The new "died in the line of duty" thread

Maybe it is linked to my professional history and current setup but I prefer FAH over any of those math projects ... wish more bio/health would show up in boinc. Not so crappy as Rosetta and more WU as GPUGrid.
While I don’t fold with cpu on FAH is because their beta version often not running on my centos stock versions and slow/no response on recompiling those.

Else: they can peacefully coexists . And for me they will.
 
Finally got my parts for the 2950x system, so i had time to take apart and figure out why my X99 was behaving weird... Found out that only 3 of 8 RAM slots work, and it is not down to overtightening. I guess i can call this state as "died". It's either the CPU or the motherboard - in any case RIP.
 
Put me down for a mobo too. When my PSU died last week it took the GA-150+ with it. Put in a spare OCZ PSU and she's deader than a doornail, its my Server too. Fck Thermaltake TR2-500 just under 2 yr old and under warranty and was plugged into a Cyberpower UPS so it wasnt a surge that took it. Thats the 2nd TR2 in last 2-3 yrs. Wont be recommending them to anyone anytime soon...
 
The database server driving hard-dc.com gave up one SSD out of its softy RAID1 , first responder tried but patient still in coma trying to get the newly implanted SSD running.
I hope the backups will work :-|

But I guess I use that opportunity to move full to a newer version of Postgres and switch the scheme to one using a partitioned tables. Based on time. That will improve the query performance.

But this means a downtime until weekend.
 
Add a MSI 1070Ti to the casualty list, died last week while folding. Bought as a refurb in feb last year so not the longest lived of cards. Not sure what i'm going to replace it with but probably a 2070
 
Nathan_P said:
Add a MSI 1070Ti to the casualty list, died last week while folding. Bought as a refurb in feb last year so not the longest lived of cards. Not sure what i'm going to replace it with but probably a 2070
Click to expand...
Warranty?
 
This is why i prefer hunting down good used deals. The 1070ti's i bought this year still had 1.5 years warranty left from EVGA when i picked them up
Just need to by the cards from manufactures that allow warranty transfer.
 
I lost a few more ARM devices this past week. Still migrating some from other locations though. Had to utilize power strips as I setn my 100 port USB charger back.
 
motqalden said:
This is why i prefer hunting down good used deals. The 1070ti's i bought this year still had 1.5 years warranty left from EVGA when i picked them up
Just need to by the cards from manufactures that allow warranty transfer.
Click to expand...

True but I bought this in feb 2018, which being as it was the height of the mining fad meant that buying anything was nearly impossible. it was a gamble and I lost on thins occasion, my other refurb was a 1080 bought in oct 2017 and its still going strong.

Lesson learnt and i will only be buying new from now on
 
^ Mining not a fad, just moved on to purpose built bare metal ASIC hardware. That said, I was thinking about adding a 2nd 1080ti to my primary setup, thinking that at two years old the prices would have come down, nope, $1200, just what I paid for this one...
 
Dual core Dell Inspiron died today. Pulling usable parts and replacing with a Dell Latitude E6410 Quad core. So, good news is that I went up 2 more cores in the grand scheme of things.
 
The i5 in my sig is no more. The OC had degraded about 200mhz In late 2018, so I replaced it with a 2700X on Black Friday 2018, moved Over the 1080ti, and left it powered off in a corner. I’ve been pulling out all the stops with hardware I’ve got laying around for Folding and Rosetta and figured I would remove the OC and let it run at stock.

Plugged it in tonight, hit the power button and “pop” the lights went off. Reset the breaker, tried a new PSU, and nothing. Either the mobo or CPU is shot, and I don’t have the hardware on hand to figure out if it is one, the other, or both.
 
jojo69 said:
yikes
Click to expand...
I ran that host on the absolute edge of what it would run at for over 5 years, so I’m not surprised it smoked. I am a little surprised it popped the breaker though - wonder what would cause that.

Normally I don’t run anything with that much of an OC for more than 18 months, but there just wasn’t anything good enough to justify an upgrade for years, so I just kept using it.
 
And another one down - I just set my core2quad 9650 to "No New Tasks". Purchased in 2008, I ran it over clocked as far as it would go until roughly June 2012 when I picked up my Ivy Bridge host. It ran 24x7 as a server since, and has been running rosetta@home 100% since March 28th. It first crashed 2 weeks ago, and is now only staying up about 12 hours at a time before crashing. I think it's probably just a bad stick of ram, but given that 2 Raspberry Pi 4s do the same amount of work for 1/10th the power draw, there isn't much reason to keep it online.
 
Endgame said:
And another one down - I just set my core2quad 9650 to "No New Tasks". Purchased in 2008, I ran it over clocked as far as it would go until roughly June 2012 when I picked up my Ivy Bridge host. It ran 24x7 as a server since, and has been running rosetta@home 100% since March 28th. It first crashed 2 weeks ago, and is now only staying up about 12 hours at a time before crashing. I think it's probably just a bad stick of ram, but given that 2 Raspberry Pi 4s do the same amount of work for 1/10th the power draw, there isn't much reason to keep it online.
Click to expand...
Put it on Facebook Marketplace as a "low end gaming machine". I cannot believe how many old systems are getting sold being labeled like that.
 
