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

[SmokinCache]

FYI for those reporting missing hashes in this miner.

4 Vega system

Cast report 7922 avg
nanopool reports last 6hr calculated at 7892

I would assume this is close enough. Am I wrong in that assumption?

[MickDaster]

FYI for those reporting missing hashes in this miner.

4 Vega system

Cast report 7922 avg
nanopool reports last 6hr calculated at 7892

I would assume this is close enough. Am I wrong in that assumption?

what are your setting and which cards exactly do you have (brand)?

[dragonmike]

Hi, can you confirm whether any type of bios-mod needs to be done to achieve these values (i.e. close to 2KH/s) on Vega56? I'm reading loads of scary stuff over in the Vega thread... and I was under the impression that BIOS modding wasn't needed to get these figures using Cast XMR?

[Ledipaa]

FYI for those reporting missing hashes in this miner.

4 Vega system

Cast report 7922 avg
nanopool reports last 6hr calculated at 7892

I would assume this is close enough. Am I wrong in that assumption?

Would be good to know whats your %. Since im pretty sure 100% of your shares were accepted. Press S on miner to see stats.

Anyway im jealous, I only get 7782 with 4 vegas. 56@64 is doing better than real 64 with same settings and clocks

[Ledipaa]

Hi, can you confirm whether any type of bios-mod needs to be done to achieve these values (i.e. close to 2KH/s) on Vega56? I'm reading loads of scary stuff over in the Vega thread... and I was under the impression that BIOS modding wasn't needed to get these figures using Cast XMR?

Pretty sure you need to increase HMB volts via bios mod to achieve these numbers. Or with much higher power draw by keeping core clock high.

[Morgon]

the need to toggle the HBCC if the hashrates are down after a reboot

You don't even need to do that. With a Windows SDK tool called 'devcon' and a batch file, you can get the full power of your devices without Crimson by disabling/re-enabling your entire suite of GPUs in one click.

See the 'Advanced Users' section of this post. You can get devcon by following these instructions. (Note that 'amd64' is simply 'x64' - that's the version you want.)

today i will do more testing with vega will keep you posted.

Yes blockchain drivers are merged with new crimson driver

It's been shown that the newer drivers do not give the same performance as the August drivers, do you have experience otherwise? The newer drivers have a 'Graphics <-> Compute' switch, but is only available for the AMD x80s and not Vega. It's my understanding that nobody is getting 1900+ H/s with anything other than the August drivers.

Looking forward for the Linux miner. Using windows headless is a pia.

What's your normal desktop? If Windows or macOS, just turn on Remote Desktop on your headless machine (assuming it is running Windows 10 Pro). Otherwise, there's always TightVNC, which can even run when in Safe Mode for using DDU.

[Sinskin]

Does anyone have problem with hash rate drop after closing Teamviewer session? I have stable 1800h/s per card mining speed with my 8x VEGA 56 mining rig when Teamviewer is on. But if I close Teamviewer and come back after couple of minutes then cards have dropped to 1500h/s. I have HDMI Dummy Plug and I have disabled lock screen from Windows and I'm using blockchain AMD drivers. Also I disable and enable all gpu's before starting XMR-miner.

OC Settings of GPU's (testing phase):
Core: 1200 / 950mv
Memory: 915 / 950mv
Power Limit: -20%

My hash rate drops from 1960 to 16xx after some period of time.  I haven't been able to pinpoint a cause yet.  I run TeamViewer as well but the hash rate drop seems random.  

Just curious, what brand of Vega 56 do you have?  Mine is a Gigabyte.  I have MSI and PowerColor coming this week and am going to do a comparison so I can figure out which brand to standardize on, if it makes any difference at all.

I have Sapphire. Now hash rate have been stable even after closing Teamviewer. I'm not sure was it related to power options that I changed. From advanced settings I changed PCI Express Link State Management setting from "Moderate power savings" to "Off" and from Display settings that monitor doesn't turn off. Also I double checked that I disabled "Lock on session end" from Teamviewer client side that was on Action menu.

[tf2addict]

Hi, can you confirm whether any type of bios-mod needs to be done to achieve these values (i.e. close to 2KH/s) on Vega56? I'm reading loads of scary stuff over in the Vega thread... and I was under the impression that BIOS modding wasn't needed to get these figures using Cast XMR?

I get 1956 H/s on my Gigabyte Vega 56.  No BIOS flash to a Vega 64.  Blockchain driver.  No HBCC memory, just use the disable/enable card trick.

The only settings I changed are P3 memory to 975 MHz, power limit to -10%, and freq. to -30%.

[R0mi]

Thank you for your reply

Don't think they are in Crossfire as that option is not available after the fourth card is installed. And before I purchased the two 56s the registry edit you speak of did not prevent them from restarting in crossfire. I will dig into this more tomorrow and see if I can see the crossfire indicated somewhere other than AMD Settings.

