The new "died in the line of duty" thread

That's gotta be damn frustrating. I went the other way and bought something that already had a few years of abuse. I figured if it survived that long, it might stand up for a couple more years.
 
I had a Gigabyte 980 Ti that had issues with two of the three fans dying on me. Just used two 120mm case fans instead and it still worked fine.
 
Started RMA for my EVGA GTX 980 Ti, cannot fold for more than 6 hours now before getting black screen / errors...
 
True there are many reports of dedicated 980 Ti folders failing... don't know if it has to be linked to their great OC potential (you don't overclock yours right ?)
EVGA approved RMA I should be able to send the card this week end
 
CV yep, and my son's part time folder SEAHAWK had to be RMA'ed. I'm not sure that there is a need for super concern with the 980Ti. We may be seeing a lot of failures, but we also bought a bunch as well. Heck, I have had multiple GPUs fail while folding, GTX570, 2x R9 270, GTX970, GTX980Ti. I have had multiple more cards that folded flawlessly for years, GTX9800, 3xGTX260, 2xGTX570, 2xGTX960, GTX970, probably some I forgot, 9200s or something. Sure our failure rates are higher than normal, but we push them [H]ard 24/7/365. Buy enough of one card and you will see higher failure rates.
 
Does anyone know if this is easily fixed? It is on an EVGA 780 SC
20170131_162920.jpg
20170131_162930.jpg
20170131_163051.jpg


Edit: This doesn't sound promising... https://forums.geforce.com/default/topic/897543/geforce-basics/geforce-gtx780-board-damage/
 
Last edited:
Gilthanis,

Get yourself one of these https://www.sparkfun.com/products/retired/10622

That looks like it once was a resistor (on further thought, as the similar box next to it has the designation of an inductor it might be one of those, the L in L13 indicated what it might be I couldn't see the number for the burnt one), probably an open circuit now. It also looks like it got hot enough to melt the solder and disconnect it from the board. You might get a replacement, and your might even get it soldered in, but IMHO it probably is not the original problem. With that much damage, my money would be on something else shorted causing the resistor to pass too much current and overheating. I know the 780 was a behemoth once, but now, you probably could buy a 1060 with the electricity you would save not running the 780, and get as much science done at the same time.

BTW, I know how to solder, I would even hazard to say I am somewhat experianced, and that looks like it is a surface mount that would be a complete PITA to solder.
 
  • Like
Reactions: N4CR
like this
I'm guessing Voltage regulator as it was another R33. Yeah.. I will probably just pull the heat sync to swap with the 780ti I have here that has a bad fan.
 
I'm still standing with my after thought of inductor. R33 is on the part, thus it is some kind of rating. L13 is silk screened on the board next to the part labeled R33, that would be the circuit designator L for inductor.

Either way though, the magic smoke is no longer inside the component, and IMHO, it's probably not worth trying to put it back in.

Oh, and I just came across this

http://www.overclock.net/t/1363440/official-nvidia-geforce-gtx-titan-owners-club/21860

and this, which indicates problems with inductors getting excedingly hot and failing while mining.

http://www.overclock.net/t/1393791/official-nvidia-gtx-780-owners-club/20890

doing some further looking, these inductors are somehow part of the voltage regulation circuit.
 
Last edited:
My bet is a 33uH inductor. Inductors are just about the only component that comes in that kind of package.
 
Every damn ass end of the year I seem to lose something lately.

Last year, 7970 ref that OCd 925>1.3GHz out of the box. That was a sad day indeed, my favourite GPU ever.

This time, I lost my P8P67Pro B3 AND a 2600k (I've only once or twice ever replaced a CPU in thousands of PC insurance claims at my old job) to a fucking unitek POS USB3 fucking external HDD which had a current leak down the fucking USB cable shield. FUCKING FUCK YOU UNITEK.
And it killed two USB3 ports on a used gigabyte z68 mATX board I bought while figuring this out. It actually shocked me from the cable ground to case.. going to shoot that fucking POS.

Now I have to wait on a fucking i3 until Zen. FFFFFFFFFFFFFFFffffffffffffffffffffffffuuuuuuuuuuuuuuuuuuuuuUUUUUUUUUUUUUUUUUUUUUUUuuuuuuuuuuuuuuuuuuuuuuuuck!!!

/rant
 
Ow, that sucks! And I agree- my 7970 didn't OC at all, but only got replaced a couple of months ago (and I'm kinda wishing I waited for Vega).
 
It's not exactly dying but Ubuntu 16.04's wi fi connection keeps crashing and killing my productivity. I need to change my OS.
 
Finally, my replacement 980 Ti Hybrid was delivered by EVGA RMA. Had to reinstall FAHClient that did not like the card swapping, and back on line. I think I got the 980 Ti Hybrid VR Edition with some useless LED lighting (would have preferred a 1070 Hybrid :))
 
I had to take my rig off line while guests were here. Not exactly dead but am staying down until I kill off Ubuntu.
 
Well, it didn't even do any duty, one DOA Galaxy 1080 EX OC:cry:

And its replacement has now also died, I think t went a few weeks ago but i'bve only just spent the time to trouble shoot it, I'm RMA'ing it but i'm thinking of asking for a credit against another brand or downsizing to a lower model, evidently the extreme overclock doesn't like folding
 
And its replacement has now also died, I think t went a few weeks ago but i'bve only just spent the time to trouble shoot it, I'm RMA'ing it but i'm thinking of asking for a credit against another brand or downsizing to a lower model, evidently the extreme overclock doesn't like folding
Sorry to hear about your trouble ... but your conclusion seems right; dump this brand and go somewhere ... with more ... build-in confidence ... like MSI or Gigabyte ...
 
