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

[mettalmag]

Code:

GPU returned incorrect data! 

how can I deal with it ?

Reduce clocks, loosen timings, or reduce undervolt.  Works on 21 470/480 cards for me, (and a 7950), all flawless.  8 are bios mod, but not extreme. Each new version is faster, and also puts more load on GPU, so it causes issues for some with more extreme tweaks.

all xfx cards are moded for ETH with low voltage mod and none of them had this problem, one card in 120, how can it be ?

Not every program is the same. Just because your video card can run League of Legends at 1350MHz, doesn't mean it won't crash if you try to run FurMark using the same settings. For everyone here having problems with crashes, blue screens, incorrect data, etc., TRY REDUCING YOUR OVERCLOCKS. It's not Claymore's job to 'fix his miner' because you refuse to find lower settings that work and tweak upwards from there. Also, every card is different. Just because one guy can copy 1375MHz memory straps and run at 2100MHz mem clock, doesn't mean that your card will be able to do the same. How about you try running 1375MHz straps at 1375MHz clock and see if it works first? If it does, that means THE PROBLEM IS YOUR OC.

Great, thank you for your advice, I'll do it on site tomorrow.

PS only game I play is EVE

Cheers

I'm having problems with the 3.0 version and my 370 GPUs. It says: OpenCL error 0 #101 - - 61. With all my 5 GPUs

What could this be?

for me updating driver version helped

[hcarpach]

I'm having problems with the 3.0 version and my 370 GPUs. It says: OpenCL error 0 #101 - - 61. With all my 5 GPUs

What could this be?

I'm seeing my problem is this one:

CL_INVALID_BUFFER_SIZE   -61   the value of the parameter size is 0 or exceeds CL_DEVICE_MAX_MEM_ALLOC_SIZE for all devices specified in the parameter context, returned by clCreateBuffer

What should I do?

[toptek]

where Linux version?

and the real turth is a drivers issue with AMD drivers and for NV i read some place  it won't be added either that's a complete rewrite of some code i read some place that's a pain to do, i wish they would tell you all the real truth and why windows works is the drivers support every card to date but a few that are to old and have no use, that's the truth why and if AMD does release a driver for all the cards like there windows drivers for Linux you will see one . and ubuntu has limits if using below 16 .

[Banksuckmyballs]

Claymore is MEGA-monster!  
RX480 4G with mod. mem timings = 70-72 sol\sec

wich timings are you using?

can you share the rom?

thanks

ROM (from MSI RX480 4G, only for HYNIX memory):
http://hushome.com/img/patched2(1950stable).rom (click rmouse and select "save link as...")
Overclocked memory from 1750 to 1950@1000mV (or 2000-2050@1250mV - this may damage your card)
Result:
http://hushome.com/img/clay3.jpg

Thanks for sharing this. How stable is it ? Should work on Sapphire 470X OC 4GB (Hynix) right ?

[nullcall]

Any chances to get v3.0 running on Linux ?

Claymore's ZCash AMD GPU Miner.
=========================

Latest version v3.0:

- improved speed by 20-30% (depends on card model). Up to 80H/s on stock 390X on "-i 2" mode.

PS. For those haters who think that I'm a thief and steal someone's code: don't use my miners, my efforts are not for you!



LINKS:

GOOGLE: https://drive.google.com/drive/folders/0B69wv2iqszefdmJickl5MF9BOEE?usp=sharing
MEGA: https://mega.nz/#F!P0ZjFDjC!Cmb-ZRqlbAnNrajRtp4zvA

NOTE: There will not be updates for Linux version for now. Please use different miner in Linux, for example, mrb's miner.

THERE IS ALSO RUSSIAN TRANSLATION OF THIS README:
https://bitcointalk.org/index.php?topic=1673724

This version is for pools.

Catalyst (Crimson) 15.12 is required for best performance and compatibility. You can get very bad results for different drivers version, or miner can fail on startup.
Set the following environment variables, especially if you have 2GB cards:

GPU_FORCE_64BIT_PTR 0
GPU_MAX_HEAP_SIZE 100
GPU_USE_SYNC_OBJECTS 1
GPU_MAX_ALLOC_PERCENT 100
GPU_SINGLE_ALLOC_PERCENT 100

For multi-GPU systems, set Virtual Memory size in Windows at least 16 GB:
"Computer Properties / Advanced System Settings / Performance / Advanced / Virtual Memory".