<snip>

That's not the case on my setup.  Two places to check: in the Registry folder for each card, the pathway is listed at the top of each SoftPowerPlay mod, check for EnableAutoCrossFire.  Is it set to "1"? If so, set it to "0".

Then check in AMD settings.  With 4+ cards crossfire can still come up as "ON" and it creates problems.

[rambetiko]

I'm really impressed by the vega XMR mining experience.

Aside from the regedit powermod needed to drop voltages down to 880mv range, and the need to toggle the HBCC if the hashrates are down after a reboot, the rest is quite amazing.

Target temperature of 48ºc:

-GPU @ 48ºc
-HBM @ 56ºc,
-Fan auto @2200rpm.  

There is barely any heat coming out.  The noise is a fraction of the 290x's noise. And it gets 1950 kh/s. (HBM cant do 1100mhz otherwise it would get 2000kh/s). And its getting it with only 1337mhz on the gpu.


Next thing I'll try to get voltages even lower (overdriventool can't get lower than 880mv), is to use the good old Offset (afterburner) and try dragging it down to ~830mv.

hello. Can you share the setting screenshot. because i hate fan sound

[spyshagg]

I'm really impressed by the vega XMR mining experience.

Aside from the regedit powermod needed to drop voltages down to 880mv range, and the need to toggle the HBCC if the hashrates are down after a reboot, the rest is quite amazing.

Target temperature of 48ºc:

-GPU @ 48ºc
-HBM @ 56ºc,
-Fan auto @2200rpm.  

There is barely any heat coming out.  The noise is a fraction of the 290x's noise. And it gets 1950 kh/s. (HBM cant do 1100mhz otherwise it would get 2000kh/s). And its getting it with only 1337mhz on the gpu.


Next thing I'll try to get voltages even lower (overdriventool can't get lower than 880mv), is to use the good old Offset (afterburner) and try dragging it down to ~830mv.

hello. Can you share the setting screenshot. because i hate fan sound

Not at home.

But I used this regedit mod file to unlock the lower voltages first:
https://drive.google.com/file/d/0B7hntsldmLZBTWwxcmN4WENrejA/view

Reset Computer and run overdriventool.  On  GPU P7 I have 1337mhz@870mv  and on HBM P3 I have 1070mhz@850mv

Target temperature is set to 48ºc (this is the setting that determines the fan speed in AUTO mode, but its important to keep HBM bellow 65ºc for better timings).  Min fan @ 1700rpm Max @ 4900rpm

That is it.  With target @ 48ºc my fan is spinning @ 2200rpm.  The fan could go inaudible If I set target @ 60ºc.   

Good luck

[SmokinCache]

FYI for those reporting missing hashes in this miner.

4 Vega system

Cast report 7922 avg
nanopool reports last 6hr calculated at 7892

I would assume this is close enough. Am I wrong in that assumption?

what are your setting and which cards exactly do you have (brand)?

From post above

Here is the system

ASUS Prime Z270-AR
Celeron G3930
16GB RAM
250GB 960 EVO in M.2_2
Vega64 PCIEX1_1
Vega64 in PCIEX16_1
Vega56 in PCIEX16_2
Vega56 in PCIEX16_3
Supernova 1600 G2 PSU
All Vegas are Sapphire and the 56s have the 64 BOIS
Monitor plugged into Intel IGPU

The PSU unit and the two 56s are all under 14 day returns at a local store that I couldn't pass up on at the price. $300 each for the 56s. The PSU is overkill at the moment for xmr since it is only pulling 710 watts at the wall, but we don't know what the future holds and I plan to add 4 more cards as I can find them at a reasonable ROI.

I have performed the Soft Power Play Edit as described here.
https://www.reddit.com/r/MoneroMining/comments/74hjqn/monero_and_vega_the_definitive_guide/

Also the batch file I use to start the miner contains the Devcon.exe disable/enable as described here. I am not loading any profiles as in the outdated procedure described.
https://bitcointalk.org/index.php?topic=2002025.msg23619746#msg23619746

These setting are not changed from the links provided above. There is still some tweaking to do as I have one card giving about 1% of total hash compute errors. I am waiting on a few parts to be able to test each card outside of this machine. I can not be certain which card it is with the reporting/control errors I have found. and I don't want three cards idle for days while I have time to run longer term tests. I am about 95% sure it is the 64 in PCIEX16_1 but I will let it eat with the failures for now.

https://i.imgur.com/nFtH66I.png

Thank you for your reply

Don't think they are in Crossfire as that option is not available after the fourth card is installed. And before I purchased the two 56s the registry edit you speak of did not prevent them from restarting in crossfire. I will dig into this more tomorrow and see if I can see the crossfire indicated somewhere other than AMD Settings.

