A little different BIOS version, byt everything else in my card is that same. Try these settings::
Works stable for about a week.
OMG i have the same hynix!,is stable?
First, I’m new to Hive, so there’s a learning curve there… lol
I have three MSI Vega 64’s with Samsung Memory that I run in Windows and have very little issues with them.
I’ll use one card as an example here.
Running RedTeamMiner I get a solid 50.8MHs ea. Which I’m happy with. My temps stay low so hopefully the cards last. Want to try a 56 bios someday but like not having any issues. Just sits there and runs.
I have OC and straps settings from Windows and was hoping that I could just use these in Hive and get up and running but no luck.
In Windows Overdrive I use GPU P7 1100 MHz 850 mV, Memory 1107 MHz 850 mV.
The strap settings are: --CL 20 --RAS 30 --RCDRD 14 --RCDWR 12 --RC 44 --RP 14 --RRDS 3 --RRDL 6 --RTP 5 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 14 --REF 17000 --RFC 249
In Hive I entered OC setting as follows:
Core Clock: 1107
Core Voltage: 850
Memory Controller Voltage: Did not enter anything
Memory Clock: 1100
Memory Voltage: 850
Fan: 100
Power Limit: Did not enter anything
Soc Frequency: Did not enter anything
Added Tweak: amdmemtweak --CL 20 --RAS 30 --RCDRD 14 --RCDWR 12 --RC 44 --RP 14 --RRDS 3 --RRDL 6 --RTP 5 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 14 --REF 17000 --RFC 249
With these setting the rig goes offline, everytime.
I set the miner only to use one card so I could try and test things out, but I seem to be missing something.
I do see Hive talking about setting the P state, 5 is default but not sure if this is relating to my P7 and P3 Overdrive settings.
I feel like a broken record reading all the posts before me, but it gets overwhelming.
Can someone point me in the right direction here?
I’ve got the same gigabyte vega56 card and the same problem with adding straps.Did You manage with that by the time?
My max for now is 46mh/s…
It is not working for me I can not add any straps to my vega56 Gigabyte (113-D0500300-X06) …any solution for that?
No. I switched my drive back to Windows for now.
I’ll be revisiting this after the holidays.
If you find anything, please let me know.
Thx
For now I get 46,72mh/s on hiveos without straps…It is pretty stable for couple of hours.
I tried downgrade- nothing
flash to another bios vega56- nothing
a lot of straps settings- always freezes the card and rig dont see it after reboot.
I managed to get 54Mhs with Hynix memory, but it gives occasional hw error, and sometimes I get “gpu dead”. So I used 960Mhs. It works stable.
Excuse me sir. Is ur Powercolor Vega 56 hynix is triple fan (Red Devil) ? Or u modbios to lower TDP?
is the cooling system fine? did you try it in linux?
yes, it´s triple fan. I belive i flashed that with some modded bios from Tech power up…dont remember for sure, sorry
Sorry don’t have the card anymore 
i have the same msi . for months i ve got 46.85 mhs… if this works il be your createst fan!!
i will try and give feedback .
cheers!!
For some reason, HiveOS registers invalid shares on one or more the GPUs if you mess with certain tweaks. I’ve had the same issue, but its interesting to note that those invalid shares are just a glitch and not actually registered as invalid shares to the mining pool. The only time I actually submitted invalid shares “officially” was when I failed to upgrade to the latest version of TRM on time when the Epoch #1 commenced. No biggy.
For my 2 Hynix cards, I have two different OC settings that have been running stable for me so far.
Try these two:
RX VEGA 56 Hynix Card 1: (This one is for my most annoying card that keeps giving me GPU Detected Dead error every couple of days)
Core Clock, MHz: 1352
Core Voltage, mV: 900
Memory Clock, MHz: 900
amdmemtweak --RC 36 --RP 14 --FAW 12 --RAS 20 --REF 65535 --RRDL 0 --RRDS 0 --RCDRD 19
RX VEGA 56 Hynix Card 2: (This card has been running stable and hasn’t had any issues)
Core Clock, MHz: 1450
Core Voltage, mV: 960
Memory Clock, MHz: 925
amdmemtweak --RC 36 --RP 14 --FAW 12 --RAS 20 --REF 65535 --RRDL 0 –
RRDS 0 --RCDRD 19
Also, try to keep the Core Temp of each card under 60 degrees. I’ve checked the crash logs and everytime I notice that GPU Detected Dead error executes a restart when temp on one of my cards exceeds 60 degrees for some weird reason. I boosted blower fans and added an oscillating fan to increase atmospheric airflow. Now stable around 52 degrees across the board.
