Void Master

Say no more fam! (XOC Updates once every three hours, but your stats on the official F@H server shows real time point earned. I always link XOC though because it gives you WAY more info)
PPD for the 2060S is about 1.2M ppd. 
 
The 2060S uses less power than the 1080 ti when both are at 100% power usage, and theoretically you can get a 2060S for about 20%less cost than the 1080ti.
 
Note though that the 1080ti will still be a bit better in games though (Closer to 5700XT and 2070S levels) also a 1080ti is still a good buy if you're not doing only folding. Kinda like Ryzen is a good buy if you're not doing only gaming  

I have two 1070ti's in Project ITNOS, and they both earn about 950k ppd currently (Under powered and slightly overclocked though). a1080 will be about 1m-1.1m ppd 
 
They should be very similar in PPD given similar WU projects. 
WHAT THE HELL IS THIS THIS ISNT ANYTHIUNG TO DO WITH FOLDING
 
LOCKED 
/s

@Void Master nvm with the mini lianli dynamic idea the mobo doesn't use 24pin iirc

You can but it's a proprietary board and a small PSU you could make a ATX shroud on the side and make the case look like a mini lian li dynamic (except the windowed panel or wait modding time)

Not while we're talking about folding it wont be!

PPD/W I think it was the 2080ti because of it's insane PPD, but I imagine that will soon be overtaken by the 3090 and 3080.
 
However it's not cost effective. PPD per watt per VALUE I believe is the 2060 Super. (since you can now get them for around $300-330 shipped on Hardwareswapped, I checked today)

I think it was the 1080 ti ?

So I can basically throw all the spare GPUs I have in one system and call it a day ? What a great choice for a multi-purpose, smart space heater.

"Loads 4 gtx 480 in quad SLI"

One of my friends actually have a 1070ti (not for sale tho), also need to figure out if I should keep the nightblade case or get a new one and mod it

Core. Memory can help with stable OC's, but F@H doesn't really use Vram at all. Ironically, it uses system RAM actually (up to 2gb per card in some cases) so my 8GB in ITNOS is almost not enough 😂

You assign threads in the Advanced Folding Client. As little as 1 and as many as all but one core for your OS. So theoretically a 3900x could have at max 23 threads dedicated to folding. 
 
Note that CPU folding is inefficient for the amount of points you get compared to how much power you use, so if your goal is lots of points, we recommend GPU folding instead. 
The 560ti is so old, it's not really worth the power to run it (a 750ti will get 75kppd, but a 560ti I'm only seeing between 20-40k ppd) I'd only use the 290x  (260k ppd) and the 1080ti (1.3m PPD)

Pretty much.
 
NVIDIA GPU's need at least one core or thread dedicated to assigning WU"s to the card though, plus at least one core to run the OS. So if you want to use three Nvidia cards, you need at least a quad core (or a dual core w/hyper threading)
Nope. @Gorgon has noted very minimal PPD losses using x1 and x4 PCI-E connections. 
Or, build a custom case to house it all

SLI has no impact on folding. Cards act independently of each other when folding, so enabling crossfire, NVLINK or SLI can actually hurt performance. 

"Loads 4 gtx 480 in quad SLI"

"Loads 4 gtx 480 in quad SLI"

"Loads 4 gtx 480 in quad SLI"

"Loads 4 gtx 480 in quad SLI"

*Loads GPU's with religious WU's*

well the 560 is priced at about $75 maybe I could get it down a little also shipping is included in the price

here you go

Feel the taste of the DERP GUN Motherland

I once had a seaggate HDD reporting errors in crystal disk info after 3 years of use. And even though I saw failure rate statistics, I didn't think it would die. It died within a year.
 
Come to think of it. Crystal disk info data is hard to make sense of. It always says good health even when read error rate clearly exceeds treshold.

I thought this thread will be locked forever lmao.

This doesn't necessarily mean that there will be good supply. You folks aren't looking at the other side of the equation.
 
Perhaps the demand will be low.