[Cast XMR] high speed XMR/CryptoNight miner for RX Vega GPUs (2 KHash/s)

[Puff Pastry Hangman]

I think I'm making a rookie error but it won't work for me, SOCKET ERROR: NO LOGIN/PASSWORD SPECIFIED

Is there some guide for how to set up the configuration? I can't work it out from the defaults.

[1714166315]

1714166315

[grimjaw99]

I think I'm making a rookie error but it won't work for me, SOCKET ERROR: NO LOGIN/PASSWORD SPECIFIED

Is there some guide for how to set up the configuration? I can't work it out from the defaults.

Does your pool require a password? Otherwise all you need is your wallet id..

Got this up and working... and yeah having a 2nd instance will knock about 200 h/s off the 1st instance... but oh well
I'm running 1600 h/s and 1800 h/s from 2 VEGA 56... still nice though.

[grimjaw99]

Well, I got a problem... I'm getting artifacts on my screen probably due to instability..
so I set my clocks down to 900 memory and still no go.
Frankly I like your miner more compared to say Claymore's but it really needs to support more cards per instance.
XMR STAK is still king right now.. good luck with your progress.

[DrgLT]

Can you post all commands cast_xmr --help wont work. My first gpu is intel so I cant get miner working. Error GPU0: Intel(R) HD Graphics 610 | 12 Compute Units, how i turn it off? Btw 6xRX Vega 64 Limited.
 I found fix --opencl 1 -G 1 but how writ it to start mining for all 6 cards?

[glph3k]

Can you post all commands cast_xmr --help wont work. My first gpu is intel so I cant get miner working. Error GPU0: Intel(R) HD Graphics 610 | 12 Compute Units, how i turn it off? Btw 6xRX Vega 64 Limited.
 I found fix --opencl 1 -G 1 but how writ it to start mining for all 6 cards?

Currently only one instance per card possible, but you can start multiple instance, each for a different card.


Working on a multi GPU version

[Shnikes101]

Always good to see another miner. Like others have said, multiple instances appear to affect hash rates. I'm getting strong initial hashes that fall over time and settle in the 1800 range.

Will definitely keep an eye on this project. Looking forward to your multi gpu version. For now though, I'm sticking with xmr-stak-amd

[tutorialevideo]

Keep up the great work the miner work awesome!

[codencoin]

Nice job.  Thanks for the specific miner. I seem to have a problem that no one else has really mentioned. I have an MSI Vega 56, and I have no problem getting 1860h/s for about 5 minutes... then it drops by 200h/s to 1660h/s... and is stable at about that rate. I can kill the miner, and start it back up and it does the same thing. I thought maybe it was a heat thing, but after playing with settings, I don't think so. Looks like perhaps a bug. Any thoughts? Maybe it is related to the other use who said they loose 200h/s on their first card when they have a second card. This system does have an integrated graphics card.

[Ystss]

Nice job.  Thanks for the specific miner. I seem to have a problem that no one else has really mentioned. I have an MSI Vega 56, and I have no problem getting 1860h/s for about 5 minutes... then it drops by 200h/s to 1660h/s... and is stable at about that rate. I can kill the miner, and start it back up and it does the same thing. I thought maybe it was a heat thing, but after playing with settings, I don't think so. Looks like perhaps a bug. Any thoughts? Maybe it is related to the other use who said they loose 200h/s on their first card when they have a second card. This system does have an integrated graphics card.

try it "-I 9"

[bostonvex]

I have 3 vega 56s. Used the one instance pr GPU method as described above with --opencl and -G.

Every instance I start gets over 1900H/s but as soon as I start a newer instance the previous instance drops to 1600H/s. The net effect is that I have one instance with 1900 and 2 with 1600 giving me around 5100H/s total.

Tried the -I 9 method and still same issue.



That's a little less than what I get with XMR-stak with code ripped from Wolf0.

Hoping the multi GPU version will fix this. I would prefer using stable, fast, non ripped off code that honors the open source license agreement. Keep up the good work.

[Goldenbunip]

I tried this on a powercolor VEGA 64.
Getting a stable 2050Hs for the last 12 hours now, which is more profitable than the 44.2Mh this card does on ETH.  Its submitting and being accepted at Nicehash just fine.

