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

[SmokinCache]

That's interesting, I don't remember if I opened anything on the rig during the process but will note it for next time. Is there any way to add settings to Cast for having it reset/restart itself when it sees a GPU falling below a certain hash speed/threshold or when it determines there is a backlog of shares that have not been submitted? I think it would be helpful to do that. Once I have worked out all the kinks I am going to create a complete walkthrough of how to setup everything using Cast XMR with the Vega cards, and using optimal settings.

After a few days now Cast finally hung. I have been checking the pool reports every 1 to 4 hours and have notice a decline from the 8000+ to all the instant and long term averages being in the 7500 to 7800 range. Before lunch I checked and instant was at 4000 so I logged in with Teamviewer. The miner was scrolling but not finding anything. As I watched it stopped reporting has rate. I hit the S key and it did it's normal catchup scrolling after a hang. lucky I caught it within minutes. Now 2 hrs later the pool trends are point back at a ~8100h/s rate.

There has bee no keystrokes in the miner cmd window in about 15 to 24hrs

I see two possible scenarios

1: The miner slowly over time hangs to a full stop and hitting "S" clears some buffer or something.

2: the miner has many "micro" hangs and clears itself most of the time.

I have noticed the occasionally during my tuning and testing that the miner at times won't respond to a CTRL-C command. hitting S first will cause it to respond to the keystroke following the status report.


When I get home tonight I am going to write an AutoHotKey script to send an "S" to the miner window once every 10 minutes and see if the hash rate fall that I witnessed happens again.

[1714024120]

1714024120

[1714024120]

1714024120

[ShirKhan76]

I already use cast-xmr since about 10 days and it is very stable with 2x Vega56s and 3x RX 580s. I only had 3 crashes and the hashrate of my 580s are much better (approx. 50 h/s per GPU) than with xmr-stak-amd miner.
But I determined that I have approx. 6% outdated jobs:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

[Shares Found: 9364 | Avg Search Time: 29.9 sec]
8803 ( 94%) Accepted
   0 (  0%) Rejected by pool
   1 (  0%) Invalid result computation failed
   0 (  0%) Could not be submit because of network error
 560 (  6%) Outdated because of job change

@glph3k: When will you release a new version of your miner that this issue will be fixed?

[MickDaster]

I already use cast-xmr since about 10 days and it is very stable with 2x Vega56s and 3x RX 580s. I only had 3 crashes and the hashrate of my 580s are much better (approx. 50 h/s per GPU) than with xmr-stak-amd miner.
But I determined that I have approx. 6% outdated jobs:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

[Shares Found: 9364 | Avg Search Time: 29.9 sec]
8803 ( 94%) Accepted
   0 (  0%) Rejected by pool
   1 (  0%) Invalid result computation failed
   0 (  0%) Could not be submit because of network error
 560 (  6%) Outdated because of job change

@glph3k: When will you release a new version of your miner that this issue will be fixed?

could you share the config file for that miner?

[tf2addict]

I typically get 6-13% outdated shares no matter what pool I'm using.   My effective hash rate with XMR Stak seems to be much closer to what it should be, even if the reported hash rate is 5% lower than Cast.  Between that and the lack of an API with Cast I am going with XMR Stak for now.  I will reexamine Cast if and when it has an API for monitoring.

[markos365]

I made an script to run devcon, overdriventool and finally cast xmr. I set the Task Scheduler to run this script at statup whether user is logged or not. When booting the pc and not login the script will run in background, without opening the cmd prompt.
This result in a consumption of around 1015W in a 6x Vega 64 rig.
If I run the script manually, so it shows the cmd prompt, the power consumption is  between 100 and 150W higher.
I don't understand why this happen. I would like to know if you can replicate this results.
Of course, the downside of this is that you cannot know the hashrate you are getting.
Devcon was not running properly.


So, an API to get your temps, fan speeds, hashrate and so on would be great. Even better if it's compatible with Awesome Miner.

[tf2addict]

So, an API to get your temps, fan speeds, hashrate and so on would be great. Even better if it's compatible with Awesome Miner.

I'm trying to design a dashboard/monitoring solution for my farm.  I absolutely require some way to get hash rates out of the rigs remotely at a minimum, fan speeds and temps would be a bonus.  Claymore has that but it does very poorly with CryptoNight on the Vegas.   My current plan is to fork XMR Stak AMD and add an API (only the CPU version apparently has an API currently).  I am going to see if I can make a Claymore-compatible API so that it would integrate into Awesome Miner.

[fr4nkthetank]

thats an interesting project, good luck with that!  But maybe the dev can add the feature, its is still very useful to have.

[rohnin]

VEGA 56  Which BIOS?

