Genesis Mining Presents: SGMiner-GM - now with Zawawa's GG! [Updated 17/01/2017]

[wacko]

All my rigs are fine with claymore (in fact I NEVER had 100% fan with claymore xmr) and ALL sgminer forks have this problem for me.
Some of my rigs are mixed amd generations, some are mixed rx470 & 480, and some are exclusively 470. (REFs, Nitros and a few Asuses

Do you have any rigs that have no RX 470 cards in them? There might be a problem with (some?) RX 470 cards not working properly either by themselves or with RX 480 cards in the same rig. That's just an idea, of course, I'm still trying to figure out this thing myself and those "theories" of mine are a long shot  , but still might be worth checking out. My "rx 480 only" rig seems to work fine so far, I'll run it with sgminer from now on and see whether it's stable for more than just a few hours.

[1715222804]

1715222804

[1715222804]

1715222804

[1715222804]

1715222804

[1715222804]

1715222804

[1715222804]

1715222804

[Prelude]

Removed 17.1.2 this morning with DDU, tried installing 16.11.4 a few times but it would crash half way through driver installation. Gave up on that, and installed 16.11.5 after another DDU cleanup. I'm going to preemptively say the issue is solved. SGMiner has been running for 7+ hours and the fans are behaving as they should. With 17.x.x they'd go max RPM within about 2 hours, then eventually the fan on one of the GPUs would actually turn its self off. That caused the GPU to overheat, reach thermal limit and turn off the rig to protect its self. For the past 2 weeks I've been trying to figure out why it would turn off, and figured the PSU was to blame. I ended up figuring out the aforementioned problem by leaving VNC connected to it on another rig.

@Prelude
What's more interesting is why it doesn't happened on your other rig. Can you please describe the setup?
What's so special about it...Are we talking XMR? are those rx400? what OS is that? Are cards modded? how about voltage?

Rig that wasn't having fan issues:

4x Reference RX480 + 2x Reference style Sapphire RX470, all 6 with modded BIOS.
Win 10 Pro X64 / 16.11.4 driver
Mining XMR with SGMiner 5.5.5-gm

Stable for weeks on end

Rig with fan issues:
4x Reference RX480 with modded BIOS.
Win 10 Pro X64 / 17.1.1 and 17.1.2 driver, both same issue
Mining XMR with SGMiner 5.5.5-gm

Would mine for 2 hours or so before 1 or more GPUs' fans went full throttle. Eventually one of the fans would turn off completely, causing over temp protection to kick in and power down rig. Claymore would work perfectly for as long as I let it, no fan issues.



 

[Ursul0]

@Prelude
Well. I hope you solved the issue for yourself with driver reinstall.
If I recall correctly sometimes SGM XMR managed to work for a day or so without 100% ... 6+ hours for sure.
Do you run it undervolted? (Becasue I do - ETH@850mv, ZEC@910mv and XMR @910-925)

[Prelude]

@Prelude
Well. I hope you solved the issue for yourself with driver reinstall.
If I recall correctly sometimes SGM XMR managed to work for a day or so without 100% ... 6+ hours for sure.
Do you run it undervolted? (Becasue I do - ETH@850mv, ZEC@910mv and XMR @910-925)

Have you tried with 16.11.4/5? I'm closing in on 9 hours now without issue. Will report back after 24h.

Yes, undervolted.  Running them at 1250MHz @ 975mV before vdroop.

[wacko]

Yeah, I also thought my 2nd rig with 3x rx 480 was stable with no issues, left it for the night, and after I returned back to it today — had 0-100% fan issue with one card. That's after ~12 hours of running. The fan doesn't actually spin 100% all the time, it goes to 100% for a minute or so, and then drops to 0% (stops completely), then a minute or so later - back to 100% again.

It might be that the card is simply not stable and that's how sgminer-gm crashes? The card that was having 100% fan issues today had the "DEAD" status in sgminer. So I dropped the clocks a bit and testing it again. I'm getting higher cryptonight hashrate with this sgminer than with any other miner, so it could be just stressing the cards more than other miners and thus the clocks/voltages that are stable with other miners might not be stable with sgminer. When a card crashes - we get this 100% fan issue. That's my new working theory anyway.

[batko]

Best setings for RX 480 --rawintesity - w - g?

