Any word on amd vega hash rates?

[zeef]

Hi Guys,

I have two  powerplay red devil vega 64, and using wattman i can hash 2050 h/s with mems in 1100mhz/950mv and in gpu -5% frequency ( 1486mhz/925mv) on P7 strap.
Im using the last drivers 18.4.1 from amd. The card is consuming 200 watts. Can sofpowertable decrease the power consuption better?
What is the actual better way of creating the softable mod , using this ?  https://github.com/halsafar/Vega64SoftPowerTableEditor/blob/master/README.md

Thanks

[1714770492]

1714770492

[1714770492]

1714770492

[Riptide_NVN]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

[heavyarms1912]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

power tables work different with the adrenalin drivers.  Clocks/volts are not the same as in BC for me.

[zeef]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

Hi

Today i put all worked with 64´s red devil.
Forget all stuff you read.
For new drivers just install it and put -5% in frequency (850mv)
Mem: 1120 mhz (850mv)
-16% on limit of energy
temp on 55c because of hmb throttling

voilá 2050++ Hs with 180watts

Regards

[wudafuxup]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

Hi

Today i put all worked with 64´s red devil.
Forget all stuff you read.
For new drivers just install it and put -5% in frequency (850mv)
Mem: 1120 mhz (850mv)
-16% on limit of energy
temp on 55c because of hmb throttling

voilá 2050++ Hs with 180watts

Regards

If you aren't you using the play tables you're doing it the half ass way :p

[Mashy81]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

Lower your intensity. I have not been able to run anywhere near 1936. The optimum intensity is going to be different with every card.

[zeef]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

Hi

Today i put all worked with 64´s red devil.
Forget all stuff you read.
For new drivers just install it and put -5% in frequency (850mv)
Mem: 1120 mhz (850mv)
-16% on limit of energy
temp on 55c because of hmb throttling

voilá 2050++ Hs with 180watts

Regards

If you aren't you using the play tables you're doing it the half ass way :p

Hi,

Have you tested with new drivers too?
What is the best method actually?

Thanks

[carlo_0000]

Using adrenaline drivers I followed the guide and was slowly tuning clock speeds on my six gpu red devil 56 rig.  Suddenly seemingly out of the blue the system was unstable and started to bootloop on me and just generally became a giant pita.

Rolling back the power tables to 75% and clocks back down to 850mhz helped a lot.  Still I am not able to make it much past 24 hours before the display driver bombs out.  Using the same 1936 intensity in the guide.

I'm desperate now and have pulled memory back to 800mhz and loaded the stock tables.  I don't know what else to do about this.  Has anyone had something like this happen?  It is really beginning to make me angry.   

I have no explanation for the sudden change in stability.  It is almost as if one or more of the cards can't handle power tables any more.  I have even re-installed my OS and made sure to keep windows 10 from updating to 1803.

The guide was written a while back and was for the blockchain drivers.  Could the power tables in the guide not work as well with the adrenaline drivers?

Lower your intensity. I have not been able to run anywhere near 1936. The optimum intensity is going to be different with every card.

yes depend on gpu  i have 1880  above and after a time miner freeze

[lexele]

The OhGodACompany team have announced that they may start working on a (devfee) hashrate fix for Vega cards. Will we finally hit the theoretical limit of 70+ MHs on Ethereum? 

Lmao wut? Complete bullshit. Even if it was true you would still make more money on CN-7, CN-LiteV1 and CN-Heavy over Ethash. 70Mh/s barely makes it somewhat an option.

Nope, you'd make more with Etash in that case.

Probably not, because everybody (well not exactly  but it's for rough explanation) will have the same raise and then at the end you won't be doing 60% more reward as expected.

[Kubiz88]

...the miners change and my intensities dont work, odnt dont work.

xmr-stak?