I had trouble with two different 7970's back in the day. RMA's had the same troubles as what I turned in. I think the refurb cards were basically all cards returned by miners and so didn't show their problems until stressed. So, they kept sending out bad cards. I eventually just sold them and moved on. No troubles with any other cards in those rigs.
 
Add an E5-2692v2 ES to the list of causalities. It was running in the board that died with the fried EPS socket last year so I wonder which took which out.Still have to test a donated 750ti as well, it works fine in windows but doesn't want to even boot in the Z9PE, but then that sums the Z9PE up, a right PITA
 
I had trouble with two different 7970's back in the day. RMA's had the same troubles as what I turned in. I think the refurb cards were basically all cards returned by miners and so didn't show their problems until stressed. So, they kept sending out bad cards. I eventually just sold them and moved on. No troubles with any other cards in those rigs.

The galaxy 970 that I had previously and the current 1070 work(ed) just fine, just the 1080 line up is crap.
 
Crappy cellphone pic in poor lighting.

Anyway, one of my 4P motherboard just died on me. Went to power it on just now and seen a quick flash (heard it also) turned around to see the motherboard/psu powered off and I immediately unplugged it. Took all the cables off and started inspecting it to see this.

Wonder if it's worth sending off to have it fixed?

G34 4P board

0615182113_HDR.jpg
 
Another 4P dies in the line of duty. These workhorses gave their all. *salute*
 
Not exactly dead as it's still CPU crunching for BOINC but the PCI-E slots stopped working on my HTPC mobo today under mysterious circumstances but mighta been coincidence: HD7770 I added for the SETI challenge was not working as planned (NVidia already in system folding) and I gave up fck with it so I decided to add a spare GTX750 I had lying around to boost a little more FAH. While doing the mandatory reboot Windows 10 decided to apply updates. Didn't repost repeatedly even after reseating the CPU. Only after pulling the CMOS battery and doing a complete BIOS reset did i get the onboard iGPU to give me a video signal whereupon Windows informed me it was rolling back the update. No amount of card/slot swapping would get either GTX to work, not even recognized in BIOS or dev.mgr. FCK! I'll keep her crunching for now but taking a hit at FAH just when I added a GTX 1060 to the mix.
 
Weird, I don't think I have ever seen all PCI-E's go... No magic smoke at all?
 
Weird, I don't think I have ever seen all PCI-E's go... No magic smoke at all?
Nope, no smoke and iGPU and all else is well. Might be a pin issue with the CPU but i doused it in alcohol and blew it down with compressed air before reseating it the last time and no gold. Had the problem crop once or twice in the past with this mobo but a reseat of the CPU always fixed it then. This time no mojo. Gotta Canyon quad CPU with mobo here listed on Kijiji for cheap, see if he goes for the offer. If so I'll get the farm boosted back up for FAH but I'll still crunch on the i7 as I have an empty box just waiting for something to power up. Might be willing to try linux on this fcked GPU box: i7 3770K, VM potential, but havent played around with it much unless you want to include XP mode on Win7 Pro or Knoppix back in the day. Ubuntu or what? Weigh in your ops, I do have a spare Win license but why suck it up for purely BOINC...
 
Weigh in your ops, I do have a spare Win license but why suck it up for purely BOINC...

i just use windows 10 unlicensed. You lose very few things and all cosmetic. Its technically not against the rules in and way and Microsoft isn't shutting anyone down
 
i just use windows 10 unlicensed. You lose very few things and all cosmetic. Its technically not against the rules in and way and Microsoft isn't shutting anyone down
Thx motq, I'll keep that in mind. Might opt for a bare bones i3 or Ryzen 3 box instead of older used Canyon quad. I can get brand spankin' new mobo and CPU for same price as he's asking with upgradeability options. DDR4 will be a huge ouch tho. If so I'll use the saved license. Tease myself with the linux idea once in awhile tho and actually have had several Ubuntu ISO's but just sat on them. Its a can of worms I'm afraid I won't put the lid on and I spend too much time on my systems as it is.
 
Thermaltake TR2 500W just died with a series of loud pops and the requisite magic smoke filled the air in the basement. Been powering the Server for the last few years 24/7 even before the Xeon upgrade. Hoping theres no mobo damage but cant be sure until I get a replacement PSU in there. Long weekend too motherfck. Gonna lose some FAH and BOINC production for a day or three. Had another almost new Thermaltake 600W die unexpectedly a couple-three years back too, think I'll steer clear of them from now on ...
 
We don't care about fah.

Some of us still do Skillz, GPU wise. Just look at the team contribs. Theres plenty of us still folding and crunching for BOINC. I had a buddy who succumbed to something thats being folded for and I wish I'd started folding sooner instead of looking for E.T. back in 2003 but all these DC projects weren't widespread or "advertised" or even existed back then. Take the monocle off ...
 
We don't care about fah.

Comments like this are not helpful, everyone in the [H]orde is free to take part in whatever DC projects they see fit, after all its about the science and not the bragging rights as to who "wins". The few still running F@H aren't constantly chasing the BOINC crew going join us we need help.....

And to echo Toconator's reply, no one cares about anything until it affects them or someone close to them.
 
Back
Top