Any word on amd vega hash rates?

[opies32765]

no

don't touch clock in wattman

so dont touch wattman clocks IF i have done the .reg mod ? is that what you mean

thanks

[Rod1970]

Flashed all cards back to original bios....
There's something funny going on with wattman.

I have a 1920 threadripper with 12 cores and hashing with 11 (1 core to keep system responsive)
I have Large Page Supoort Enabled and hash at around 90 per core = ~1000

Wattman settings for the vega56's    945 mem / -30 core  - 20 power
(Got it to mine at 1080 mem once with one card at 1.000 volt mem and it hashed nearly 2000 ?!?! then it crashed after 15 secs.)

if i turn on HBCC it does mines at 900+880 = ~ 1780 per card
Stak settings 1522+1512
The card with the monitor attached is set to 900 mem to keep things stable...

All nice and well , rig is hashing around ~7500 with threadripper and 4 x vega 56's @ ~950 watts.

Point is, when i reboot cards hash @ around 550+550 = 1100. And everything in wattman looks the same.
But for some reason HBCC has changed and doesnt show any  changes  in wattman.
Settings look exactly as before rebooting.

If i adjust the slider , (any way) , it starts hashing @ 900 + 880 again.

So wattman does not correctly set HBCC by itself after reboot, and have to set is for all 4 cards again each time.

 i  have 32gb quad channel mem and 70.0000 swap file on fast NVME SSD.

First have to start stak cpu and then stak gpu , otherway around it bugs / hangs...

Any tips and tricks ?

[rednoW]

This HBCC bug after reboot is common.
Even worse ... sometimes hash drops without reboot and you need to reapply HBCC again.

[dharmeshchikkegowda]

I found these two videos on you tube.  both are fails

https://www.youtube.com/watch?v=Z48TzK529Eo
https://www.youtube.com/watch?v=SvLhEzSyouY

Guess we will need to wait for an update from Claymore before we see...

Anybody tried any other algos?

[melloyellow]

Flashed all cards back to original bios....
There's something funny going on with wattman.

I have a 1920 threadripper with 12 cores and hashing with 11 (1 core to keep system responsive)
I have Large Page Supoort Enabled and hash at around 90 per core = ~1000

Wattman settings for the vega56's    945 mem / -30 core  - 20 power
(Got it to mine at 1080 mem once with one card at 1.000 volt mem and it hashed nearly 2000 ?!?! then it crashed after 15 secs.)

if i turn on HBCC it does mines at 900+880 = ~ 1780 per card
Stak settings 1522+1512
The card with the monitor attached is set to 900 mem to keep things stable...

All nice and well , rig is hashing around ~7500 with threadripper and 4 x vega 56's @ ~950 watts.

Point is, when i reboot cards hash @ around 550+550 = 1100. And everything in wattman looks the same.
But for some reason HBCC has changed and doesnt show any  changes  in wattman.
Settings look exactly as before rebooting.

If i adjust the slider , (any way) , it starts hashing @ 900 + 880 again.

So wattman does not correctly set HBCC by itself after reboot, and have to set is for all 4 cards again each time.

 i  have 32gb quad channel mem and 70.0000 swap file on fast NVME SSD.

First have to start stak cpu and then stak gpu , otherway around it bugs / hangs...

Any tips and tricks ?

Are you running your core clock at 1115?

[l1xx]

I made it work on 5 it seems : http://prntscr.com/grm9ed

upped the virtual ram to 80. Disabled On board graphics , connected Monitor  with HDMI to GPU0.
Without Team Viewer going the hashrate is slightly higher - around 10500.
Only 4 cards accepted some change on the slider. The 5th had HBCC on, and on/off was not doing anything. On the 6th one HBCC was disabled. It was not even possible to enable it.

Sadly ... that lasted only until the first restart.
After it - we are back to 4. And I can't get em to apply any HBCC settings on at least 3 cards.

[cryptoinvestor_x]

I made it work on 5 it seems : http://prntscr.com/grm9ed

upped the virtual ram to 80. Disabled On board graphics , connected Monitor  with HDMI to GPU0.
Without Team Viewer going the hashrate is slightly higher - around 10500.
Only 4 cards accepted some change on the slider. The 5th had HBCC on, and on/off was not doing anything. On the 6th one HBCC was disabled. It was not even possible to enable it.

Sadly ... that lasted only until the first restart.
After it - we are back to 4. And I can't get em to apply any HBCC settings on at least 3 cards.

I wasn't even able to boot with 6 Vega as it would BSOD on driver install for the 6th card. I could get 5 hashing but similar results as you. 5th card has an HBCC slider but changes had no effect.