'
xmr stak
and xmr aeon stak

Try SRBminer with intensity 112. its working fine. or use CastXmr.
Why xmr-stak?!

because of its integration with the JJ's hash monitor. I never figured out the cast xmr;s UI that sets hash rate thresholds.

anyway I did use cast xmr and got similar hashrate. Not sure where the problem is. Using win 10 1709

Go back and look on the Cast thread for StellarX88's modified JJ hash monitor script.  That works with Cast + JJ hash monitoring, and works very well.

JJ has released closed source precious Guardian too to work around this too. As of now I am still using the one I modified myself for both Xmr-stak & Cast-Xmr. Let me know if you need assistance to set it up on your rig.

Hi Stellax88,

I have seen your post on cast-xmr on hash monitoring. are you still using it after the monero 7 fork?

can you share your script?

[nordmann666]

two cards on my system and mining monero

Ref. Vega MSI 56 (1400core and 1075mem) with 3 monitors pluged in and water cooled
Ref. Vega 64 XFX (1400core and 1100mem)


with SRBMiner - vega 56 got 1900hs (goes down to 1800hs after hours) and vega 64 got 1960hs (stable)
with cast_xmr  - vega 56 got 2000hs (stable) and vega 64 got 1880hs

very confusing...i can identify the cards at fan rpm and temp (water cooled = 0RPM / 40°C) - so no bug with wrong ID or something




another confusing point

GPU-Z shows me:
vega 56 use 7800MB mem
vega 64 use 7358MB mem


both card HBM active but 0MB

anyone?!

[R0mi]

...the miners change and my intensities dont work, odnt dont work.

xmr-stak?

'
xmr stak
and xmr aeon stak

Try SRBminer with intensity 112. its working fine. or use CastXmr.
Why xmr-stak?!

because of its integration with the JJ's hash monitor. I never figured out the cast xmr;s UI that sets hash rate thresholds.

anyway I did use cast xmr and got similar hashrate. Not sure where the problem is. Using win 10 1709

Go back and look on the Cast thread for StellarX88's modified JJ hash monitor script.  That works with Cast + JJ hash monitoring, and works very well.

JJ has released closed source precious Guardian too to work around this too. As of now I am still using the one I modified myself for both Xmr-stak & Cast-Xmr. Let me know if you need assistance to set it up on your rig.

Hi Stellax88,

I have seen your post on cast-xmr on hash monitoring. are you still using it after the monero 7 fork?

can you share your script?

Yes, the script absolutely works after the fork.  At least it does with Cast; you simply insert the --algo=[whatever you are trying to mine].

StellarX88 has posted links at least a few times in the Cast thread to his script.  Search for his posts there.

[djkice]

I have 4x STRIX V56 using BC drivers

I'm mining CN-Heavy

changing p7 1411 950 and p3 910 910 getting around 1400h/s a card

on my reference I use PPT, i seem to have issues with using similar PPT. I did modify mem values as I know that I can't achieve 1100mem on custom cards.

is there suitable PPT for STRIX by any chance ?

[rednoW]

My personal opinion that for heavy better to use 18.3.14 driver. Blockchain driver is not stable on Heavy with max intensities

[Eisenkopf89]

I have 6x Sapphire Nitro+ Vega 56, anyone knows why two of them can't reach hashrate of others?
https://prnt.sc/jbcymv

You will have to give us more information for us to help you. Some guide questions listed below should help:

1. Are you sure that you did overclock them correctly? Seems like they are at default clock settings.
2. Did you install the soft power play tables? This is a must for maximum profitability too
3. What driver version are you using?

1. I OC them all in Wattman but these two drop hashrate after a short period of time (1hour or so).
2. No, please provide with info how to do this.
3. Blockchain drivers, I tried new drivers but my cards was crashing like every few hours with them so I stick to blockchain.

I also tried flashing them to bios 64, but that again didn't work well they were crashing for some reason. Do I need to flash them again to 64?

Thanks

I have the same Problem with my third Vega 56. 2 runs stabel @ 1900 H/s and the other one is just 1200 H/s. And what is power play tables?