This miner is free-to-use, however, current developer fee is 2.5%, every hour the miner mines for 90 seconds for developer.
If you don't agree with the dev fee - don't use this miner.

This version is for recent AMD videocards only: 7xxx, 2xx, 3xx and 4xx, 2GB or more.

There are builds for Windows x64 and for Linux x64 (tested on Ubuntu 12.04). No 32-bit support. No NVidia support.



COMMAND LINE OPTIONS:

-zpool    ZCash pool address. Only Stratum protocol is supported.

-zwal    Your ZCash wallet address. Also worker name and other options if pool supports it.
   Pools that require "Login.Worker" instead of wallet address are not supported directly currently, but you can use "-allpools 1" option to mine there.

-zpsw    Password for ZCash pool, use "x" as password.

-i   mining intensity. Possible values: 0, 1, 2. 0 - lowest intensity and CPU usage, 2 - maximal intensity. You can also specify values for every card, for example "-i 0,2,1".

-allpools Specify "-allpools 1" if miner does not want to mine on specified pool (because it cannot mine devfee on that pool), but you agree to use some default pools for devfee mining.
   Note that if devfee mining pools will stop, entire mining will be stopped too.

-di    GPU indexes, default is all available GPUs. For example, if you have four GPUs "-di 02" will enable only first and third GPUs (#0 and #2).
   Use "-di detect" value to detect correct GPU order for temperatures management (requires non-zero "-tt" option); note that it will not work properly if you do not want to assign all GPUs to miner.
   You can also turn on/off cards in runtime with "0"..."9" keys and check current statistics with "s" key.

-ftime   failover main pool switch time, in minutes, see "Failover" section below. Default value is 30 minutes, set zero if there is no main pool.

-wd    watchdog option. Default value is "-wd 1", it enables watchdog, miner will be closed (or restarted, see "-r" option) if any thread is not responding for 1 minute or OpenCL call failed.
   Specify "-wd 0" to disable watchdog.

-r   Restart miner mode. "-r 0" (default) - restart miner if something wrong with GPU. "-r -1" - disable automatic restarting. -r >20 - restart miner if something
   wrong with GPU or by timer. For example, "-r 60" - restart miner every hour or when some GPU failed.
   "-r 1" closes miner and execute "reboot.bat" file ("reboot.bash" or "reboot.sh" for Linux version) in the miner directory (if exists) if some GPU failed.
   So you can create "reboot.bat" file and perform some actions, for example, reboot system if you put this line there: "shutdown /r /t 5 /f".

-dbg   debug log and messages. "-dbg 0" - (default) create log file but don't show debug messages.
   "-dbg 1" - create log file and show debug messages. "-dbg -1" - no log file and no debug messages.

-logfile debug log file name. After restart, miner will append new log data to the same file. If you want to clear old log data, file name must contain "noappend" string.
   If missed, default file name will be used.

-li   low intensity mode. Reduces mining intensity, useful if your cards are overheated. Note that mining speed is reduced too.
   More value means less heat and mining speed, for example, "-li 10" is less heat and mining speed than "-li 1". You can also specify values for every card, for example "-li 3,10,50".
   Default value is "0" - no low intensity mode.

-tt   set target GPU temperature. For example, "-tt 80" means 80C temperature. You can also specify values for every card, for example "-tt 70,80,75".
   You can also set static fan speed if you specify negative values, for example "-tt -50" sets 50% fan speed. Specify zero to disable control and hide GPU statistics.
   "-tt 1" (default) does not manage fans but shows GPU temperature and fan status every 30 seconds. Specify values 2..5 if it is too often.
   Note: for Linux gpu-pro drivers, miner must have root access to manage fans, otherwise only monitoring will be available.

