Claymore's CryptoNote AMD GPU Miner v11.3

[Redcard]

While using this miner I noticed the fan speed of the main card(rx 470) suddenly jumped to 100%! WTF??

I'd say its a driver bug. however I have it only in sgminer. I use mostly 16.9.2 and never seen it on claymore, but on sgminer I get it all the time on all the rigs I tested

EDIT: some weird combination of XMR algo code and drivers

In addition to this problem is I'm not able to fix the crazy fan at normal speed unless I reset the system.
And after one amd driver crash and and reboot the driver has changed from 16.10.3 to 16.6.. but this is another issue.

same here, only the reboot solves this crazy run... it shows over 5000RPM on refs.

Have you tried to set fan rpm max limit with wattool or wattman before it rise to 100% rpm?

[PlasticStool]

10:28:48:978   2e80   Stratum - Connected (unsecure)
10:28:48:978   2e80   New connection: reset total hashes counters
10:28:49:062   2e80   got 100 bytes
10:28:49:062   2e80   buf: {"id":1,"jsonrpc":"2.0","error":{"code":-1,"message":"invalid address or username used for login"}}

10:28:49:062   2e80   parse packet: 94
10:28:49:062   2e80   Error in server response
: {"id":1,"jsonrpc":"2.0","error":{"code":-1,"message":"invalidaddressorusernameusedforlogin"}}
10:28:49:077   2e80   buf was cleared
10:28:49:077   2e80   Stratum - reading socket failed 10038, disconnect
10:28:49:077   2e80    Job timeout, disconnect, retry in 20 sec...


As username I use xmr address generated in Kraken but why the miner doesn't accept it??

I got this same error. 
It was because when I C&P the address into the batch file it removed the space between the address and -p x

[Ursul0]

While using this miner I noticed the fan speed of the main card(rx 470) suddenly jumped to 100%! WTF??

I'd say its a driver bug. however I have it only in sgminer. I use mostly 16.9.2 and never seen it on claymore, but on sgminer I get it all the time on all the rigs I tested

EDIT: some weird combination of XMR algo code and drivers

In addition to this problem is I'm not able to fix the crazy fan at normal speed unless I reset the system.
And after one amd driver crash and and reboot the driver has changed from 16.10.3 to 16.6.. but this is another issue.

same here, only the reboot solves this crazy run... it shows over 5000RPM on refs.

Have you tried to set fan rpm max limit with wattool or wattman before it rise to 100% rpm?

it fucks all the limits

[Redcard]

I got this same error. 
It was because when I C&P the address into the batch file it removed the space between the address and -p x

Just check those two addr:
4GdoN7NCTi8a5gZug7PrwZNKjvHFmKeV11L6pNJPgj5QNEHsN6eeX3DaAQFwZ1ufD4LYCZKArktt113 W7QjWvQ7CWFNeDcndLq3GwnXRLM
46evMe84XVyeAD9XsQ3G93WtcVQeLpC4P4UKEPMwEhdYRwAo8KKtTrf5DQMitTK8TnWo7EMUxvXjXjW djutnMbf823AbDkR

The longer is from Kraken and the 2nd is from mymonero.com. The space could be a problem as you said but in my case I think the address is strange.  

[Ursul0]

these are the safe numbers for my average ref 480 8GB with 1500 straps:
it does ~740.

[GPU]
Mode=Manual
P1_CLK=483
P1_VID=800
P2_CLK=778
P2_VID=811
P3_CLK=906
P3_VID=834
P4_CLK=1020
P4_VID=854
P5_CLK=1060
P5_VID=857
P6_CLK=1080
P6_VID=888
P7_CLK=1160
P7_VID=917
[Memory]
Mode=Manual
P1_CLK=1930
P1_VID=917
[Fan]
Mode=Manual
MinRPM=1400
MaxRPM=4900
TargetTemp=70

[rinus]

these are the safe numbers for my average ref 480 8GB with 1500 straps:
it does ~740.