I have been running the 64 BIOS in my 56s since startup before discovering this miner. I could never get either of them over the 1960-70 H/s. That was at 1100MHz. Today I flip the switches to the stock BIOS just to see what it would do. So far it is impressive. The entire system is pulling 20 to 25 less watts and the two 56s are now at the 2000H/s mark. You can look at their rates in my previous post. Here they are now.

https://i.imgur.com/qSWIVNO.png


We will see what the next 12 hrs shows

Is your system still using RX Vega 64 drivers for your Vega 56 cards?  What does GPU-Z show for your Vega 56s?

I'm on stock BIOS Vega 56s and used the 800hbm regedit but I'm unable to increase mem speeds higher than 950mhz and unable to get past 1960h/s.  So I'm wondering if you're able to push higher than 950 because you may still have Vega 64 drivers installed.

[vdkcoins]

Finally got single MSI Vega 56 to run stable with xmr-cast-vega.  My key was getting the temp right <65 to stay stable and keep MEM < 950
Block chain drivers and reset driver prior to running.  Very stable.

Freq: -8%
Mem: 945
Power +10
Rate:  1980
Fan: ~3400 min
Temp: 64
Ave Search 45 secs, no errors

or

Freq -8%
Mem: 935
Power +0
Rate 1950
Fan: ~2500 min
Temp: 62

Not bad, both very stable, but I have not been able to get to the fabled 2000


Here's my question.  I mined 24 hrs with nicehash
Nicehash reports between 1.33 and 2.2 at any given time and is quite variable on their BTC profitability curve through the day.
anyone have good/bad NiceHash experience or can explain this phenomenon?

Or, suggest a more profitable pool to point to?

Thanks

[techo]

I have 5 rx 570 and 1 rx 580. I am running amd stak gpu without any error but Cast xmr gives me error.

"Error cl_device_not_found calling clGetDeviceIDs for number of devices"

This is my config file: "cast_xmr-vega -S pool_adress -u my_wallet -p x -G 0,1,2,3,4,5"

What should i do?

[mikejennings]

I have 5 rx 570 and 1 rx 580. I am running amd stak gpu without any error but Cast xmr gives me error.

"Error cl_device_not_found calling clGetDeviceIDs for number of devices"

This is my config file: "cast_xmr-vega -S pool_adress -u my_wallet -p x -G 0,1,2,3,4,5"

What should i do?

Massively increase the amount of virtual memory. I ran into this problem myself today with 3xRX580s.

Though virtual memory doesn't actually seem to be used, there's some check when mining begins to see if there's enough room. Not sure why. But it exists.

[mikejennings]

Finally got single MSI Vega 56 to run stable with xmr-cast-vega.  My key was getting the temp right <65 to stay stable and keep MEM < 950
Block chain drivers and reset driver prior to running.  Very stable.

Freq: -8%
Mem: 945
Power +10
Rate:  1980
Fan: ~3400 min
Temp: 64
Ave Search 45 secs, no errors

or

Freq -8%
Mem: 935
Power +0
Rate 1950
Fan: ~2500 min
Temp: 62

Not bad, both very stable, but I have not been able to get to the fabled 2000


Here's my question.  I mined 24 hrs with nicehash
Nicehash reports between 1.33 and 2.2 at any given time and is quite variable on their BTC profitability curve through the day.
anyone have good/bad NiceHash experience or can explain this phenomenon?

Or, suggest a more profitable pool to point to?

Thanks

Try these; this gets me to 2000 h/s on my single Vega 56 card unmodded.

Core Frequency: -10%
Memory Clock: 965 MHz
Power Limit: +4
Fan: Maximum

Run using -I 9 switch on the command line as well to maximize intensity.

Nicehash is a different beast of a pool. Try mineXMR/supportXMR or the like if you want lower fees and to be paid in XMR, not BTC.

[bytiges]

I typically get 6-13% outdated shares no matter what pool I'm using.   My effective hash rate with XMR Stak seems to be much closer to what it should be, even if the reported hash rate is 5% lower than Cast.  Between that and the lack of an API with Cast I am going with XMR Stak for now.  I will reexamine Cast if and when it has an API for monitoring.

I can confirm. There's 4-5% old shares in my case too, on average. This totally destroys the supposed cast-xmr.
Not sure if the displayed hash rate is real or fake, but pool numbers don't lie.
Unfortunately, I cannot use stak because it's not stable for my overclock settings.

[ShirKhan76]

I already use cast-xmr since about 10 days and it is very stable with 2x Vega56s and 3x RX 580s. I only had 3 crashes and the hashrate of my 580s are much better (approx. 50 h/s per GPU) than with xmr-stak-amd miner.
But I determined that I have approx. 6% outdated jobs:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

[Shares Found: 9364 | Avg Search Time: 29.9 sec]
8803 ( 94%) Accepted
   0 (  0%) Rejected by pool
   1 (  0%) Invalid result computation failed
   0 (  0%) Could not be submit because of network error
 560 (  6%) Outdated because of job change

@glph3k: When will you release a new version of your miner that this issue will be fixed?

could you share the config file for that miner?