-ttli   reduce entire mining intensity (for all coins) automatically if GPU temperature is above specified value. For example, "-ttli 80" reduces mining intensity if GPU temperature is above 80C.
   You can see if intensity was reduced in detailed statistics ("s" key).
   You can also specify values for every card, for example "-ttli 80,85,80". You also should specify non-zero value for "-tt" option to enable this option.
   It is a good idea to set "-ttli" value higher than "-tt" value by 3-5C.
   NOTE: Check "KNOWN ISSUES" section. GPU indexes in temperature control sometimes don't match GPU indexes in mining!

-tstop   set stop GPU temperature, miner will stop mining if GPU reaches specified temperature. For example, "-tstop 95" means 95C temperature. You can also specify values for every card, for example "-tstop 95,85,90".
   This feature is disabled by default ("-tstop 0"). You also should specify non-zero value for "-tt" option to enable this option.
   NOTE: Check "KNOWN ISSUES" section. GPU indexes in temperature control sometimes don't match GPU indexes in mining!
   If it turned off wrong card, it will close miner in 30 seconds.
   You can also specify negative value to close miner immediately instead of stopping GPU, for example, "-tstop -95" will close miner as soon as any GPU reach 95C temperature.

-fanmax   set maximal fan speed, in percents, for example, "-fanmax 80" will set maximal fans speed to 80%. You can also specify values for every card, for example "-fanmax 50,60,70".
   This option works only if miner manages cooling, i.e. when "-tt" option is used to specify target temperature. Default value is "100".

-fanmin   set minimal fan speed, in percents, for example, "-fanmin 50" will set minimal fans speed to 50%. You can also specify values for every card, for example "-fanmin 50,60,70".
   This option works only if miner manages cooling, i.e. when "-tt" option is used to specify target temperature. Default value is "0".

-cclock   set target GPU core clock speed, in MHz. If not specified or zero, miner will not change current clock speed. You can also specify values for every card, for example "-cclock 1000,1050,1100,0".
   Unfortunately, AMD blocked underclocking for some reason, you can overclock only.

-mclock   set target GPU memory clock speed, in MHz. If not specified or zero, miner will not change current clock speed. You can also specify values for every card, for example "-mclock 1200,1250,1200,0".
   Unfortunately, AMD blocked underclocking for some reason, you can overclock only.

-powlim set power limit, from -50 to 50. If not specified, miner will not change power limit. You can also specify values for every card, for example "-powlim 20,-20,0,10".

-cvddc   set target GPU core voltage, multiplied by 1000. For example, "-cvddc 1050" means 1.05V. You can also specify values for every card, for example "-cvddc 900,950,1000,970". Supports latest AMD 4xx cards only in Windows.

-mvddc   set target GPU memory voltage, multiplied by 1000. For example, "-mvddc 1050" means 1.05V. You can also specify values for every card, for example "-mvddc 900,950,1000,970". Supports latest AMD 4xx cards only in Windows.

-mport   remote monitoring/management port. Default port is 3333, specify "-mport 0" to disable remote monitoring/management feature.
   Specify negative value to enable monitoring (get statistics) but disable management (restart, uploading files), for example, "-mport -3333" enables port 3333 for remote monitoring, but remote management will be blocked.
   You can also use your web browser to see current miner state, for example, type "localhost:3333" in web browser.

-colors enables or disables colored text in console. Default value is "1", use "-colors 0" to disable coloring.



CONFIGURATION FILE

You can use "config.txt" file instead of specifying options in command line.
If there are not any command line options, miner will check "config.txt" file for options.
If there is only one option in the command line, it must be configuration file name.
If there are two or more options in the command line, miner will take all options from the command line, not from configuration file.
Place one option per line, if first character of a line is ";" or "#", this line will be ignored.