[GPU]
Mode=Manual
P1_CLK=483
P1_VID=800
P2_CLK=778
P2_VID=811
P3_CLK=906
P3_VID=834
P4_CLK=1020
P4_VID=854
P5_CLK=1060
P5_VID=857
P6_CLK=1080
P6_VID=888
P7_CLK=1160
P7_VID=917
[Memory]
Mode=Manual
P1_CLK=1930
P1_VID=917
[Fan]
Mode=Manual
MinRPM=1400
MaxRPM=4900
TargetTemp=70

all these numbers , are that settings u put in somewhere? and what do you mean by " 1500 " straps ?

" im still learning a lot about mining "

[Redcard]

it fucks all the limits

Confirmed but there must be some solution. Like waterblock instead of fan after all

[Ursul0]

it fucks all the limits

Confirmed but there must be some solution. Like waterblock instead of fan after all

sure I do have one 370 that drives my desktop which is watercooled:)


It just cannot believe there's no fan...

[Ursul0]

these are the safe numbers for my average ref 480 8GB with 1500 straps:
it does ~740.

[GPU]
Mode=Manual
P1_CLK=483
P1_VID=800
P2_CLK=778
P2_VID=811
P3_CLK=906
P3_VID=834
P4_CLK=1020
P4_VID=854
P5_CLK=1060
P5_VID=857
P6_CLK=1080
P6_VID=888
P7_CLK=1160
P7_VID=917
[Memory]
Mode=Manual
P1_CLK=1930
P1_VID=917
[Fan]
Mode=Manual
MinRPM=1400
MaxRPM=4900
TargetTemp=70

http://www.overclock.net/t/1609782/watttool-a-simple-tool-that-combines-overclocking-with-vrm-monitoring-tweaking-for-rx-400-series

all these numbers , are that settings u put in somewhere? and what do you mean by " 1500 " straps ?

" im still learning a lot about mining "

for rx400 series use watttool with profiles for each card/each algo.
straps are memory timings that you need to tighten in bios (bios mod)

[rinus]

these are the safe numbers for my average ref 480 8GB with 1500 straps:
it does ~740.

[GPU]
Mode=Manual
P1_CLK=483
P1_VID=800
P2_CLK=778
P2_VID=811
P3_CLK=906
P3_VID=834
P4_CLK=1020
P4_VID=854
P5_CLK=1060
P5_VID=857
P6_CLK=1080
P6_VID=888
P7_CLK=1160
P7_VID=917
[Memory]
Mode=Manual
P1_CLK=1930
P1_VID=917
[Fan]
Mode=Manual
MinRPM=1400
MaxRPM=4900
TargetTemp=70

all these numbers , are that settings u put in somewhere? and what do you mean by " 1500 " straps ?

" im still learning a lot about mining "

for rx400 series use watttool with profiles for each card/each algo.
straps are memory timings that you need to tighten in bios (bios mod)

are there manuals on how to reflash these bios ?

[rinus]

these are the safe numbers for my average ref 480 8GB with 1500 straps:
it does ~740.

[GPU]
Mode=Manual
P1_CLK=483
P1_VID=800
P2_CLK=778
P2_VID=811
P3_CLK=906
P3_VID=834
P4_CLK=1020
P4_VID=854
P5_CLK=1060
P5_VID=857
P6_CLK=1080
P6_VID=888
P7_CLK=1160
P7_VID=917
[Memory]
Mode=Manual
P1_CLK=1930
P1_VID=917
[Fan]
Mode=Manual
MinRPM=1400
MaxRPM=4900
TargetTemp=70

all these numbers , are that settings u put in somewhere? and what do you mean by " 1500 " straps ?

" im still learning a lot about mining "

for rx400 series use watttool with profiles for each card/each algo.
straps are memory timings that you need to tighten in bios (bios mod)

are there manuals on how to reflash these bios ?

ok i did some research and i understand how to mod the bios, you copy the value from 1500 to the higher frequencies ...

but what dos P1 clk and p1 vid , p2 clk p2 vid ... mean, are these all different gpu cards? why is the first set so low and the last so high? should you not set them the same for max hash ?

thx for the advice , appreciate it

[werdwerdus]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

[rinus]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

What power usage per card do u have at 740 hash?

[Cubirez]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

What power usage per card do u have at 740 hash?

Should be Sub 100W / card at these voltages.

[xneoenx]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

What power usage per card do u have at 740 hash?

I am at around 360W with 3 cards (RX480) and sys power. 410W Including CPU Mining (i7 4790K).

[rinus]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

What power usage per card do u have at 740 hash?

I am at around 360W with 3 cards (RX480) and sys power. 410W Including CPU Mining (i7 4790K).

Ok I will order 6 sapphire rx480 nitro 8giga cards. They are better for tweaking yes? Good cooling... With these I safely get 740 hash and sub 800 w in total yes?

[Ursul0]

those are the different power levels for the gpu. so the lowest is for 2D mode when you're no utilizing the GPU, the higher ones are for more demanding application calls.

What power usage per card do u have at 740 hash?

I am at around 360W with 3 cards (RX480) and sys power. 410W Including CPU Mining (i7 4790K).

Ok I will order 6 sapphire rx480 nitro 8giga cards. They are better for tweaking yes? Good cooling... With these I safely get 740 hash and sub 800 w in total yes?

good cards. you'll probably be able to get it below 800 on XMR, in any case I'd suggest to run it of 1.2KW PSU

[xneoenx]

Ok I will order 6 sapphire rx480 nitro 8giga cards. They are better for tweaking yes? Good cooling... With these I safely get 740 hash and sub 800 w in total yes?

Yes. They are good cards, with the best (lenient) mosfets and good cooling. They didnt go cheap like Gigabyte or Powercolor. But remember you may still need to tweak you bios / strap timings to run it efficiently.

[rinus]

Ok I will order 6 sapphire rx480 nitro 8giga cards. They are better for tweaking yes? Good cooling... With these I safely get 740 hash and sub 800 w in total yes?

Yes. They are good cards, with the best (lenient) mosfets and good cooling. They didnt go cheap like Gigabyte or Powercolor. But remember you may still need to tweak you bios / strap timings to run it efficiently.

Yeah was not sure to get the cheaper 4giga ones but I am convinced buying the 4giga nitro are the best way to go. I will reflash the BIOS with the 1500 straps and undervolt...

[Qazo]

Can someone please help?

why is this happening randomly ? before this it was rock stable for >24 hours

11:55:18:642   e74   Round
11:55:20:064   948   GPU 0: HashCnt = 1024, time = 1395 ms, hashrate = 734 h/s
11:55:20:080   948   Round
11:55:20:892   55c   GPU0 t=49C fan=46%, GPU1 t=55C fan=41%
11:55:20:892   55c   watchdog - thread 0, hb time 812
11:55:20:892   55c   watchdog - thread 1, hb time 2250
11:55:20:892   55c   watchdog - thread 2, hb time 80437
11:55:20:892   55c   WATCHDOG: Thread 2 hangs in OpenCL call, you need to restart miner
11:55:20:908   55c   watchdog - thread 3, hb time 81891
11:55:20:908   55c   WATCHDOG: Thread 3 hangs in OpenCL call, you need to restart miner
11:55:20:908   55c   WATCHDOG: GPU memory error, you need to restart miner
11:55:21:517   e74   GPU 0: HashCnt = 1024, time = 1412 ms, hashrate = 725 h/s
11:55:21:533   e74   Round
11:55:22:174   55c   Restarting OK, exit...


and when it restart minner PC just hangs and I need to manually restart PC. It's not problem when I'm here but I soon plan to move miner in other place where I will have much cheaper electricity but I will have no access to minner and this will be a huge problem for me.


2x480 8gb OC ~ 1470 h/s with 305 W from wall
1350 MHz, 1080 mV
2200 MHz, 1000 mV

reject rate about 1.3%, suprnova pool