[melloyellow]

I made it work on 5 it seems : http://prntscr.com/grm9ed

upped the virtual ram to 80. Disabled On board graphics , connected Monitor  with HDMI to GPU0.
Without Team Viewer going the hashrate is slightly higher - around 10500.
Only 4 cards accepted some change on the slider. The 5th had HBCC on, and on/off was not doing anything. On the 6th one HBCC was disabled. It was not even possible to enable it.

Sadly ... that lasted only until the first restart.
After it - we are back to 4. And I can't get em to apply any HBCC settings on at least 3 cards.

I wasn't even able to boot with 6 Vega as it would BSOD on driver install for the 6th card. I could get 5 hashing but similar results as you. 5th card has an HBCC slider but changes had no effect.

I have the same situation.  I can't get HBCC to activate on more than 4 cards.  The sliders work for all but there's no effect for anything over 4 cards.

[kyogi14]

Just wanted to share some improvement I found.  I was having this issue where every 20 minutes to a few hours or so my hashrates would drop from 1800-1900 to 1500-1600.  I would go check the rig or login via teamviewer and simply restarting xmr-stak-amd would restore my hashrate.  This was driving me crazy as I hated losing that 600h/s across my two vega's.  My monitor was plugged in to the integrated graphics port on the motherboard and windows was putting the monitor to sleep after inactivity and I figured out this sleep mode was dropping the hashrate on my vega's.  Turning off the display sleep has preserved my hashrate so far.  I'm going to order a dummy plug to keep my hashrates up without having a monitor connected.

[mutual.consent]

Just wanted to share some improvement I found.  I was having this issue where every 20 minutes to a few hours or so my hashrates would drop from 1800-1900 to 1500-1600.  I would go check the rig or login via teamviewer and simply restarting xmr-stak-amd would restore my hashrate.  This was driving me crazy as I hated losing that 600h/s across my two vega's.  My monitor was plugged in to the integrated graphics port on the motherboard and windows was putting the monitor to sleep after inactivity and I figured out this sleep mode was dropping the hashrate on my vega's.  Turning off the display sleep has preserved my hashrate so far.  I'm going to order a dummy plug to keep my hashrates up without having a monitor connected.

I had the same issue and found the same fix on another forum. My main issue now is getting a good hash to power ratio. Without applying any soft power table mods I can get 1600 H/s at total system power draw of 200w at the wall for one Vega 64 (idle power draw is about 55w). That's with two threads in xmr-stak (2016 and 1600 intensity), -50% power limit, -30% GPU, and 900 MHz on the ram. Can anyone please advise whether this is a good hash to power ratio or whether I can do better?

[kyogi14]

I made it work on 5 it seems : http://prntscr.com/grm9ed

upped the virtual ram to 80. Disabled On board graphics , connected Monitor  with HDMI to GPU0.
Without Team Viewer going the hashrate is slightly higher - around 10500.
Only 4 cards accepted some change on the slider. The 5th had HBCC on, and on/off was not doing anything. On the 6th one HBCC was disabled. It was not even possible to enable it.

Sadly ... that lasted only until the first restart.
After it - we are back to 4. And I can't get em to apply any HBCC settings on at least 3 cards.

I wasn't even able to boot with 6 Vega as it would BSOD on driver install for the 6th card. I could get 5 hashing but similar results as you. 5th card has an HBCC slider but changes had no effect.

I have the same situation.  I can't get HBCC to activate on more than 4 cards.  The sliders work for all but there's no effect for anything over 4 cards.

I see this same issue even if the other cards are not vega, as soon as I add a 5th amd card I lose hbcc control on all. I wonder if it requires more ram or virtual memory? I’m running 2x Vega 56 and 2x580 and shits the bed when adding a 470 to the mix. I have 16gb ram and 48gb virtual. Hbcc would display as on, but hash rates show it’s not really on.

[Hellae]

Guys, don't take this comment the wrong way but right now we have old questions, which were already answered before, being asked, and old discoveries being shared again.

If you really want to help this thread, start by reading what others shared already and come up with new questions/ideas/discoveries.

[ingyenfrag]

Noob question, but how do you run multiple (2) threads in xmr stak amd and how do you adjust the intensitys of the second one?
I only get 6300 H/s with  6*vega 56  -30%, 950 HBM, -20% power. 

[zenstrive]

Maybe it's good for a mod to freeze this thread for a while and put every relevant info on first page

[MingMining]

This is just surprisingly strange (in good way  ), but I have no idea what is going on.

So on my test rig, there is only 4gb memory and I can only reach 1300 h/s on my vega 56. Today I tested it on another box and it reached 1800 h/s and I was very happy. But I cannot let it running long on this box coz it is just so noisy so I put it back to the old rig. After setting up wattman and starting the runner, I pressed h to see the hashrate. It is 1900 h/s and it is stable! I did not change anything! Still this rig has only 4g memory.