SAMPLE USAGE

 flypool:
   ZecMiner64.exe -zpool eu1-zcash.flypool.org:3333 -zwal t1RjQjDbPQ9Syp97DHFyzvgZhcjgLTMwhaq.YourWorkerName -zpsw x

 nanopool:
   ZecMiner64.exe -zpool zec-eu1.nanopool.org:6666 -zwal t1RjQjDbPQ9Syp97DHFyzvgZhcjgLTMwhaq.YourWorkerName -zpsw x

 miningpoolhub
   ZecMiner64.exe -zpool us-east.equihash-hub.miningpoolhub.com:20570 -zwal YourLogin.YourWorkerName -zpsw YourWorkerPassword

 nicehash ZEC pool:
   ZecMiner64.exe -zpool stratum.zcash.nicehash.com:3357 -zwal t1RjQjDbPQ9Syp97DHFyzvgZhcjgLTMwhaq -zpsw x

 nicehash Equihash pool:
   ZecMiner64.exe -zpool equihash.usa.nicehash.com:3357 -zwal 1LmMNkiEvjapn5PRY8A9wypcWJveRrRGWr -zpsw x

 suprnova:
   ZecMiner64.exe -zpool zec.suprnova.cc:2142 -zwal YourLogin.YourWorkerName -zpsw YourWorkerPassword

 coinmine:
   ZecMiner64.exe -zpool zec.coinmine.pl:7007 -zwal YourLogin.YourWorkerName -zpsw YourWorkerPassword

 coinsforall:
   ZecMiner64.exe -zpool coinsforall.io:3357 -zwal t1RjQjDbPQ9Syp97DHFyzvgZhcjgLTMwhaq.YourWorkerName -zpsw x



FAILOVER

Use "epools.txt" file to specify additional pools. This file has text format, one pool per line. Every pool has 3 connection attempts.
Miner disconnects automatically if pool does not send new jobs for a long time or if pool rejects too many shares.
If the first character of a line is ";" or "#", this line will be ignored.
Do not change spacing, spaces between parameters and values are required for parsing.
If you need to specify "," character in parameter value, use two commas - ,, will be treated as one comma.
You can reload "epools.txt" file in runtime by pressing "r" key.
Pool specified in the command line is "main" pool, miner will try to return to it every 30 minutes if it has to use some different pool from the list.
If no pool was specified in the command line then first pool in the failover pools list is main pool.
You can change 30 minutes time period to some different value with "-ftime" option, or use "-ftime 0" to disable switching to main pool.



REMOTE MONITORING/MANAGEMENT

Miner supports remote monitoring/management via JSON protocol over TCP/IP sockets, HTTP is supported as well.
Start "EthMan.exe" from "Remote management" subfolder (Windows version only).
Check "Help" tab for built-in help.



KNOWN ISSUES

- AMD cards: GPU indexes in temperature control sometimes don't match GPU indexes in mining. Miner has to enumerate GPUs via OpenCL API to execute OpenCL code, and also it has to enumerate GPUs via ADL API to manage temperature/clock.
And order of GPUs in these lists can be different. There is no way to fix GPUs order automatically, thanks to AMD devs.
But you can do it manually. For example, if you have two cards, you can change their order by adding "-di 10". Another example, reverse order for six cards: "-di 543210".
Also you can do it automatically (experimental feature) with "-di detect" option.
- Windows 10 Defender recognizes miner as a virus, some antiviruses do the same. Miner is not a virus, add it to Defender exceptions.
  I write miners since 2014. Most of them are recognized as viruses by some paranoid antiviruses, perhaps because I pack my miners to protect them from disassembling, perhaps because some people include them into their botnets, or perhaps these antiviruses are not good, I don't know. For these years, a lot of people used my miners and nobody confirmed that my miner stole anything or did something bad.
  Note that I can guarantee clean binaries only for official links in my posts on this forum (bitcointalk). If you downloaded miner from some other link - it really can be a virus.
  However, my miners are closed-source so I cannot prove that they are not viruses. If you think that I write viruses instead of good miners - do not use this miner, or at least use it on systems without any valuable data.



TROUBLESHOOTING

1. Install Catalyst v15.12 (for AMD cards).
2. Disable overclocking.
3. Set environment variables as described above.
4. Set Virtual Memory 16 GB.
5. Reboot computer.
6. Check hardware, risers.
7. Set some timeout in .bat file before starting miner at system startup (30sec or even a minute), and try "-ethi 4" to check if it is more stable. It can help if miner is not stable on some system.


