damric

When your ACX fans disintegrate, you got to pull out the Wraith Stealth fans.

That one have voltage control or is that all just on stock voltage. What a beast!

Thank you sir!

Yeah I never have had problems with good old barbs with zip ties. I'm probably going to go back to using those. The cheap vinyl seems to have become too soft, though it's never been overheated or anything. I'm going to check out better materials when I rebuild.

Trust me I laugh at all of it, because there are funny parts everywhere. Poor Michele. She felt so bad that she went and bought me a new work light since my old one got lost. I think she went and bought me like 3 or 4 small presents today when she went out shopping. That light came in handy for inspecting the PCB. When I went to look at my radiators, man they were caked with dust, another hose fitting slipped out. I guess I'm lucky I didn't have disaster sooner with these cheap vinyl hoses. I just pulled the whole loop apart since it already drained on the floor I'm going to wash and flush everything and I threw out all of that trash vinyl hose. I'm going to get something a bit sturdier.

OMG. I think I fixed it. Ok so I noticed I lost some coolant from my reservoir on my ETF rig when I also noticed it was OFF. I pulled it apart last night and this morning. After looking it over it appeared that it leaked from a loose hose right on the CPU block (probably because I adjusted the frame's feet on Thursday and shook it around). I cleaned up the board but it still wouldn't act right. I suspected it leaked directly into the CPU socket, though it was hard to tell because it must have been a tiny amount of coolant and it must have evaporated quickly. The coolant I use is mostly distilled water, but it's most definitely conductive because of the metal additives I mixed. The titanium dioxide left a little white residue on the plastic socket cover where it dried. I knew I had to get that AM4 socket guard off so I could inspect and clean any residue. I've never been able to remove one of those without breaking it, but at this point I figured I already broke the mobo and maybe the CPU so what the hell might as well try. I watched a video or two on youtube where this little girl carefully takes a precision screwdriver to remove the plastic socket cover, and she made it look so easy. 5 minutes later I'm looking at plastic fragments of what used to be my socket cover but yes, there was some coolant stain on the metal socket connections. I cleaned them off with a fancy toothpick brush and some 91% isopropyl alcohol. I also cleaned the CPU pins. Next I took that CPU and what was left of the plastic pieces of socket cover and just kind of mashed it all together hoping for the best. I stuck the aluminum piece of a wraith stealth (I took all the fans off and use these for other stuff) on there, connected the power cables, keyboard and monitor, and jump started it with a screwdriver and bam into the BIOS we go. I dodged another bullet I think. I need to go ahead and assemble the rest of it, but while it's still mostly apart I'm going to clean my radiators really good.

@BWG when you get a chance, adjust my points and multi. I turned on the Vega 64 just after I posted last night. I'm using it at 50% power to keep the house from heating up so much and it's still pumping out 800k ppd.

@BWG @Supercrumpet Moving to Vega 64 until the end of the month. Entering my ETF key on that card now. Next month I'd like to fold on GTX 750, which while it isn't nearly as good of an overclocker as my old HD 7850, it's very efficient and won't heat this room up running it 24/7.

I've got some 95% iso. I'll try tomorrow. I need to see if I can pull the plastic socket guard off without breaking it so I can see what's going on in there.

I've done that. I pulled the CPU out of the socket, and sure enough there was a slight stain on the corner of the socket and also on that corner of the CPU. I'll reevaluate tomorrow when I get some more time.

I pulled the loop apart. There was a slightly loose hose connection on the CPU block. I think water got into the socket because it just doesn't want to come back to life. I'm going to throw the air cooler back on the HD 7850 tomorrow and see if I can get it to fold on my second slot of my main rig. This ETF stuff is really an endurance challenge.

Seriously can't wait for that board. As far as this board, I think it still works. I'm not sure what happened. There was some moisture near on one of the cpu fittings so slow leak maybe. I've got to tear everything down and rebuild it to really know for sure. It might be a loose fitting, but that water block was dirt cheap, well all of the water cooling parts on that thing were cheap and sketchy, which kind of made it exciting. I might have to switch to folding on my vega 64 just to complete this month.