Anyone can explain this? Thanks.

[zenstrive]

This is just surprisingly strange (in good way  ), but I have no idea what is going on.

So on my test rig, there is only 4gb memory and I can only reach 1300 h/s on my vega 56. Today I tested it on another box and it reached 1800 h/s and I was very happy. But I cannot let it running long on this box coz it is just so noisy so I put it back to the old rig. After setting up wattman and starting the runner, I pressed h to see the hashrate. It is 1900 h/s and it is stable! I did not change anything! Still this rig has only 4g memory.

Anyone can explain this? Thanks.

Monitor attached to the GPU  both times? Vega has this quirk that when monitor is attached it works faster

[STT]

This HBCC bug after reboot is common.
Even worse ... sometimes hash drops without reboot and you need to reapply HBCC again.

A reboot is not needed, logging out will allow the driver to reset and be reapplied.  I found about 2 or 3% difference with hbcc or without but I guess that varies by use.

When setting volts I found it best to lower power limit but then later raise it to allow more performance, so long as the volts above are set low it should still consume less

[fmz89]

Flashed all cards back to original bios....
There's something funny going on with wattman.

I have a 1920 threadripper with 12 cores and hashing with 11 (1 core to keep system responsive)
I have Large Page Supoort Enabled and hash at around 90 per core = ~1000

Wattman settings for the vega56's    945 mem / -30 core  - 20 power
(Got it to mine at 1080 mem once with one card at 1.000 volt mem and it hashed nearly 2000 ?!?! then it crashed after 15 secs.)

if i turn on HBCC it does mines at 900+880 = ~ 1780 per card
Stak settings 1522+1512
The card with the monitor attached is set to 900 mem to keep things stable...

All nice and well , rig is hashing around ~7500 with threadripper and 4 x vega 56's @ ~950 watts.

Point is, when i reboot cards hash @ around 550+550 = 1100. And everything in wattman looks the same.
But for some reason HBCC has changed and doesnt show any  changes  in wattman.
Settings look exactly as before rebooting.

If i adjust the slider , (any way) , it starts hashing @ 900 + 880 again.

So wattman does not correctly set HBCC by itself after reboot, and have to set is for all 4 cards again each time.

 i  have 32gb quad channel mem and 70.0000 swap file on fast NVME SSD.

First have to start stak cpu and then stak gpu , otherway around it bugs / hangs...

Any tips and tricks ?

i'm doing 7500 at 650w(all system), 3xV56+1xV64
v56 stable at 1850, can go 1920+ need more voltage, i didnt test again
while v64 at 1970 with lower voltage,
vega truly awesome card

[posasky]

Flashed all cards back to original bios....
There's something funny going on with wattman.

I have a 1920 threadripper with 12 cores and hashing with 11 (1 core to keep system responsive)
I have Large Page Supoort Enabled and hash at around 90 per core = ~1000

Wattman settings for the vega56's    945 mem / -30 core  - 20 power
(Got it to mine at 1080 mem once with one card at 1.000 volt mem and it hashed nearly 2000 ?!?! then it crashed after 15 secs.)

if i turn on HBCC it does mines at 900+880 = ~ 1780 per card
Stak settings 1522+1512
The card with the monitor attached is set to 900 mem to keep things stable...

All nice and well , rig is hashing around ~7500 with threadripper and 4 x vega 56's @ ~950 watts.

Point is, when i reboot cards hash @ around 550+550 = 1100. And everything in wattman looks the same.
But for some reason HBCC has changed and doesnt show any  changes  in wattman.
Settings look exactly as before rebooting.

If i adjust the slider , (any way) , it starts hashing @ 900 + 880 again.

So wattman does not correctly set HBCC by itself after reboot, and have to set is for all 4 cards again each time.

 i  have 32gb quad channel mem and 70.0000 swap file on fast NVME SSD.

First have to start stak cpu and then stak gpu , otherway around it bugs / hangs...

Any tips and tricks ?

i'm doing 7500 at 650w(all system), 3xV56+1xV64
v56 stable at 1850, can go 1920+ need more voltage, i didnt test again
while v64 at 1970 with lower voltage,
vega truly awesome card

settings please...
very appreciated.

[rednoW]

So, at last I grab my new Vega56, flashed it with liquid bios and apply the same patch I used with my pure Liquid Vegas.
But I decided to clock mem for vega56 to 1090 instead of 1100 as it has not so good cooling as liquid.
Screen bellow, vega56 is running first 2 threads in miner, in hwinfo64 vega56 is the last card.
For Liquid Vegas I set a little bit higher voltage as they run xmr-stak with more intensity (to be sure stable).