Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 455. (Read 2347664 times)

i have 2 evga 1070 (6173) burned.. out of a rig of 5. 

5173 is the Black Edition right ? without backplate

Did you check the memory brand ?

Yeah I bought three of those and resold them back due to very poor cooling performance (also talked about it in this thread a few times). The fans sit at 100% all the time due to the shroud wrapping the edges of the card completely and preventing air from exhausting from the heatsink. That can contribute to failures due to fans having to work harder to move any air at the same RPMs due to resistance. Cover the outlet on a fan and see how long it lasts, it's pretty close to simply holding down the fan and letting the motor burn out.

Not sure about the metal caps, other brands have metal caps as well and they don't seem to effect the longevity of the fans. Matter of a fact the only brand I have that doesn't have caps is Gigabyte and they definitely eventually go bad. The fans might just be bad in general, mixed with running at 100% and the exhaust being covered.

I own some three fan Zotacs and haven't had problems yet. Those don't have a shroud that completely covers the edge... we'll see. Fan issues usually take a half a year before the start developing.

ZOTAC GeForce® GTX 1070 AMP Edition
https://www.zotac.com/us/product/graphics_card/zotac-geforce-gtx-1070-amp-edition

The metal stickers movements which are on the fans effect the fan speed and balance. The same problem exist of my friend's zotac 1070 rig. There is no replacement fan to buy.

Gigabyte fans are pron to failure across all their cards (they're the half height cheap fans)

Which specific Zotac cards do you have? They have more then one model and I haven't had any problems with the three fan ones yet.

I have problem with zotac gtx 1070 amp fans. I have 7 cards and 4 of them have fan problems. 2 gigabyte 1070 g1 gaminghave fan problem too. Their nearest fans to the vga ports are not working. Zotac's fans are really crappy. if the metal stickers that on the fans move, fans begin to works slow and live balance problems. I contacted ith zotac but they don't want to send fan to replace them. Ne revisions have micron and rma processes are too slow in here.

It appears there are different revisions between 5173s as well, there are different component layouts. I checked a lot of them with my infrared gun, this specific revision appears to be the one with trouble. I checked another one of the boards and it was sitting at 112c in one spot(!), 108c on one of the components. None of my other cards get anywhere close to this.

Unfortunately the boards aren't labeled with a revision number so it's hard to tell which one is which, without looking at the layouts. Some of my older 5173s with different component layouts don't seem to have hotspots like this (theirs sit under 100c).

Had another EVGA 5173 burn out on me. I only had one failure until I started OCing their memory. The only thing I can think of is they didn't engineer the cards capable of handling full TDP including the memory. They were operating just fine for about a month and a half until I OC'd the memory like two weeks ago. Now I'm getting a chain of failures.

However, I have seen delayed failures which have nothing to do with OCs at all (MSI cards).

Failures goes like this > rig turns off > either refuses to turn back on until faulty card is removed OR shower of sparks that shoot out of the burned out component on the EVGA card.

This is the second one I've had fail specifically at this spot. I've had others fail, but it's under the heatsink and frontplate (above the PCB) so I can't determine where the sparks are coming from. I assumed it had to do with the thermal pad cooling issue, however, there is no way a thermal pad would help this little guy on the back of the card.

As I've mentioned the only failures I've been having are 5173s, not 6173s or FTW cards. It could be due to the backplate? It's hard to tell if the backplate actually makes contact with this specific component. Either way it looks like multiple things are going wrong with these cards. This failure is on the back of the card and there is no way a thermal pad would help. Read on.

These specific cards are kept very close to a air intake which is currently about 20f. There is no way they're overheating, this can only be an eventual stress failure due to utilizing too much voltage for either a underrated component or one that is of very low quality. However the front failure I'm pretty sure is different from the back failures I've been having. It's not just the fans spinning to low as well (back components aren't even cooled by the fan). I have them in machines further away from the intake and they do the same thing.





If you have these cards and you're running them at low TDP to save on efficiency you're playing with fire, quite literally. I run mine full TDP (112%). You probably can avoid failures by running them at 100% as well. I haven't tested this, but I assume so.

Either way, highly avoid getting these cards. So far I haven't had any failures with any of the other models though.

I just saw this post and registered to answer, even though it may be too late for the poster who asked.  If you try to run ccminer with the required address, wallet, etc., and it gives you a JSON error, you probably have ccminer installed from the repositories, and you are inadvertently running that one.  Instead, open a console in your new directory and run it as ./ccminer.  The ./ make all the difference in the world.  Then you'll be running the version inside that folder which should not cause the JSON error.  I just had and solved this same issue myself.

Using miners others don't have access to. -_-

The best to mine is what the others dont

What you guyz mining atm ? (most profitable)

Is it worth Zec/Zcl with EWBF (0.1.0b) or Nicehash ? (Suprnova, coinmine or Nicehash ? )

What is the best pool for Zec/Zcl and wich one between those two atm?

Thanks

Nope... No cigarette smoke. I have multiple fan failures over multiple Gigabyte cards (970 and 1070).

My guess is your cards have always remained at low RPMs and it never became a issue. It doesn't work that way if you have multiple cards in close proximity of each other (about 1" between cards) and multiple rigs.

So consider yourself lucky. XD

maybe your environment is toxic are you smoking weird products in the same room ? I know the fans dont like that at all.

Ive gigabyte cards since years, and they were sold in the exact same state as new at least when i sold them. Only some small dust easily cleaned with an air spray. Now remains only 2 750Ti Black Edition (which are working fine, non stop)

I could spend time to fix a faulty product, but I'm not taking time to add something that should be installed in the first place. That's like taking apart Gigabyte cards to oil them because of their broken fans. -_-

They'll fail, they'll get called out for what they do and the products they make. I wont buy them again and people reading this should become informed and not buy them as well.

Weird thing about this is it only seems to effect the 5173 (cheapest) models. I have 6173s and the FTW models and none of them have seem to have had this issue besides. even though they mention it effects all 1070 models. Although I don't have as many of those as the 5173s.

Best forever! thanks! .

https://github.com/KlausT/ccminer-cryptonight/releases

Anyone link me best miner for algo CryptoNote for nvidia and windows 10 x64 , compiled , thanks .

I would suggest you to add thermal pads to vrms, and cool them actively

a very small little pump, i would take it in consideration when it return to 2k+ and heading to 10k

in theory it should be pumped at least another time