980Ti from RFGuy_KCCO finally kicked the bucket. Now until inventory of the latest AMD Gen cards become available, I am down to extremely old cards.
 
Hey, didn't I send you an RX470 or something like that? lol
 
X370 Prime Pro did not seem to like running a 3900x....
This board has had a lot of miles on it as it's been running 24 / 7 since i bought it. upgraded the CPU from R1700 to 3900x 2 weeks ago and friday she died when my VRM blew up.
Was running Rosetta@home when it died :(


20210313_102457.jpg

20210313_102644resized.jpg
 

Attachments

  • 20210313_102641.jpg
    20210313_102641.jpg
    263.9 KB · Views: 0
Motherboard died on the dual 2011 setup. Gotta do more testing on chips but I actually saw the flame when it happened.
 

Attachments

  • IMG_20211015_162928.jpg
    IMG_20211015_162928.jpg
    256.5 KB · Views: 0
  • IMG_20211015_155606.jpg
    IMG_20211015_155606.jpg
    436.4 KB · Views: 0
jojo69 said:
like you can buy anything worth a damn right now...
Click to expand...
Get gud scrub. CPUs are pretty easy to come by, gpus are a bit harder.
Also I don't see Coleslaw getting current gen, just maybe upping to 2011-3.
 
Don't worry, I sent him another copy of the exact same board, just slightly less broken.
 
Sadly, the Antec NeoHE 650 which had been in almost constant use for 13 years or so and much of it crunching died tonight. Took my main system and my server down to the shed to blow the dust out (and probably lost at least a pound in each system) and when I hooked everything back up my server wouldn't power on. After much troubleshooting I hooked up an even older Antec NeoHE 500 and it started right up so my server is up and running again.

There was some luck, though. Both power supplies are semi-modular and an amazing and unprecedented never before happening thing happened. I looked and the pinout for the modular connectors (SATA and Molex) were the same. There was no need to completely rewire the system.

Okay, it's a bit of an exaggeration to say it's never happened before that the modular connectors had the same pinout but it's pretty damn rare.
 
Patriot said:
Get gud scrub. CPUs are pretty easy to come by, gpus are a bit harder.
Also I don't see Coleslaw getting current gen, just maybe upping to 2011-3.
Click to expand...
Sadly EXT64 's identical slightly less broken board also no worky. I don't plan on adding DDR4 based rigs unless on the super cheap as I've been waiting to see what DDR5 era has to bring. I only have one DDR4 box. Honestly, the DDR3 stuff works just fine on a lot of DC stuff. The 2011 v1/v2 boards are just still to this day crazy stupid expensive. If you want to send me any "good" setups, I'm all ears...lol.
 
Gilthanis said:
Sadly EXT64 's identical slightly less broken board also no worky. I don't plan on adding DDR4 based rigs unless on the super cheap as I've been waiting to see what DDR5 era has to bring. I only have one DDR4 box. Honestly, the DDR3 stuff works just fine on a lot of DC stuff. The 2011 v1/v2 boards are just still to this day crazy stupid expensive. If you want to send me any "good" setups, I'm all ears...lol.
Click to expand...
Well...I thought that I had resurrected one of the 2 sockets and put a low power draw chip in it. It lasted 5 days and now the other socket is burned up as well. RIP my friend. You certainly served your time in battle.
 

Attachments

  • PXL_20211220_201818816.MP.jpg
    PXL_20211220_201818816.MP.jpg
    297.9 KB · Views: 1
racercarl said:
Had one boxen die today, the caps in the motherboard are leaking, hangs up when trying to start in a different part everytime. :(

I wonder if the caps off my other dead motherboard will work, must check this out as it would be to easy of a fix :D
Click to expand...
A long long time ago in a galaxy far way to small i started this thread. Yesterday I retired the last of the old folding boxen, it has served as my main box from when i joined up until it died last week, the only thing I salvaged was my [H] badges. the other 37 box all went away when Stanford changed the work units, they were to slow. almost all went to the Boys and Girls Club in my town.
 
racercarl said:
A long long time ago in a galaxy far way to small i started this thread. Yesterday I retired the last of the old folding boxen, it has served as my main box from when i joined up until it died last week, the only thing I salvaged was my [H] badges. the other 37 box all went away when Stanford changed the work units, they were to slow. almost all went to the Boys and Girls Club in my town.
Click to expand...

16 years goes by way to fast…. Seems like a generation ago, but it’s actually been many generations in the [H]ardFolding world. <salute>
 
Not yet died but limping. This SSD (PNY CS1311) is showing symptoms of reaching end-of-life with occasional reboot and bad results returned. I bought this back in 2016. It is based on TLC, so not the best NAND cells. Almost 69TB writes spanning over 2.8 years running various distributed computing (DC) projects. Yes, some DC projects can really kill your SSD very quickly.

I'm still pushing it but will need to replace this very soon....

I can't find any published data by PNY on the endurance life for this model. Please share if anyone know this.

P.S. This is my xeon haswell system and has been running rock solid since day one. I have not yet eliminated if the issue is related to cpu, mb or ram. The only one that stands out is the SSD and it says 0% health. There is a possibility that I could be wrong since I don't really know what is the expected SSD endurance for this model. Is CrystalDiskInfo result reliable?


1641240762881.png
 
You must log in or register to reply here.
Back
Top