Great. Another one shits the bed. I got Michele back up and running and this afternoon I check on MY ETF folding rig and it's down. I have no idea what happened. It just booted up and attempted windows auto-repair, so something disk related maybe, who knows. Great. Worse problem than I thought. reservoir a tad low...ofc it leaked near the CPU. This should be fun.

I'll send you good juju

Shared album - Luke Gelsinger - Google Photos PHOTOS.APP.GOO.GL

I don't have time to mess with it until maybe Monday, but at least it still has some overclock, but only 1050MHz which is the driver limit. I've fixed this issue before, I just forget how. There's a zero somewhere in the registry where there needs to be a 1 or something. It's always an assinine process trying to bypass the driver limits imposed by AMD.

It broke my unofficial overclocking mode in msi afterburner. So I am going to be limited to 1050MHz core until I figure out the workaround again. At least Client back up folding again

Thank you so much for that. I've had this problem a few times now over the past few months. I got the machine back up. I think I only lost a few hours. about to be folding again here soon.

Windows forced an update on my folding rig. I'm down a couple hours while I troubleshoot while this computer is not waking up.

It's not just a GTX 960 that is good at this. You should see crazy PPD I'm getting at 20 watts with a GTX 750 and it's set to "no preference". You should ban that card now, except banning cards is dumb.

There is a way to create handicaps without using LARS at all. Determine the multiplier by FLOPS based on stock clock speeds. For graphics cards that information is in the techpowerup database for just about any GPU ever made. CUDA seems more efficient but as long as the categories are split by AMD and Nvidia it shouldn't matter. We would need to find the data for CPUs or use linpacks to determine.

From what I have read on F@H, the server tries to assign WUs as best it can to a processor's ability. If they are sending more Moonshot units to GTX 960s because they are good at it, so be it. We need that Covid-19 research. It's not like anyone here can't buy a Maxwell card on ebay if they wanted to. Tomorrow there might be something that runs better on X GPU. We going to piss and moan over that? As far as I can see, only 1 person broke the rules in the ETF manual, and I sure hope he plugs his machine back in because folding 24/7 all month long is the rules. Just my 2 cents as a folding newbie and veteran overclocker.

I agree LARS is inaccurate. But I think we can take steps to make it more accurate. Make entrants fold at stock clocks for however long to get a baseline before a card can be entered. I use the "no preference" setting, but I don't see how setting it to a specific cause is cheating. Anyone can do that. Anyone can buy an ebay GTX 960 and moonshot all day for all I care. I think limiting 1 Nvidia card per team should handle that. Get rid of the "wildcard" and add a team slot for CPU folding. I don't pull any weight here, just my observations as a folding newbie and veteran overclocker.

What is the proposed rules change? No overclocking? Rules changes affect everyone and might make MORE people drop out, depending on what they are. I'm an overclocker. Folding is just a benchmark to me. I only joined this foray because I thought of it as an overclocking challenge, but as time went on, I got more interested in the folding aspect, and so did the rest of my family. I'm curious as to what other's perceptions were when they read the contest page and the ETF manual here on ExtremeHW. Did some team members think that running a 300W card with 3% OC headroom would be wise for this challenge? The scores have been tracked for months, and once my tweaker card was in working order it dominated, but that was plain for all to see. We also knew that GTX 960s would also be very competitive since they also generally have a lot of OC headroom, especially the models with voltage tuning. I've been researching, and there are some cards that will most probably perform even better in this challenge than HD 7850 or GTX 960. For instance, look at GTX 465. It has been cut down from the GTX 480 die twice over, which eliminates the weakest clocking cores. I would expect more than 900MHz easily, so 50% OC since stock is 600MHz. That equates to more GFLOPS, which powers folding, so it would be 50% more PPD.

Please finish it out. It's wonky, but it's for a good cause.