Settings
Core -30% (1150mhz) (no difference in hash all the way upto 1800mhz)
Mem 1100
Power limit +50%

Strings used (after the usual site and wallet) -G 1 -I 7
intensity at 7-9 are all the same hash rates so left at 7. I didn't need the --nicehash addon to make this work.

The system has another card as GPU0 which is mining ETH/LCR on the NiceHash Legacy software just fine alongside this.

HWinfo says about 120W core. Which is the same setting and power usage as I use on Claymore for 44.2Mhs on ETH.
The card is water-cooled with a EK full cover and running at 45C.

Now to see how stable this is over the next month or so.

[Nokia555]

Is there a API coming also? So i can list it on EthMonitoring EthControl software control for users.

[bostonvex]

I tweaked the following:

• Frequency: -4%
• MemClock: 920 (best my silicon can do without artifacts)
• Power Limit: -10%
• args:  -I 9 --opencl 1 -G 1 (or G 2 or G 0 depending on instance)

Now getting between 1850 - 1890 H/s per card. Going to let it run today and see if it remains stable.

[glph3k]

There is an update available:

Cast XMR Version 0.5 (2017/10/19)

  - support for multiple Vega GPUs. Specify multiple GPUs comma separated with the -G switch like "-G 0,1,2,3"
  - monitor temperature and fan speed for each GPU
  - subtle performance improvements in core

Download Cast XMR 0.5 for Windows (64bit)


Now hash with full 4 KHashes/s with 2 GPUs:

[bostonvex]

There is an update available:

Cast XMR Version 0.5 (2017/10/19)

  - support for multiple Vega GPUs. Specify multiple GPUs comma separated with the -G switch like "-G 0,1,2,3"
  - monitor temperature and fan speed for each GPU
  - subtle performance improvements in core

Download Cast XMR 0.5 for Windows (64bit)


Now hash with full 4 KHashes/s with 2 GPUs:

Woohoo!

[rednoW]

There is an update available:

Cast XMR Version 0.5 (2017/10/19)

  - support for multiple Vega GPUs. Specify multiple GPUs comma separated with the -G switch like "-G 0,1,2,3"
  - monitor temperature and fan speed for each GPU
  - subtle performance improvements in core

Download Cast XMR 0.5 for Windows (64bit)


Now hash with full 4 KHashes/s with 2 GPUs:

I'd better add fan speed - with 64-72 gpu temp your mem temp will be to high, I think.

[huntingthesnark]

There is an update available:

Cast XMR Version 0.5 (2017/10/19)

  - support for multiple Vega GPUs. Specify multiple GPUs comma separated with the -G switch like "-G 0,1,2,3"
  - monitor temperature and fan speed for each GPU
  - subtle performance improvements in core

Download Cast XMR 0.5 for Windows (64bit)


Now hash with full 4 KHashes/s with 2 GPUs:

Congrats sir. Bloody good show.

[codencoin]

Nice job.  Thanks for the specific miner. I seem to have a problem that no one else has really mentioned. I have an MSI Vega 56, and I have no problem getting 1860h/s for about 5 minutes... then it drops by 200h/s to 1660h/s... and is stable at about that rate. I can kill the miner, and start it back up and it does the same thing. I thought maybe it was a heat thing, but after playing with settings, I don't think so. Looks like perhaps a bug. Any thoughts? Maybe it is related to the other use who said they loose 200h/s on their first card when they have a second card. This system does have an integrated graphics card.

try it "-I 9"

Didn't change... updated to 0.5 version also.

[07:25:14] Initializing GPU, loading kernel ...
[Suspicious link removed]pute Driver detected.
GPU0: gfx901 | 64 Compute Units
[07:25:15] Connecting to pool.supportxmr.com:7777 ...
[07:25:15] Connected to pool.
[07:25:15] New job received.
[07:25:15] Difficulty changed. Now: 25000.
[07:25:19] GPU0 | 64°C | Fan 2078 RPM | 1679.1 H/s
[07:25:23] Found Nonce, submitting...
[07:25:23] Share Accepted -> +1
[07:25:23] Shares: 1 Accepted, 0 Rejected | Total Hash Rate: 1739.5 H/s | Avg Search Time: 8.0 sec
[07:25:25] GPU0 | 65°C | Fan 2282 RPM | 1887.0 H/s
[07:25:30] GPU0 | 65°C | Fan 2479 RPM | 1875.1 H/s

.....