<snip>

That's not the case on my setup.  Two places to check: in the Registry folder for each card, the pathway is listed at the top of each SoftPowerPlay mod, check for EnableAutoCrossFire.  Is it set to "1"? If so, set it to "0".

Then check in AMD settings.  With 4+ cards crossfire can still come up as "ON" and it creates problems.

I changed all registry keys to 0. As I stated above this had no effect on the two Vegas retarting with crossfire and ULPS (shuts second card down vegas will have green light on tach instead of red) enabled and on. It has had no effect on vard control or reporting of variables. Nor did it change anything in AMD settings. Still no crossfire option available.

 https://i.imgur.com/aqMglpl.png

FYI for those reporting missing hashes in this miner.

4 Vega system

Cast report 7922 avg
nanopool reports last 6hr calculated at 7892

I would assume this is close enough. Am I wrong in that assumption?

Would be good to know whats your %. Since im pretty sure 100% of your shares were accepted. Press S on miner to see stats.

Anyway im jealous, I only get 7782 with 4 vegas. 56@64 is doing better than real 64 with same settings and clocks

https://i.imgur.com/c6oFLFY.png

[SmokinCache]

Devcon.exe update for those using iGPU (I guess)

I noticed some instability issues after I started using the devcon disable/enable trick. I have since changed to using devcon.exe restart and the stability issues are gone. I use a single line in the batch file now.

devcon.exe restart PCI\*DEV_687F*

It restarts the cards one at a time. As noted in the original posting https://bitcointalk.org/index.php?topic=2002025.msg23619746#msg23619746 there may be issues with disabling all cards when using iGPU. I was disabling one and then enable then do three at a time but it was taking to long for my liking.

Hope this helps someone. And I can only report my experiences it may not be the same on other systems.

[cryptohit]

Hello,

I'm new to this forum, but have been trying the Cast XMR in comparison to Stak for the last few days, and definitely would say it's the better of the those two. Am running Cast on 2 rigs, and noticed that after about 12-15 hrs of running the rig would disappear from my pool's stats. When I check the rig, it's running, the Cast software is still going (appearing to hash) however, the hashrate and rig itself seems disconnected from the pool. When I hit "q" to seamlessly quit, it starts posting several lines of "Outdated" share results, as if it had been backed up and couldn't post these before.

Anyone else experiencing this? (and for info I am mining XMR on supportxmr pool, running 8 Vega 56's on each rig)

Any suggestions on what might be causing it?

Thanks.

[SmokinCache]

Hello,

I'm new to this forum, but have been trying the Cast XMR in comparison to Stak for the last few days, and definitely would say it's the better of the those two. Am running Cast on 2 rigs, and noticed that after about 12-15 hrs of running the rig would disappear from my pool's stats. When I check the rig, it's running, the Cast software is still going (appearing to hash) however, the hashrate and rig itself seems disconnected from the pool. When I hit "q" to seamlessly quit, it starts posting several lines of "Outdated" share results, as if it had been backed up and couldn't post these before.

Anyone else experiencing this? (and for info I am mining XMR on supportxmr pool, running 8 Vega 56's on each rig)

Any suggestions on what might be causing it?

Thanks.

I have seen this once. I sent 1150 instead of 1050 and crashed the system. I was in a rush to leave and restarted everything quickly and left. It only ran about 15 minutes before hanging. When I got home and hit "s" and the cntr-c it did what you described. I was still using wattman then to reset the cards after a reboot. I chalked it up to me forgetting something on the restart because I was in a rush.

[cryptohit]

Thanks for the quick reply. I tested this a few times throughout the last 3 days, and it is consistent (after about 10-14 hours of straight mining). Of the 4 times, one of them had the rig slow down it's hashrate from 1900 h/s to 1500 h/s. I don't use Wattman for anything EXCEPT fan speed control.

[SmokinCache]

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

[cryptohit]

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

Wow, that's awesome, I have not tried the stock BIOS. I am running Flashed BIOS for the 64's and am getting about 1900+ per GPU. This is also using the 1100MHz reg settings.

https://i.imgur.com//a/3fCrQ

[SmokinCache]

Thanks for the quick reply. I tested this a few times throughout the last 3 days, and it is consistent (after about 10-14 hours of straight mining). Of the 4 times, one of them had the rig slow down it's hashrate from 1900 h/s to 1500 h/s. I don't use Wattman for anything EXCEPT fan speed control.

That 1500h/s rate sounds like the cards need to be reset. Opening GPUz, HWiNFO64, or anything like this will cause the need for a reset. In fact as I was typing this I opened file explorer the look at the spelling of HWiNFO64 and it cause all of mine to got to 1500 and 1700. Hmmm

[cryptohit]

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.