[Prelude]

Yeah, I also thought my 2nd rig with 3x rx 480 was stable with no issues, left it for the night, and after I returned back to it today — had 0-100% fan issue with one card. That's after ~12 hours of running. The fan doesn't actually spin 100% all the time, it goes to 100% for a minute or so, and then drops to 0% (stops completely), then a minute or so later - back to 100% again.

It might be that the card is simply not stable and that's how sgminer-gm crashes? The card that was having 100% fan issues today had the "DEAD" status in sgminer. So I dropped the clocks a bit and testing it again. I'm getting higher cryptonight hashrate with this sgminer than with any other miner, so it could be just stressing the cards more than other miners and thus the clocks/voltages that are stable with other miners might not be stable with sgminer. When a card crashes - we get this 100% fan issue. That's my new working theory anyway.

No, that's not the issue. It's a bug with sgminer and some driver versions. I say the bug is with sgminer because claymore doesn't have issues with the fans.

I'm over 24h without fan issues now, problem solved. Use 16.11.4 or 16.11.5. Hopefully the sgminer bug can get sorted, I'd like to use the latest drivers on my gaming rig.

[marvykkio]

new 17.2.1 amd

[wacko]

No, that's not the issue. It's a bug with sgminer and some driver versions. I say the bug is with sgminer because claymore doesn't have issues with the fans.

I'm over 24h without fan issues now, problem solved. Use 16.11.4 or 16.11.5. Hopefully the sgminer bug can get sorted, I'd like to use the latest drivers on my gaming rig.

Nope, installed 16.11.4 this morning, and it was pretty stable for the whole day, but just a few minutes ago got 100% fan with one of my rx 480/4 nitro cards. There might be some correlation between the driver version and how often this 100% fan thing happens, but it's not just the driver. I still think it's card(s) being not 100% stable. I'm running HWiNFO now in the background, looking for "GPU memory errors". Had to actually drop my memclocks quite significantly before those errors stopped occurring. I thought I was 100% stable at 2150 memclock (1750 strap timings, Samsung chips) for most of the cards, had absolutely zero issues with other miners, but I never checked that counter in HWiNFO before. This sgminer-gm problem made me look for solutions and I've just discovered this HWiNFO thing only today, so not sure yet how it works and whether it's normal to have some errors or there shouldn't be any at all.

[Ursul0]

No, that's not the issue. It's a bug with sgminer and some driver versions. I say the bug is with sgminer because claymore doesn't have issues with the fans.

I'm over 24h without fan issues now, problem solved. Use 16.11.4 or 16.11.5. Hopefully the sgminer bug can get sorted, I'd like to use the latest drivers on my gaming rig.

Nope, installed 16.11.4 this morning, and it was pretty stable for the whole day, but just a few minutes ago got 100% fan with one of my rx 480/4 nitro cards. There might be some correlation between the driver version and how often this 100% fan thing happens, but it's not just the driver. I still think it's card(s) being not 100% stable. I'm running HWiNFO now in the background, looking for "GPU memory errors". Had to actually drop my memclocks quite significantly before those errors stopped occurring. I thought I was 100% stable at 2150 memclock (1750 strap timings, Samsung chips) for most of the cards, had absolutely zero issues with other miners, but I never checked that counter in HWiNFO before. This sgminer-gm problem made me look for solutions and I've just discovered this HWiNFO thing only today, so not sure yet how it works and whether it's normal to have some errors or there shouldn't be any at all.

HWinfo shows errors which were corrected as well as those that were not.
Dagger-Hashimoto is the most tolerant for these kind of errors (probably due to memory READ intensive operations) Samsung memory is the most stable in this regard...
this is how two UV 470 refs (hynix and samsung) look after 10hours of operations both doing dual ETH@28+& DCR@320+

Basically for ETH mem error is not an issue.
For XMR it's a different story, yet I'm getting rejects even when mem errors are at 0...

[ronin82srb]

Need help!
Hi guys, I had a similar problem with 100% fans, mining on win10pro with 6 x msi rx470 4gb, 4 of them on ga970 with sempron x2, 2 of them on h81m with g1820, both rigs have 4gb ram. I used 16.9.2 drivers on both mashines for eth and there were no problems ever. First time switch to xmr I got 100% fan speed on both rigs runing claymore 9.7. Instaled 16.11.4 driver and now fans are ok - no 100%, but sempron mashine frezes after 1 or 2 hours mining, while Intel mashine works perfectly. All 6 gpus are moded, gpu 1100, mem 1900, core volt -480mV, low intensity mode.
What to do next?