Hello everyone,
Happy New Year and best wishes to all.
I allow myself to ask my question because I need help
I have a 64 gigabit vega with the id
I can not find the equivalent of bios 54 apart from this one
is it possible to take a bios or the device reference id: is the same but the subsystem identifier is slightly different. ?
Thank you
sorry for my English…
Here’s an update for all you folks having issues with RX VEGA 56/64 Cards on HiveOS, mining ERG on autolykos2 algorithm.
Before the nerf, Vega 56/64 cards were pushing around 190 - 200Mhz.
After the nerf, Vega 56/64 cards now push around 135 - 145MHz.
However! Overall network hashrate decreased and there hasn’t been any decrease in payouts since the nerf. So it’s nothing to worry too much about. It wouldn’t make an impact if an optimization came out across the board for Vega 56/64 GPUs. The percentage of resources you allocate to each block remains the same, thus resulting in virtually the same exact payouts because overall network hashrate would increase respectively. Its obviously correlated. Before the nerf, total network hashrate oscillated around 25Th/s. Now we’re looking at around 15TH/s for total network hashrate, thats around a 40% decrease in overall network hashrate.
NOW. Following the Vega GPU update or nerf or whatever you want to call it, a 180-200MHz nerf to 130 - 145MHz, yields on average a 30% decrease in hashrate power. With that being said, your percent contribution should be slightly higher per block than it was before, thus an increase in payout rewards. Correct me if I’m wrong, idk. this is how I’m looking at it.
SO BASICALLY what I’m trying to say is to STOP overstressing about your hashrates and just find a stable config that works for your cards. If you’re achieving anywhere between 130 - 145 MHz and not crashing, or producing invalid shares, you’re fine. You won’t be earning any less than you were before the emergency release. Also, don’t worry too much about GPU Detected DEAD error. No matter what you do, there’s always gonna be that one pesky card that just hates being alive. Check my previous post to see how I tackled that issue.
One last thing, After following the performance and payout intervals over the course of the last couple of months since the nerf, I haven’t made any more or less ERG than before the update.
P.S. Sorry for the rant. Hope this helped put some minds at ease.
Impressive for Hynix memory, however some cards can do even better:
Hello everyone,
I’m just starting out with an MSI vega 56 airboost (no luck! But I’ve had it for a long time).
After several days I have a hard time stabilizing it.
The best I have at the moment is something like:
Core 1040Mhz
Vdd 840Mhz
Mem 910Mhz
amdmemtweak --CL 20 --RC 36 --RP 13 --WR 13 --CWL 7 --FAW 12 --RAS 23 --REF 65535 --RFC 248 --RTP 6 --RRDL 5 --RRDS 3 - -WTRL 9 --WTRS 4 --RCDRD 19 --RCDWR 11
(posted by Oni)
I still have a few invalid shares, 95% ratio …
Before applying the tweak I could go up to 960Mhz without crashing (but with at least the same amount of invalid shares).
I have not yet tested the ApocWarz solution (a few messages above), but consumption scares me.
Can invalid shares be linked to Vdd directly? (I no longer have a “dead gpu”, so I tell myself it’s ok).
In what I am describing I am between 46 and 47Mhz. I can climb more but a lot of error. Maybe I’m doing it wrong, or the card is really bad.
Hello,
Finally I have something really stable, but low.
Thanks to ApocWarz for his stats.
I would like to understand something, now that I am stable (and that I have lowered the core and vdd), my average hashrate in 2miners is lower than advertised (it was the opposite in recent days).
Does the speed of the core have an influence on the shares validated at this point?
(I have barely 42Mh / s in 2miners life).
(Notice that there has been a TRM update in the meantime)
