update. The 3950x was pumping out a unit every 30 minutes, pretty nuts!
But now, for some reason I can't get it to work in such a way that there is no decrease in GPU performance.
The last few CPU units I did were taking 2 hours each on 12 threads on the 3950x. 300k ppd but a loss of 200k ppd on the GPU, not worth it imo.
I'm not the only one having this problem, I saw something about it on the folding forum too with the same processor. Even with Process Lasso locking tasks to individual cores, which seems to work, I still see the TPF go up after starting the CPU client. I still need to give Linux another try, but all the hassle to get that working does not motivate me at all
Pretty insane isn't it? Working years and years to get where you are, and totally crushing those numbers in only a few months..
I tried finishing on my cpu paired with my rx580 at work which I can closely monitor during the week, no discernable difference in TPF aside from the typical fluctuations. Perhaps I have to try with an nvidia card? 10% is more than enough to be worth exploring most certainly!
I'm actually scraping together a phenom ii/ddr 2 machine for my man cave htpc, so it seems I'm moving backwards haha.
Not old, refined, like a fine bottle of wine.Thanks for making me feel old
Nice! Grats!
Congrats!
Congrats! 🥂