FAQ:

Q: It seems miner mines devfee more than 90 seconds.
A: No it does not, please read: https://bitcointalk.org/index.php?topic=1670733.msg16777816;topicseen#msg16777816

Q: Why do I see more shares for devfee than in my mining for the same time?
A: Most pools support variable diff, they change "target share" after some time after connection. For example, you have very powerful rig, after connection you will send shares very often. It takes some CPU time to check your shares so after some time pool will send higher share target and miner will send less shares (but they will have more value). When pool updates share target you will see "Pool sets new share target" line in the miner. This way pool can adjust the number of shares that miner sends and balance its load.
So check the log or console text to see current target for main mining thread and for devfee thread. For example:
DevFee: Pool sets new share target: 0x0083126e (diff: 500H) - this is for devfee mining connection
Pool sets new share target: 0x0024fa4f (diff: 1772H) - this is for main mining connection
As you can see, target share for main mining is higher in about 3.5 times, so for main mining miner sends in 3 times less shares (but they have 3x more value) than for devfee mining.


==========
This miner uses some code of SILENTARMY OpenCL GPU Zcash Equihash solver:

# License

The MIT License (MIT)
Copyright (c) 2016 Marc Bevand

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"),
to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

[AcaStar]

If you guys look at GPGPU Benchmarks
http://www.anandtech.com/bench/GPU16/1512

You will notice that the 480 8GB will at max be 20% faster than the 290 under full optimization.

Comparing the results


480 8GB - 120MH/s
290 - 107MH/s
280x - 83.3MH/s
7850 - 45.5MH/s

Using the same ratios that leads to Claymore software

Nano - 98 Sol/s <- should be around here Optimally
Fury X - 96Sol/s <- should be around here Optimally
480 8GB - 96Sol/s <- should be around here Optimally.
290 - 75 Sol/s
280x - 55Sol/s
7850 - 30Sol/s

We are FVCKED if some developer released a Optimized Nvidia ZEC miner however because if its released and uses a similar hashing algo as Claymore then...

GTX 1080 - 262MH/s
GTX 1070 - 224.4MH/s
980Ti - 213.6MH/s

Converting to what we have now


GTX 1080 - 183Sol/s
GTX 1070 - 157Sol/s
980Ti - 150Sol/s


Thank god all the Nvidia miner devs are greedy and keeping everything PRIVATE.

On my rig with windows 7 ultimate and 16.10.2 drivers with  6 x MSI r9 380  -80/800/1500 setup on afterburner gettting about 35-37 h/s in total about 220 with new claymore v3 zcash miner mode i-2. Is that Ok or i can do some other setup and get more from this cards?

[xeridea]

If you guys look at GPGPU Benchmarks
http://www.anandtech.com/bench/GPU16/1512

You will notice that the 480 8GB will at max be 20% faster than the 290 under full optimization.

Comparing the results


480 8GB - 120MH/s
290 - 107MH/s
280x - 83.3MH/s
7850 - 45.5MH/s

Using the same ratios that leads to Claymore software

Nano - 98 Sol/s <- should be around here Optimally
Fury X - 96Sol/s <- should be around here Optimally
480 8GB - 96Sol/s <- should be around here Optimally.
290 - 75 Sol/s
280x - 55Sol/s
7850 - 30Sol/s

We are FVCKED if some developer released a Optimized Nvidia ZEC miner however because if its released and uses a similar hashing algo as Claymore then...

GTX 1080 - 262MH/s
GTX 1070 - 224.4MH/s
980Ti - 213.6MH/s

Converting to what we have now


GTX 1080 - 183Sol/s
GTX 1070 - 157Sol/s
980Ti - 150Sol/s


Thank god all the Nvidia miner devs are greedy and keeping everything PRIVATE.

Except you don't understand how these things work.  Cards can vary wildy from task to task (most GPGPU loads are compute limited). 1080 is great at computing, but sux horribly for Ethereum and Zcash, since they are memory hard, and it has very loose timings n GDDR5X memory.  1070 is no better than a 480 with bios mod for Ethereum, yet costs twice as much. I think it is partially due to how global memory is addressed.  All nvidia miners so far for Zcash are a lot slower than AMD miners, and the cards cost a lot more.  Theoretically if 1070 could be 50% faster, at 100% more cost, 1070 still loses.

