Pages:
Author

Topic: Any word on amd vega hash rates? - page 24. (Read 202722 times)

sr. member
Activity: 378
Merit: 258
February 22, 2018, 02:50:47 PM
Very long ago, some experienced here said they would be a big failure if they did not reach 80mhs in eth with vega cards.
XMR will soon switch to new algo, and look like vega will stay as one big epic fail.
Like their custom version, came out too late in too small a volume.
jr. member
Activity: 69
Merit: 2
February 22, 2018, 02:33:49 PM
Hi guys I got a prob with my 6 vega 56. After I follow sum instruction in this forum like disable-enable GPU, powertable mode with P7 stock 1407 :950 MaV (Bernest setup) , and then run overdriveNtools after that run CAST XMR , everything work fine until a few hours later my GUI Cast XMR just stop hashing/got frezze this really annoying I need to run again-again-again til I got really tired about this case anyone please help how to fix this case. T T  Cry Cry Cry

crossfire disabled and EnableUlps too?
newbie
Activity: 26
Merit: 0
February 19, 2018, 03:59:21 PM
Hi guys,
Is there anybody who has a soft power play table for vega rx 64 liquid edition? I mean specially for liquid, because there is a fan which has max rpm 3300. With normal sppt are max rpm very high and I am worried about the damage of fan (higher voltage/current than is max for this fan).
I am lookong for version with 800MHZ for HBM.
Thx

800mhz?  Remind me how to export mine for me?  I think the tables that everyone is using are lc tables, though.  Definitely check this thread.  My 3 Sapphire vega64 lc are pulling 545W from the wall, so probably 150W each.  I've got the fans on pretty low (low minimum and high max temp of 60 i think), but the rig seems to be fine (goes at least a week, no errors at least).  Also, max temp on the lc is 70.  Anything higher will cause you an error in overdriventool. 
newbie
Activity: 14
Merit: 0
February 19, 2018, 12:51:07 PM
Hi guys I got a prob with my 6 vega 56. After I follow sum instruction in this forum like disable-enable GPU, powertable mode with P7 stock 1407 :950 MaV (Bernest setup) , and then run overdriveNtools after that run CAST XMR , everything work fine until a few hours later my GUI Cast XMR just stop hashing/got frezze this really annoying I need to run again-again-again til I got really tired about this case anyone please help how to fix this case. T T  Cry Cry Cry
newbie
Activity: 141
Merit: 0
February 18, 2018, 04:22:32 PM
In case someone  interested...

I figured out, Blockchain drivers works more stable if Radeon Setting process will be killed on Win10. You can configure everything in Radeon Setttings app, close app, and run miner. I found miners running much much stable after this procedures.
newbie
Activity: 23
Merit: 0
February 18, 2018, 02:02:25 PM
Quote

I couldn't manage to make Windows recognize more than 3 cards using Asus Z270-A. What motherboard are you using?
Thanks.

Asus Z270-A works well with 8 vegas. Update bios, do all the bios tweaks(Gen2 everywhere and 4g) and dont use 5th pci slot - windows wont boot with gpu in it. Use windows 1709 - fall creators update or older.
https://i.imgur.com/WspQjU8.jpg
newbie
Activity: 23
Merit: 0
February 18, 2018, 01:33:06 PM
Got an MSI Air Boost V56, does 935/935 hash @ 950 HBM through PPT mods. Draws way less than 150w (130w if hwinfo's anything to go by). XMR-STAK intensity is 1940/1940 IIRC. Ambient temperature usually 65f or lower, cards never go above 57c under those conditions.

I have 2x Asus STRIX V56s and 10x Sapphire Pulses coming in. Think I should just sell them and find some 1080/1080Tis for relatively normal prices? Pretty tired of maintenance and tweaking on the AMD side. It's fun for the first couple cards, then it just becomes burdensome. Alternatively I could just hoard the cash and wait for Volta/Ampere?

what ppt did you use for MSI Air Boost i have 3 but they are unstable crash randomly more than 10 times a day and i have a lot of bad shares. any help is welcome.
dont exceed 905hz memory and u will get ~1820h/s stable. this is maximum u can get with hynix memory.
@905 1 of the gpu's freeze after making 1st share i am at 930;900 but very unstable
what should i use in p3 memory 905:900 is ok or should i feed a little bit more the gpu with power
doesnt matter what u put after memory clock. u can adjust mem voltage only via vbios flashing not via ondriventool.
member
Activity: 275
Merit: 11
February 17, 2018, 12:26:11 PM
Hi guys,
Is there anybody who has a soft power play table for vega rx 64 liquid edition? I mean specially for liquid, because there is a fan which has max rpm 3300. With normal sppt are max rpm very high and I am worried about the damage of fan (higher voltage/current than is max for this fan).
I am lookong for version with 800MHZ for HBM.
Thx
newbie
Activity: 2
Merit: 0
February 16, 2018, 01:02:28 AM
Hi, Greetings! can you share your msi rx vega 56 air boost settings with wattman please, I got one of this card and mining it with xmr but not able to drive the perfect settings.
member
Activity: 99
Merit: 10
February 15, 2018, 08:35:12 AM
Does anyone know a miner that supports X17 for Vega?
newbie
Activity: 74
Merit: 0
February 15, 2018, 04:31:34 AM
i'm selling a lot of vega 56/64 in marketplace

Pm if interested
legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
February 14, 2018, 09:23:11 PM
Anyone got a guide on how to get an FE to work with xmr-stak under linux?

It runs the rx 560 fine, but I get an error when I do clinfo.