[07:45:27] GPU0 | 69°C | Fan 2441 RPM | 1886.6 H/s
[07:45:33] GPU0 | 69°C | Fan 2441 RPM | 1885.7 H/s
[07:45:38] GPU0 | 69°C | Fan 2441 RPM | 1887.6 H/s
[07:45:44] GPU0 | 68°C | Fan 2444 RPM | 1852.2 H/s
[07:45:49] GPU0 | 69°C | Fan 2401 RPM | 1629.1 H/s
[07:45:53] GPU0 | 69°C | Fan 2401 RPM | 1597.5 H/s
[07:45:55] Found Nonce, submitting...
[07:45:55] Share Accepted -> +1
[07:45:55] Shares: 47 Accepted, 0 Rejected | Total Hash Rate: 1884.1 H/s | Avg Search Time: 26.4 sec
[07:45:58] GPU0 | 70°C | Fan 2404 RPM | 1607.9 H/s
[07:46:02] GPU0 | 68°C | Fan 2404 RPM | 1613.7 H/s

[bostonvex]

There is an update available:

Cast XMR Version 0.5 (2017/10/19)

  - support for multiple Vega GPUs. Specify multiple GPUs comma separated with the -G switch like "-G 0,1,2,3"
  - monitor temperature and fan speed for each GPU
  - subtle performance improvements in core

Download Cast XMR 0.5 for Windows (64bit)


Now hash with full 4 KHashes/s with 2 GPUs:

Runs well on 3 GPUs and getting better hashrate than on XMR-STAK.

Awesome stuff!

[bostonvex]

Nice job.  Thanks for the specific miner. I seem to have a problem that no one else has really mentioned. I have an MSI Vega 56, and I have no problem getting 1860h/s for about 5 minutes... then it drops by 200h/s to 1660h/s... and is stable at about that rate. I can kill the miner, and start it back up and it does the same thing. I thought maybe it was a heat thing, but after playing with settings, I don't think so. Looks like perhaps a bug. Any thoughts? Maybe it is related to the other use who said they loose 200h/s on their first card when they have a second card. This system does have an integrated graphics card.

try it "-I 9"

Didn't change... updated to 0.5 version also.

[07:25:14] Initializing GPU, loading kernel ...
[Suspicious link removed]pute Driver detected.
GPU0: gfx901 | 64 Compute Units
[07:25:15] Connecting to pool.supportxmr.com:7777 ...
[07:25:15] Connected to pool.
[07:25:15] New job received.
[07:25:15] Difficulty changed. Now: 25000.
[07:25:19] GPU0 | 64°C | Fan 2078 RPM | 1679.1 H/s
[07:25:23] Found Nonce, submitting...
[07:25:23] Share Accepted -> +1
[07:25:23] Shares: 1 Accepted, 0 Rejected | Total Hash Rate: 1739.5 H/s | Avg Search Time: 8.0 sec
[07:25:25] GPU0 | 65°C | Fan 2282 RPM | 1887.0 H/s
[07:25:30] GPU0 | 65°C | Fan 2479 RPM | 1875.1 H/s

.....

[07:45:27] GPU0 | 69°C | Fan 2441 RPM | 1886.6 H/s
[07:45:33] GPU0 | 69°C | Fan 2441 RPM | 1885.7 H/s
[07:45:38] GPU0 | 69°C | Fan 2441 RPM | 1887.6 H/s
[07:45:44] GPU0 | 68°C | Fan 2444 RPM | 1852.2 H/s
[07:45:49] GPU0 | 69°C | Fan 2401 RPM | 1629.1 H/s
[07:45:53] GPU0 | 69°C | Fan 2401 RPM | 1597.5 H/s
[07:45:55] Found Nonce, submitting...
[07:45:55] Share Accepted -> +1
[07:45:55] Shares: 47 Accepted, 0 Rejected | Total Hash Rate: 1884.1 H/s | Avg Search Time: 26.4 sec
[07:45:58] GPU0 | 70°C | Fan 2404 RPM | 1607.9 H/s
[07:46:02] GPU0 | 68°C | Fan 2404 RPM | 1613.7 H/s

What are your setting for:
1) Frequency
2) Memclock
3) Power limit
4) fan speed