Started ~12 hours ago on a FX-8100 sys and have made ~2800 pts.
Patience is more than a virtue with this, but I’ll leave it running for a few days.
Oh yeah, it is the browser based and not desktop version.
1 Like
Winter is the best time to fold
1 Like
It’s an old HP desktop that I don’t mind if it burns out. I’m going to use it for some experimenting with a custom liquid cooling setup in a custom case and will more than likely die in a fire at some point. Just have to wait until summer is finally here to test the ~36+ Celsius Room Temps.
The folding is running at max in the browser.
Current temps (air cooling with stock CPU heatsink):
Room: ~27 Celsius;
CPU: ~60 Celsius;
Before a good cleanout the CPU was running at ~70 Celsius. If all goes well, I’ll be setting up a couple of dedicated machines for this task. (Obv. not old HP machines)
1 Like
would love to see pictures, maybe a build thread?
Yeah I might end up doing something like a build thread here or somewhere else. It won’t be for a while though. Many numbers to crunch and metal to bend.
1 Like
Abysmally Low PPD on Win10/Ryzen 7 1700X
I have no clue why, but I am getting ridiculously bad folding performance on my CPU lately. I was just using the GPU for a while, so I don’t know when this molasses infected my system. The client works fine for my GTX 1070 (about 90,000 PPD), but when using the CPU slot, I get much less than 10,000 PPD. It’s using all the cores and burning all the power. Am I missing some little obvious thing? When just using the Chrome browser plug-in, shows around 46,000 PPD.
I seem to remember that Ryzen should be getting much higher numbers.
I reinstalled the client to no avail.
Ryzen 7 1700X Stock Settings except vitualization enabled.
Asrock X370 Taichi
32GB RAM
Folding@home client 7.4.4
Windows 10 Pro Build 15063.632
Any ideas?
Have you entered your passkey in the FaH client?
I can already answer that. The 1070 gets around 100K PPD without a passkey, 600-700K PPD with passkey.
That being said, the CPU score doesn’t really look abnormal to me. CPUs just don’t have anywhere near as much power as GPUs.
1 Like
I guess I missed the chapter on passkeys. 
So the answer is no, I didn’t.
I got my passkey and entered it. Things are much better now, thanks!
PPD = 56K.
That’s a huge difference.
1 Like
Win10 VM with 630 in pass through, folding shows a GPU failure- are there nvidia drivers I should roll back to?
1 Like
I was getting a fail on straight Win 10 with a GTX 1070 using the 388.13 driver. So I arbitrarily rolled back to 385.69 and things seem to work now.
1 Like
Dunno it this counts as necro, think it isn’t as this should be an ongoing topic?
Apologies if it is.
Didn’t know we had a team until this post just got quoted. So installed and got my desktop up and running again, we have some ranks to climb on the leaderboards!
3 Likes
Go team go!
3 Likes
Folding pc has been shelved will be a nas soon 
2 Likes
I have been on BOINC running IBM World Community Grid non-stop since joining in 2007.
Might be persuaded … what does Folding At Home attack?
1 Like
It depends on the project, but I think that the big ones are Alzheimer’s and Parkinson’s right now. Basically, we know that proteins are very important to these diseases, as they are likely protein-mediated. The problem is that we don’t know what the proteins look like. Many of them are too big to resolve structures of in traditional ways (X-Ray crystallography, Cryo-Electron Microscopy, etc). The other problem is that we don’t know if the structures that we do see, from those that we can resolve structures of, are what the proteins look like when they are pathogenic. Proteins are dynamic, they move around between multiple conformations. One of the issues with Alzherimer’s is that proteins that normally behave OK begin to cluster together and form a plaque. This means something is happening the cell thermodynamically that makes it more favorable for the proteins to assume some conformation that enables their aggregation and causes the disease. We can’t resolve that because we don’t know what those conditions are or how to replicate them.
That brings us to the problem that Folding@Home helps to solve. We know the amino acid sequence of the protein, and we know that there’s a finite number of possible arrangements of that sequence (1x10^n, (n= number of amino acids) - Kinetic impossibilities). The problem is that there’s so many possibilities, and then not only do we need to figure out what they are, but also which are thermodynamically favorable, and how they work. Not to mention that every angle of every amino acid residue can do things like change the dielectric constant of the water around it, and that adds additional layers of complexity. The truth is we just don’t have the computing power centralized anywhere in the world that is capable of figuring these problems out ab initio, and so we need a solution like F@H to distribute that computing load.
We probably won’t cure any diseases with F@H, but what we will do is contribute towards our understanding of ab initio protein folding models and our understanding of what the thermodynamics of water look like inside a globular protein.
4 Likes
3 Likes
Enable “advanced mode” to participate in Covid-19 research…
1 Like
I already had my folding rig set to “any”. Just checked, and my GPUs are working on Covid-related WUs.
Might let my main rig’s GPU join in.
1 Like