Ubuntu 16.04, you need to install rocm. Also rocm needs the Vega to be on a PCIE 3.0 slot.
https://github.com/RadeonOpenCompute/ROCm

Code:
wget -qO - http://repo.radeon.com/rocm/apt/debian/rocm.gpg.key | sudo apt-key add -
sudo sh -c 'echo deb [arch=amd64] http://repo.radeon.com/rocm/apt/debian/ xenial main > /etc/apt/sources.list.d/rocm.list'
sudo apt-get install libnuma-dev
sudo apt-get update
sudo apt-get install rocm-dkms rocm-opencl



Add this to /etc/rc.local

Code:




#DOWNCLOCK MEM and CHIP
/opt/rocm/bin/rocm-smi --setsclk 0;
/opt/rocm/bin/rocm-smi --setmclk 0;
#DOWNCLOCK MEM and CHIP

su - -c "screen -S miner -dm /home//mine.sh"

sleep 120

#VEGA FE Begin

#MEM OC
echo 1 > /sys/class/drm/card0/device/pp_mclk_od;
echo 16 > /sys/class/drm/card0/device/pp_mclk_od;
#MEM OC
#DOWNCLOCK
/opt/rocm/bin/rocm-smi -d 0 --setsclk 5;
/opt/rocm/bin/rocm-smi --setmclk 3;
/opt/rocm/bin/rocm-smi --setsclk 0;
#DOWNCLOCK END
#VEGA FE END

screen -dsm bash /home//temp-loop-vega.sh

you need the following bash script to control FE temps for optimum HBM2 operation

contents of temp-loop-vega.sh

Code:
very_high_speed=250
high_speed=240
new_speed=175
min_speed=100
card=0
 while [ 1 ]
         do
                 temp_card=`/opt/rocm/bin/rocm-smi |  awk -v card="$card" '{if($1==card)print $2}' | awk -F "." '{print $1}'`
                 if [ $temp_card -gt 64 ]
                         then new_speed=$(( $new_speed + 5 ));
                                if [ $new_speed -gt $high_speed ]
                                then
                                        new_speed=$high_speed;
                                        echo -e "new speed=" $new_speed
                                fi
                                 /opt/rocm/bin/rocm-smi -d ${card} --setfan ${new_speed}
                                 sleep 1
                 fi
                 if [ $temp_card -lt 65 ]
                        then new_speed=$(( $new_speed - 1 ));
                        if [ $new_speed -lt $min_speed ]
                                then
                                        new_speed=$min_speed;
                                        echo -e "new speed=" $new_speed
                                fi
                         /opt/rocm/bin/rocm-smi -d ${card} --setfan ${new_speed}
                         sleep 1
                 fi

         done
newbie
Activity: 8
Merit: 0
February 14, 2018, 09:12:51 PM
My setup:

8 x Nitro+ Vega 56 + Cast XMR

AMD Blockchaindriver Global WattMan:

GPU Freq.: -30
Memory Freq.: 930
Temp: Max-75
Power Limit: -22
Speed(RPM): Max-Max

Windows 10 Virtual Memory Paging: 60000

Disable & enable every gpu card after every reboot before to start mining.

@Result:
http://ipfs-gateway.ipbc.io/ipfs/QmNVqW3DqXMJWimuGUvi5qXmyRkvXLECKCgrfAuwZXhmzA/3.jpg




I couldn't manage to make Windows recognize more than 3 cards using Asus Z270-A. What motherboard are you using?
Thanks.
jr. member
Activity: 88
Merit: 1
February 14, 2018, 09:01:51 PM
does anyone know how to mod timing straps on vega without vbios? There is a way.
newbie
Activity: 80
Merit: 0
February 13, 2018, 05:42:50 AM
Hi

DId you meet 'invalid AMD results' error for vega56? It is about 10% in my result  Angry Angry

Try to use cast-xmr with the option --fastjobswitch.
This should solve your problem.
newbie
Activity: 1
Merit: 0
February 13, 2018, 01:37:02 AM
Hi

DId you meet 'invalid AMD results' error for vega56? It is about 10% in my result  Angry Angry
full member
Activity: 729
Merit: 114
February 10, 2018, 11:38:35 AM
Did someone try latest Claymore's 11 dual modes (eth+blake2s or eth+keccak) with vega? Should be good but hot )) Is it stable?

44 MH/s + 1950 blake2s @ 1240/1100 clocks.  I switched back to cryptonight as the card was warmer and burned more power.
newbie
Activity: 60
Merit: 0
February 10, 2018, 10:51:33 AM
Did someone try latest Claymore's 11 dual modes (eth+blake2s or eth+keccak) with vega? Should be good but hot )) Is it stable?

The 11.0 miner is not fully optimised for VEGA yet.
legendary
Activity: 1510
Merit: 1003
February 08, 2018, 04:38:52 PM
Did someone try latest Claymore's 11 dual modes (eth+blake2s or eth+keccak) with vega? Should be good but hot )) Is it stable?
full member
Activity: 391
Merit: 105
February 08, 2018, 03:05:20 PM
I have quite the conundrum guys.. So I was flashing my 56 like normal (done it a million times to multiple cards B4) and accidentally flashed a corrupted 60KB rom into the GPU and bricked the first bios switch. No biggie, has happened to me before.

I booted into windows using the 2nd bios switch and popped it back into the first to fix the corrupted bios. Here is where I fudged up  Angry. I accidentally flashed the locked bios from the 2nd switch into the first switch (like a damn tard) which fixed the bricked 1st switch but now I can't flash either of them since they both got the locked version of the bios.

Is there any way to purposefully corrupt a bios to where I can recover it and flash a different one. Or is there a way to bypass the "can not erase rom" message?

Does the -f switch in the command line not work?

I actually figured it out. I'm dumb lol just ingnore that last post xD. That's for trying to help though  Smiley
Pages:
Jump to: