Claymore's ZCash/BTG AMD GPU Miner v12.6 (Windows/Linux)

[Raja_MBZ]

Woah  Between v11.1 and v12.0 i'm getting a huge hash drop. With 11.1 i'm getting ~205H on my R9 290. When I just fired up the v12.0 i'm only getting ~69-~75H. Am I missing something? I just checked 11.1 again and it's fine so it is definitely something in 12.0 :S..hmm.

Well maybe you're missing some overclocking stuff here? Have you checked the start file and the driver configuration?

I guess something is wrong with your configuration of inner features of miner, such as undervolting using miner directly. Or maybe you're over-clocking too much which is hard to handle for the newer version because of its optimization.

[1714198391]

1714198391

[1714198391]

1714198391

[1714198391]

1714198391

[GH]

Woah   Between v11.1 and v12.0 i'm getting a huge hash drop. With 11.1 i'm getting ~205H on my R9 290. When I just fired up the v12.0 i'm only getting ~69-~75H. Am I missing something? I just checked 11.1 again and it's fine so it is definitely something in 12.0 :S..hmm.

I think lowering the intensity (-i) will cure your problem. Aggressive settings from v11.x do not work with every card type under v12.

[mistercoin]

Woah  Between v11.1 and v12.0 i'm getting a huge hash drop. With 11.1 i'm getting ~205H on my R9 290. When I just fired up the v12.0 i'm only getting ~69-~75H. Am I missing something? I just checked 11.1 again and it's fine so it is definitely something in 12.0 :S..hmm.

Well maybe you're missing some overclocking stuff here? Have you checked the start file and the driver configuration?

I guess something is wrong with your configuration of inner features of miner, such as undervolting using miner directly. Or maybe you're over-clocking too much which is hard to handle for the newer version because of its optimization.

Nope, not overclocking or undervolting at all. Everything is stock. Lowering the intensity does help but it is still much slower than 11.x without any extra settings.

[RobertD]

Woah  Between v11.1 and v12.0 i'm getting a huge hash drop. With 11.1 i'm getting ~205H on my R9 290. When I just fired up the v12.0 i'm only getting ~69-~75H. Am I missing something? I just checked 11.1 again and it's fine so it is definitely something in 12.0 :S..hmm.

Well maybe you're missing some overclocking stuff here? Have you checked the start file and the driver configuration?

I guess something is wrong with your configuration of inner features of miner, such as undervolting using miner directly. Or maybe you're over-clocking too much which is hard to handle for the newer version because of its optimization.

Nope, not overclocking or undervolting at all. Everything is stock. Lowering the intensity does help but it is still much slower than 11.x without any extra settings.

Your 11.1 config is not ok. 200h/s is really low for a 290, it should be 260-300h/s with stock bios, 350+h/s with 390x bios. 290x @ 1075mhz do 400h/s.

Have you tried using DDU to remove Radeon drivers and installing 15.12? Most problems are fixed by removing and installing drivers.  Have you checked the temperatures? Have you check the voltages?

I have lots of 290/290x working 24/7 with v11.1 and now v12.0. v12.0 does not like undervolting the GPU and some cards needed to be slowed down from 1100MHz to 1075MHz.

[GH]

Woah  Between v11.1 and v12.0 i'm getting a huge hash drop. With 11.1 i'm getting ~205H on my R9 290. When I just fired up the v12.0 i'm only getting ~69-~75H. Am I missing something? I just checked 11.1 again and it's fine so it is definitely something in 12.0 :S..hmm.

Well maybe you're missing some overclocking stuff here? Have you checked the start file and the driver configuration?

I guess something is wrong with your configuration of inner features of miner, such as undervolting using miner directly. Or maybe you're over-clocking too much which is hard to handle for the newer version because of its optimization.

Nope, not overclocking or undervolting at all. Everything is stock. Lowering the intensity does help but it is still much slower than 11.x without any extra settings.

Your 11.1 config is not ok. 200h/s is really low for a 290, it should be 260-300h/s with stock bios, 350+h/s with 390x bios. 290x @ 1075mhz do 400h/s.

Have you tried using DDU to remove Radeon drivers and installing 15.12? Most problems are fixed by removing and installing drivers.  Have you checked the temperatures? Have you check the voltages?

I have lots of 290/290x working 24/7 with v11.1 and now v12.0. v12.0 does not like undervolting the GPU and some cards needed to be slowed down from 1100MHz to 1075MHz.

Yes, of course, thats right! Stock BIOS reference 290 @ 1000 MHz can at least do 320 H/s with v12. Don't remember the values for v11.1 but the difference for sure was not 60%.

[forzendiablo]

v 12 needs alot of CPU, normally i could run -8 now just -5

more sols still but i think theres CPU leak somewhere

[grendel25]

Since V12 it seems to run for 3+ hours and then the computer starts rebooting in 5-15 cycles... and then it ran for 45 minutes and then for over an hour again... weirdness 

[lordzc]

The V12 speed is very nice. But the stability is minimal.

[Masked_Immortal]

I switched back to ver 11.1 because of stability, i tried lowering intensity, clock and increase voltage but no success.
the interesting point was the crashed gpus (390) did not work normal even after windows restart, they were hashing around 40 sols where other gpus did 380.
so i was forced to uninstall and reinstall gpu in device manager.

[spyshagg]

I have been running v12 + recommended drivers since the middle of last week non-stop on two rigs, not a single crash yet.

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ZecMiner64.exe -zpool ssl://zec.suprnova.cc:2242 -zwal xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx -zpsw xxxxxxxxxxx -allpools 1 -tstop -95 -i 8

[golden0707]

I have been running v12 + recommended drivers since the middle of last week non-stop on two rigs, not a single crash yet.

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ZecMiner64.exe -zpool ssl://zec.suprnova.cc:2242 -zwal xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx -zpsw xxxxxxxxxxx -allpools 1 -tstop -95 -i 8

Hi. What is your rig config ?

[theoulis]

Latest version v12.0:

- improved speed, about 405H/s on stock 390X, 450H/s on stock Nano, 290H/s on stock 280X, 295H/s on stock RX480.

can you please point out what drivers and what OS are you using to measure these?
in my win7 64 and latest AMD drivers i don't see those numbers unless i OC my cards!

[RobertD]

/****************/
/* Feature Request */
/****************/

People are having stability problems caused by unstable power supplies. Claymore could you miner monitor the GPU supply voltage? I would be nice to warn people when it drops bellow 11.3V and display an errror when it drops bellow 11.0V.

My 290x with server power supplies @ 12.24V show 11.64V in GPUZ. It seems GPUZ is measuring the voltage after some protection diode ( 0.7V voltage drop ). Since ATX spec allow 12V rail down to 11.5V, so 11.5V-0.7V = 10.8V.

The GPU supply voltage could also be used to tweak the mining intensity and GPU clock.

[Claymore]

Latest version v12.0:

- improved speed, about 405H/s on stock 390X, 450H/s on stock Nano, 290H/s on stock 280X, 295H/s on stock RX480.

can you please point out what drivers and what OS are you using to measure these?
in my win7 64 and latest AMD drivers i don't see those numbers unless i OC my cards!

It also depends on your GPU manufacturer/memory/model.
My hardware/software:

390X MSI GAMING 8G. "-i 8" option.
Fury Nano PowerColor (with "-powlim 50" option). "-i 7 option".
280X (don't remember details).
Windows 7, drivers 15.12.

RX480 Sapphire Nitro+ 8G. "-i 8 option".
Windows 7, drivers 16.12.2.

Note that in some cases you may need to reduce "-i" value if your rig has a lot of GPUs and/or not enough RAM/swap size.

[RobertD]

I switched back to ver 11.1 because of stability, i tried lowering intensity, clock and increase voltage but no success.
the interesting point was the crashed gpus (390) did not work normal even after windows restart, they were hashing around 40 sols where other gpus did 380.
so i was forced to uninstall and reinstall gpu in device manager.

You didn't try reducing the voltage and clock.

Try reducing your power tune to -10%, -20%, -50%. Start GPUZ and monitor the min/max GPU supply voltage. If this helps then the problem is your power supply.

[mez]

Again:

Do not use -li and -ttli options for now, there is a bug there:

More info:
https://bitcointalk.org/index.php?topic=1670733.msg17848866#msg17848866

I will release an update in 1-2 days.
Added to OP.

we used to be able to trust your time table, but right now its shite.
Your losing a lot of respect over this.

[adaseb]

Again:

Do not use -li and -ttli options for now, there is a bug there:

More info:
https://bitcointalk.org/index.php?topic=1670733.msg17848866#msg17848866

I will release an update in 1-2 days.
Added to OP.

we used to be able to trust your time table, but right now its shite.
Your losing a lot of respect over this.

No he is not.

Nobody cares if he releases it a day or two late. Its a very small bug, who cares.

Seems its not easy being a dev these days... people complaining about the dumbest things.

[tutorialevideo]

if you don`t like the V12 you can always use the older version

[bbkoopsta]

well i had to go back to v11.1 because im having a huge hash drop in v12 i hope claymore will fix it soon

[Claymore]

v12.1:

- slightly improved speed for some cards, about 410H/s on stock 390X, 450H/s on stock Nano, 290H/s on stock 280X, 300H/s on stock RX480.
- slightly improved stability. If it is still unstable for you, reduce "-i" value.
- fixed issue with wrong hashrate when "-li" or "-ttli" options enabled.