[wacko]

Basically for ETH mem error is not an issue.
For XMR it's a different story, yet I'm getting rejects even when mem errors are at 0...

Thanks. I kept lowering memclock on all my 4 cards until I've got 0 errors. Had to go down quite significantly. Now there's 0 errors, but I've still had 2 cards out of 4 having that 0-100% fan issue this night. So I just give up, spent quite a lot of time on all these tests, but no point mining with sgminer at a bit higher hashrate if the rigs keep going down every day (I think the cards having 0-100% fan issue either stop mining when the problem shows up, or continue mining at very low rate). Funny thing, after playing with different drivers versions it looks like my "platform" id has changed, it's now "1" instead of "0", and I can't even use the wolf's miner anymore. It doesn't have an option to select the platform, does it?

gpu 1100, mem 1900, core volt -480mV

What was the stock voltage? -480 mV is a lot, even for 1100 MHz, some of your cards are probably not stable. GPU isn't stressed that much when mining ETH so you were able to get away with this low voltage before, but it might be not enough for other algos like cryptonight.

[Prelude]

No, that's not the issue. It's a bug with sgminer and some driver versions. I say the bug is with sgminer because claymore doesn't have issues with the fans.

I'm over 24h without fan issues now, problem solved. Use 16.11.4 or 16.11.5. Hopefully the sgminer bug can get sorted, I'd like to use the latest drivers on my gaming rig.

Nope, installed 16.11.4 this morning, and it was pretty stable for the whole day, but just a few minutes ago got 100% fan with one of my rx 480/4 nitro cards. There might be some correlation between the driver version and how often this 100% fan thing happens, but it's not just the driver. I still think it's card(s) being not 100% stable. I'm running HWiNFO now in the background, looking for "GPU memory errors". Had to actually drop my memclocks quite significantly before those errors stopped occurring. I thought I was 100% stable at 2150 memclock (1750 strap timings, Samsung chips) for most of the cards, had absolutely zero issues with other miners, but I never checked that counter in HWiNFO before. This sgminer-gm problem made me look for solutions and I've just discovered this HWiNFO thing only today, so not sure yet how it works and whether it's normal to have some errors or there shouldn't be any at all.

Weird, mine has been perfect since changing. Maybe the Nitro is different than the reference card in some regard. Consider using claymore until this gets sorted. You'll be a bit slower, but avoiding downtime will more than make up for that.

[wacko]

Yep, switched to claymore for now. Gonna reinstall windows today and hopefully will get wolf's miner working since it was faster for me than claymore.

[ronin82srb]

Thanks for fast reply Wacko, my fault, my cards are on -48mV , under mining gpuz shows 0.9813V, in iddle shows 1.0V. I dont remember what was the stock core voltage, I tried to set core to "0" mV in msi afterburner - no fix, still frezing after few hours , also tried to underclock the memory, but no difference except the lower hasrate. First I will underclock the gpu for 50mhz to see how stable it works. Thanks again!

[Tmdz]

Does anyone know how to make sgminer-GM work on nicehash?

It gets 100% rejected shares.

On any other pool it works perfect. (eg nano or dwarf)

[citronick]

Does anyone know how to make sgminer-GM work on nicehash?

It gets 100% rejected shares.

On any other pool it works perfect. (eg nano or dwarf)

There is a sgminer-gm v5.5.5-4 by NH team, that is suppose to fix the cryptonight issues..... but I still get 100% rejects too (under Linux). Any ideas?

[Ursul0]

regarding this 100% fan issue....
well... I don't want to be overly optimistic but it does look like 16.11.4 fixes the problem! (DDU and manual drv install)  Thanks @Prelude!
Really weird, cause I'm 100% sure I tested 16.11.5 and the bug was there.
Anyway running since yesterday GG fork on several machines and it's all good.

EDIT: now when I'm finally able to run xmr on SGminer I'm sure Claymore would soon update his miner...

[Prelude]

regarding this 100% fan issue....
well... I don't want to be overly optimistic but it does look like 16.11.4 fixes the problem! (DDU and manual drv install)  Thanks @Prelude!
Really weird, cause I'm 100% sure I tested 16.11.5 and the bug was there.
Anyway running since yesterday GG fork on several machines and it's all good.

EDIT: now when I'm finally able to run xmr on SGminer I'm sure Claymore would soon update his miner...

Awesome, glad to hear it. Both my RX rigs are working perfectly with 16.11.4/5

[go6ooo1212]

Which are the usual scores with RX470/80 on monero