Of course, here we go:

cast_xmr-vega -G 0,1,2,3,4 -S pool.supportxmr.com:7777 -u 45gLm8HYupeW6f8LFL479GiZsQS8nXEj2ScaaJvB8gZQMJRG2bxEUZrCQsGPQzc4tHgu89tyoyBMr9Q fDK6xBnybPB4fTo7.1ce888860aab407 -p xxx -I 9

[Morgon]

Try these; this gets me to 2000 h/s on my single Vega 56 card unmodded.

Core Frequency: -10%
Memory Clock: 965 MHz
Power Limit: +4
Fan: Maximum

Yeah but how much of that are you giving to the power company? How many watts per card under those settings?

I have 5 rx 570 and 1 rx 580.

Cast does not work with x70 cards.

[Jackkyy]

I already use cast-xmr since about 10 days and it is very stable with 2x Vega56s and 3x RX 580s. I only had 3 crashes and the hashrate of my 580s are much better (approx. 50 h/s per GPU) than with xmr-stak-amd miner.
But I determined that I have approx. 6% outdated jobs:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

[Shares Found: 9364 | Avg Search Time: 29.9 sec]
8803 ( 94%) Accepted
   0 (  0%) Rejected by pool
   1 (  0%) Invalid result computation failed
   0 (  0%) Could not be submit because of network error
 560 (  6%) Outdated because of job change

@glph3k: When will you release a new version of your miner that this issue will be fixed?

could you share the config file for that miner?

Of course, here we go:

cast_xmr-vega -G 0,1,2,3,4 -S pool.supportxmr.com:7777 -u 45gLm8HYupeW6f8LFL479GiZsQS8nXEj2ScaaJvB8gZQMJRG2bxEUZrCQsGPQzc4tHgu89tyoyBMr9Q fDK6xBnybPB4fTo7.1ce888860aab407 -p xxx -I 9

pls send your settings and hashrate

thanks

[mikejennings]

Yeah but how much of that are you giving to the power company? How many watts per card under those settings?

At 4% overvolt it's about 208-212W per card. At 0.119 kWh pricing it's plenty profitable, about 1.02 XMR/month under current difficulty while the power cost is ~$18/month for the card itself. Given that 1XMR = ~$120 today, that's plenty of margin.

As for castXMR outdated jobs, I am running about 2.2% outdated jobs. Will have to compare this to xmr-stak-amd soon but it's far lower than what others are reporting here. I find the hashrate reported on supportXMR to be generally correct, adjusting for the fact that supportXMR's pooling/graphing doesn't really show perfectly with any miner that I've found. I get the right amount of shares, which is all that matters.

[ShirKhan76]

I already use cast-xmr since about 10 days and it is very stable with 2x Vega56s and 3x RX 580s. I only had 3 crashes and the hashrate of my 580s are much better (approx. 50 h/s per GPU) than with xmr-stak-amd miner.
But I determined that I have approx. 6% outdated jobs:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

[Shares Found: 9364 | Avg Search Time: 29.9 sec]
8803 ( 94%) Accepted
   0 (  0%) Rejected by pool
   1 (  0%) Invalid result computation failed
   0 (  0%) Could not be submit because of network error
 560 (  6%) Outdated because of job change

@glph3k: When will you release a new version of your miner that this issue will be fixed?

could you share the config file for that miner?

Of course, here we go:

cast_xmr-vega -G 0,1,2,3,4 -S pool.supportxmr.com:7777 -u 45gLm8HYupeW6f8LFL479GiZsQS8nXEj2ScaaJvB8gZQMJRG2bxEUZrCQsGPQzc4tHgu89tyoyBMr9Q fDK6xBnybPB4fTo7.1ce888860aab407 -p xxx -I 9

pls send your settings and hashrate

thanks

hashrate is already shown in my entire post:

[Hash Rate Avg: 6365.5 H/s]
  852.8 H/s GPU0
  862.3 H/s GPU1
  877.1 H/s GPU2
  1923.9 H/s GPU3
  1849.4 H/s GPU4

First 3 GPUs are 580s and last two Vega56s.

Which settings do you mean?

[meroh]

I'm running 4 Vega 56 with stock BIOS and was able to get 1950 h/s on 3 of them, but my 4th is only running at 1720 h/s. I am using block chain drivers, tried disabling and re-enabling everything and all of the cards are over clocked to the exact same settings. Anyone know of any reason why this might be happening?

[tf2addict]

I'm running 4 Vega 56 with stock BIOS and was able to get 1950 h/s on 3 of them, but my 4th is only running at 1720 h/s. I am using block chain drivers, tried disabling and re-enabling everything and all of the cards are over clocked to the exact same settings. Anyone know of any reason why this might be happening?

What brands?  I have a PowerColor (1850 H/s) and it does 100 H/s less than my GigaByte and Sapphire (1950 H/s) on the same settings.