[djkice]

My personal opinion that for heavy better to use 18.3.1 driver. Blockchain driver is not stable on Heavy with max intensities

do you use PPT on custom cards also or are you on reference cards?

my reference running on blockchain run fine on high intensity about 1280h/s 140w 1407/1100 ? stable 18+ hours

ill try the drivers on my custom cards as the above settings seems to crash and not detect on vega on reboot. I then need to reboot once more for it to show up. (could this be a bios thing perhaps?) I set all to gen2 and above 4g enabled.

EDIT: Tried 18.3.1 - lose around 250h/s a card

[rednoW]

My personal opinion that for heavy better to use 18.3.1 driver. Blockchain driver is not stable on Heavy with max intensities

do you use PPT on custom cards also or are you on reference cards?

my reference running on blockchain run fine on high intensity about 1280h/s 140w 1407/1100 ? stable 18+ hours

ill try the drivers on my custom cards as the above settings seems to crash and not detect on vega on reboot. I then need to reboot once more for it to show up. (could this be a bios thing perhaps?) I set all to gen2 and above 4g enabled.

EDIT: Tried 18.3.1 - lose around 250h/s a card

my ref cards are on blockchain drivers mining xmr 2000-2100h/s with srbminer.
My non-refs, including Hynix-based Sapphire Pulse Vega 56, are running 18.3.1 and Heavy algo. Hynix are 1450+h/s, Samsung 1500+h/s

All cards are using SoftPowerPlay tables. For blockchain drivers and ref cards it is custom table I posted here some time ago with increased SoC clock.
For non-refs and 18.3.14 drivers I prepare SoftPowerPlay tables easily with OverdriveNtool on basis of card's own bios.

[djkice]

yeah i think i used ur ref for example ages ago.

if you can give us a quick guide on non-ref cards using bios that would be amazing

[Riptide_NVN]

I removed and reinstalled w/DDU 18.4.1 adrenaline and am trying to optimize Wattman. Power consumption with a single XFX Vega64 LC card and a Ryzen 1700 is around 310w. Is this good and if not how can I improve? RAM is set to 1100mhz 905mv. GPU is set to 1212/1407mhz and 905/925mv. Power limit set to -20%.  Radiator fan is spun to around 2500rpm which keeps core temp in the mid 40s.  HBM is around 5-10c warmer.

I am using XMR-STAK and have been unable to get higher than 1960 h/s. Both indexes are set to 1824 intensity with comp mode false. Every time I play with these hash seems to get worse not better and I see driver crashes show up in event viewer when closing XMR out,

Am I at the limit then? TIA

[StellarX88]

...the miners change and my intensities dont work, odnt dont work.

xmr-stak?

'
xmr stak
and xmr aeon stak

Try SRBminer with intensity 112. its working fine. or use CastXmr.
Why xmr-stak?!

because of its integration with the JJ's hash monitor. I never figured out the cast xmr;s UI that sets hash rate thresholds.

anyway I did use cast xmr and got similar hashrate. Not sure where the problem is. Using win 10 1709

Go back and look on the Cast thread for StellarX88's modified JJ hash monitor script.  That works with Cast + JJ hash monitoring, and works very well.

JJ has released closed source precious Guardian too to work around this too. As of now I am still using the one I modified myself for both Xmr-stak & Cast-Xmr. Let me know if you need assistance to set it up on your rig.

Hi Stellax88,

I have seen your post on cast-xmr on hash monitoring. are you still using it after the monero 7 fork?

can you share your script?

Hi, sorry for late reply, didn't browse this thread for a bit. Yes, I am still running the modified JJ script monero 7 fork, even though with the new driver chances of hash drop is lower, it's still a failsafe to run it. Hit me up on Skype @ mo.jin.ee if you need direct assistance/modification with the script.

This is the version that I shared: https://drive.google.com/open?id=1TnicWxpB1TtGQCe_cuXND8ha-XsgQsU1

I have a somewhat updated and modified version pre Precious Guardian that I am running now, but I don't think there's much difference there.