[bradli]

Gigabyte 390 8gb G1 they start to mine with Claymor 3.0 and crash few share after start. I have tryed -i 0 and 1 and 2 same results any solutions or sugestions ? All other miners work no problem  

change driver to 15.12

[Linit]

same rigs 6x390
rig 1 | win 8.1 - 68H/s per card
rig 2 | win 10 - crash after first share --- problem with all win 10 rigs

Install 15.12 drivers.

for those on win 10.. copy the zecminer64 exe from ver 2.1 to the 3.0 folder overwriting the file.  Then try it.. Seems the gains are had in the files associated with zecminer :0  plus no crash.

seems win 10 has some wierd issue with giving this version access.  Not sure why it didnt do it on first load up but second or third time it gave me a error that it didnt have rights to the cards..

best Regards
d57heinz

You know you are the Man ? :-) Worked now with much more Speed on my Main System without Crashes. My two Rigs works with original v3.0 too.

Thanks,

Gary

ALSO I WORK. 

[mishax1]

Which one is better for mining today? (without problems or the need to upgrade bios and stuff)

MSI - RX-480-GAMING-8G
Sapphire - 11260-07-20G - SAPPHIRE NITRO Plus
Gigabyte - GV-RX480G1-GAMING-8GD

Thanks..    If anyone got one of these and mind sharing his H/s I would really appreciate it! 

[tutorialevideo]

Testing on moded 470 4 GB getting arround 52-58/gpu. with -i 1

[hcarpach]

New problem: "Application has been blocked from accessing Graphics hardware". What's the deal here?

[safonov]

270X 4GB - 38-39

[GH]

So basically it only works best on the newer cards RX470/RX480.

The rest are crashing

You must be doing something wrong. Running stable AND fast on all my OCd and partially BIOS modded R9 270, 270x, 290, 290x, 370, 380, 380x, 390, (+ RX470, 480) here...

[Tmdz]

So basically it only works best on the newer cards RX470/RX480.

The rest are crashing

You must be doing something wrong. Running stable AND fast on all my OCd and partially BIOS modded R9 270, 270x, 290, 290x, 370, 380, 380x, 390, (+ RX470, 480) here...

Try it with -i 0 maybe the cpu is too slow causing problems, -i 2 doesn't give that much extra anyway.

Then update to the newest drivers and see how it working.

No problems for a few hours on 290 rig and 470 rig.  My 290 rig can get unstable easily and hard crash.

edit: Under volting can also help on stability.

[joaocha]

Claymore, it is possible to scale better the intensity -i ?

Like 0 to 10?

Some of cpus, runs fine with 1 (60%) but bottlenec at 2 (100%)

[Swimmer63]

Claymore - love the idea of another 20% increase.  But having trouble with 290's and 390's.  Either locking up or (390) or getting 0.00 H/s (290).

[semajjames]

so 15.12 drivers makes more memory available than 14.6 ?

GPU #0:Hawaii, 4096 MB available, 40 compute units
GPU #0 recognized as Radeon 290

GPU #0 Hawaii, 3072 MB available, 40 compute units
GPU #0 recognized as Radeon 290

[offon]

anyone with nanopool getting error? ZEC: Stratum - socket send failed 10054, disconnect

[Claymore]

Claymore - love the idea of another 20% increase.  But having trouble with 290's and 390's.  Either locking up or (390) or getting 0.00 H/s (290).

I have rigs with 270X, 280X, 285, 390, 390X, 480 cards, i3 CPU, all they work. However, I will test everything more carefullt within a day and release an update if I find something useful.
Note that "-i 2" may fail on 2GB cards or if you don't have enough RAM and/or virtual memory.

Claymore, it is possible to scale better the intensity -i ?
Like 0 to 10?
Some of cpus, runs fine with 1 (60%) but bottlenec at 2 (100%)

Ok I'll check it.