Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[skablast]

Mining- x16r  isn't hard with the current hardware, some of the 16 algos stresses the core, other the memory.
I don't see how Dual mining can be achieved and making sense.

[angelesmix]

Mining- x16r  isn't hard with the current hardware, some of the 16 algos stresses the core, other the memory.
I don't see how Dual mining can be achieved and making sense.

ok.
What about Keccak-C ?

[Mashy81]

Update my 2 rigs to 11.5.
Presently surprised. Stable and I am able to run higher memory clocks with 11.5.
Vega 44mhs 1108 core and 1120 memory, rx570 now 30.6mhs 1140 core and 2080 memory, rx550 now 14mhs  1150 core 1930 memory.
All up I gained 10mhs over 2 rigs. Could push clocks higher on 550 and 570s but not worth the extra power
Best update I have seen from clamore thanks dude keep up the great work.

570s and 550s bios modded
Latetest amd drivers.

now you wait for incorrect shares to pop up. i hope not.

No incorrects So far and also averaging 7% above reported hashrate on nanopool

[nbruce]

Update my 2 rigs to 11.5.
Presently surprised. Stable and I am able to run higher memory clocks with 11.5.
Vega 44mhs 1108 core and 1120 memory, rx570 now 30.6mhs 1140 core and 2080 memory, rx550 now 14mhs  1150 core 1930 memory.
All up I gained 10mhs over 2 rigs. Could push clocks higher on 550 and 570s but not worth the extra power
Best update I have seen from clamore thanks dude keep up the great work.

570s and 550s bios modded
Latetest amd drivers.

Updated from which version?

[pogsnet]

Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v11.5 is very nice,

Running very fine on RX 480 win10 64, Catalyst 17.12.2

I got more hashrate and more stable, less wrong share

[pogsnet]

Sorry I take back what I said.


I got much better hashrate than v11.0 but v11.5 is less stable

[dragonmike]

Has anyone experienced hashrate drop in what seems to be a quite consistent matter?
I'm not talking about my rock solid RX570 rigs, they've been absolute beasts for a year now. But I'm struggling with the Vega rig.

All GPUs will hash at ~41MH/s until one decides to drop to 26 for a minute. Then goes back up to 41. This happens regularly.

Surely if it were temp throttling (which seems unlikely as I have the fans running at pretty much 100%) the hashrate reduction would be more gradual... and not bam!... down to 26?

Any clues? I'm on v11.4, latest Adrenalin, Vega56@64, safe softPP table with specific clocks and voltages set in Claymore miner.

Got the same issue but with my rx580 rig, all 7 cards running stable at 31 mh/s. From time to time gpu0 suddenly crashes to a rate around 5-10 mh/s and after a few minutes hits back to 31. There are no error logs showing up, also no incorrect shares and temps are fine. Card has also the same setup as other cards (no oc), kinda strange since it just appeared lately, running currently v 11.5

here you can see it: https://imgur.com/iQeWpKJ

same issue here

are you using stock rom  with no oc?

In addition to the above, Mr Claymore, your miner reboots my rig every hour. Almost without fail. There's no error in the log, no watchdog issue with GPUs OpenCL issues, nothing. I'm baffled.

[nemzy]

rx 470 strix 4gb - crashing all the time with new 11.5
11.0 working perfect , 20 days and no restart. 11.5 crashing !
windows 8 and 10 - same result
yes i have 174-175 mh/s but is not stable 

[SKBitCoinX]

is it crashing when it runs DEV FEE for Claymore?  re-loading the DAG? 
I had same issue and I think i have it working, will know later today, so far no crash on Pirl.

rx 470 strix 4gb - crashing all the time with new 11.5
11.0 working perfect , 20 days and no restart. 11.5 crashing !
windows 8 and 10 - same result
yes i have 174-175 mh/s but is not stable 

[Claymore]

Has anyone experienced hashrate drop in what seems to be a quite consistent matter?
I'm not talking about my rock solid RX570 rigs, they've been absolute beasts for a year now. But I'm struggling with the Vega rig.

All GPUs will hash at ~41MH/s until one decides to drop to 26 for a minute. Then goes back up to 41. This happens regularly.

Surely if it were temp throttling (which seems unlikely as I have the fans running at pretty much 100%) the hashrate reduction would be more gradual... and not bam!... down to 26?

Any clues? I'm on v11.4, latest Adrenalin, Vega56@64, safe softPP table with specific clocks and voltages set in Claymore miner.

Got the same issue but with my rx580 rig, all 7 cards running stable at 31 mh/s. From time to time gpu0 suddenly crashes to a rate around 5-10 mh/s and after a few minutes hits back to 31. There are no error logs showing up, also no incorrect shares and temps are fine. Card has also the same setup as other cards (no oc), kinda strange since it just appeared lately, running currently v 11.5

same issue here
are you using stock rom  with no oc?

In addition to the above, Mr Claymore, your miner reboots my rig every hour. Almost without fail. There's no error in the log, no watchdog issue with GPUs OpenCL issues, nothing. I'm baffled.

1. Unstable speed - I see this issue sometimes on different cards, but I have no idea about the reason - miner executes jobs on GPUs in the same way all the time, but sometimes GPU becomes slower for some time. I noticed that often different "-ethi" value helps, try "-ethi 4" or "-ethi 12" and see if it helps. I think that it must be something related to AMD drivers but I cannot confirm it yet.
2. I have a few rigs on v11.5 and they are working for days until I restart miner to perform some new tests, but I always use non-overclocked cards and default bioses, I have to do it to exclude everything except miner in case of any problems. When you use miner on overclocked/unstable rig you cannot be sure that the reason of the problems is the miner, but of course you will blame it Try to reduce "-ethi" value, if it does not help you can use the version that works fine for you.

[RooiWillie]

is it crashing when it runs DEV FEE for Claymore?  re-loading the DAG? 
I had same issue and I think i have it working, will know later today, so far no crash on Pirl.

I have not had a crash yet. But in once instance, after mining the DEV Fee, I had a serious drop in hash rate across all GPUs (RX470, RX480, 2 x GTX 1050ti).

And it only happened once thus far.

I have also noticed the fan speed is reported as 0 once in a while, then Polly restarts the rig, although it was quite unneccesary. When I used multi pool miner some time ago, the API also reported zero fan speeds at random intervals.

Finally, stale shares are ranging from 4% to 15%. But overall hash rate increase seems to outweigh this by about 0.0001 ETH per day.

[Vann]

My rigs run very stable with v11.5 and the latest Adrenaline drivers. I think for the most part, those having problems are using dodgy or a copy/paste hack job Bios mods. I'm running the Polaris Bios Editor 'one click timing patch' on my cards and it works great. The only change I made to the PBE bundled timings was to replace the UberMix v3.1 Samsung timing, which is too aggressive with this one.

Code:

777000000000000022CC1C00106A5D4DD0571016B90D060C0060070014051420FA8900A0030000001011333DC0303A17

https://github.com/jaschaknack/PolarisBiosEditor

[RooiWillie]

2. I have a few rigs on v11.5 and they are working for days until I restart miner to perform some new tests, but I always use non-overclocked cards and default bioses, I have to do it to exclude everything except miner in case of any problems. When you use miner on overclocked/unstable rig you cannot be sure that the reason of the problems is the miner, but of course you will blame it Try to reduce "-ethi" value, if it does not help you can use the version that works fine for you.

I can confirm that the miner is behaving well with my cards with the following OCs:

NVidia GTX 1050ti: +600 memory, -100 gpu clock and 75% power (samsung memory)
AMD RX470 4GB: 1500 mem straps, 96 TDP, 1125 gpu clock, 2000MHz memory
AMD RX480 4GB: 1500 mem straps, 104 TDP, 1125 gpu clock, 2000MHz memory

[Marvell2]

Is there a benefit to using the new ethermine ssl connection?
i.e lower fee? better shares?

[alex_85]

i use -ethi 16 , any disadvantage?

[RooiWillie]

i use -ethi 16 , any disadvantage?

You did not mention GPU, but I am using -ethi 10 on Red Devil RX470 4GB and -ethi 14 on MSI Armor RX480 4GB

[asb1321]

Hi all,

Does anyone knows how to reduce stale shares? Currently I have 10-15% stale shares mining ETH but no invalid shares.

Any help would be appreciated.

Thanks

[alex_85]

You did not mention GPU, but I am using -ethi 10 on Red Devil RX470 4GB and -ethi 14 on MSI Armor RX480 4GB

I am with Sapphire RX580 Nitro+ 8GB, i use -ethi 16 but actually i dont know why.....i just dont see any difference wih or without -ethi parameter.
I tried ethi 4,8,12,16 - no difference

[SKBitCoinX]

I haven't got a crash or miner restart yet from the DAG on DEV FEE running the below config
I tried a couple different options on ETH mining mode only to start, just for troubleshooting.

weird thing is I haven't seen the DEV FEE, dev mining start show up on screen yet.. not sure if thats in the backgroudn from the options im rumming in config now but anyone have some feedback let me know.

you will see the lines with # that i have turned OFF for now as i was trying different things.... -allpools 1 -allcoins -1 -allcoins 1  would always cause miner restart after DAG on DEV FEE...

-epool stratum+tcp://lb.geo.pirlpool.eu:8002
-ewal MYWalletAddress
-epsw x
-eworker RX-Miner
#-allpools 1
-allcoins etc
#-allcoins exp
#-allcoins -1
#-dpool stratum+tcp://yiimp.ccminer.org:4252
#-dwal
#-dpsw x
-esm 0
#-mode 0
#-gser 2
#-ethi 10
-etha 0
-asm 1
-mode 1
-fanmin 50
-tstop 85
-mport 0
-r 0

Has anyone experienced hashrate drop in what seems to be a quite consistent matter?
I'm not talking about my rock solid RX570 rigs, they've been absolute beasts for a year now. But I'm struggling with the Vega rig.

All GPUs will hash at ~41MH/s until one decides to drop to 26 for a minute. Then goes back up to 41. This happens regularly.

Surely if it were temp throttling (which seems unlikely as I have the fans running at pretty much 100%) the hashrate reduction would be more gradual... and not bam!... down to 26?

Any clues? I'm on v11.4, latest Adrenalin, Vega56@64, safe softPP table with specific clocks and voltages set in Claymore miner.

Got the same issue but with my rx580 rig, all 7 cards running stable at 31 mh/s. From time to time gpu0 suddenly crashes to a rate around 5-10 mh/s and after a few minutes hits back to 31. There are no error logs showing up, also no incorrect shares and temps are fine. Card has also the same setup as other cards (no oc), kinda strange since it just appeared lately, running currently v 11.5

same issue here
are you using stock rom  with no oc?

In addition to the above, Mr Claymore, your miner reboots my rig every hour. Almost without fail. There's no error in the log, no watchdog issue with GPUs OpenCL issues, nothing. I'm baffled.

1. Unstable speed - I see this issue sometimes on different cards, but I have no idea about the reason - miner executes jobs on GPUs in the same way all the time, but sometimes GPU becomes slower for some time. I noticed that often different "-ethi" value helps, try "-ethi 4" or "-ethi 12" and see if it helps. I think that it must be something related to AMD drivers but I cannot confirm it yet.
2. I have a few rigs on v11.5 and they are working for days until I restart miner to perform some new tests, but I always use non-overclocked cards and default bioses, I have to do it to exclude everything except miner in case of any problems. When you use miner on overclocked/unstable rig you cannot be sure that the reason of the problems is the miner, but of course you will blame it Try to reduce "-ethi" value, if it does not help you can use the version that works fine for you.

[RooiWillie]

I am with Sapphire RX580 Nitro+ 8GB, i use -ethi 16 but actually i dont know why.....i just dont see any difference wih or without -ethi parameter.
I tried ethi 4,8,12,16 - no difference

I have to be honest, the only reson I added in the -ethi argument was to avoid the miner having to detect what value to use, especially with the rig restarting once in a while because of the 0 fan speed picked up by Polly.

For me, faster startup means faster $$$.