You won’t get 54 with Phoenix miner use team red miner with B mode.
Anybody can copy them you just need to use amdmemtweak which is already in hiveos.
To run it at startup you need to ssh the rig and edit xinit.user.sh add amdmemtweak + the straps save the file close and enable user interface mode in the maintenance tab
Thank you, That worked. But I am having a problem where I get this error when I run Teamredminer regardless of the OC. Looked everywhere and could not find a fix yet. Any ideas?
and if I use Lolminer i can go up to 48 MH with memtweak but GPU 0 dies!!
I get the same on recent version but 0.6-195@210127 works fine
Same thing here with H110 D3A MOBO and Asus vega 64.
I managed to get 43MH though with the stock presets of Hive.
If you got anywhere let me know please, I am spending days and got nowhere yet.
At a point, I flashed to 56 and managed to get 47MH but it killed the card in PCI 1.
I think it has to do with ASUS cards only.
Anyone with ASUS Vega 64 managed to get more than 43MH!
If anyone did, PLEASE share experience.
Thank you
Anyone have any experience trying this with a liquid cooled Vega 64? Same straps and stuff might work on there?
You are a lifesaver 
, I did downgrade hive os and used TRM ver 7.16 and finally I got the Vegas to work on Hive.
But whenever i apply the memtweak amdmemtweak --1 --CL 20 --RAS 26 --RCDRD 12 --RCDWR 12 --RC 38 --RP 12 --RRDS 3 --RRDL 5 --RTP 6 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 13 --REF 24000 --RFC 239 , the GPU crashes
I also tried amdmemtweak --1 –CL 20 --RAS 26 --RCDRD 12 --RCDWR 12 --RC 38 --RP 12 --RRDS 3 --RRDL 5 --RTP 6 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 13 --REF 24000 --RFC 239
and this 
amdmemtweak --2 --CL 19 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 12 --RRDS 3 --RRDL 3 --REF 15600 --RFC 250
Now I managed to teach 46.9 without any tweaks.
I saw in the release notes for the latest version a mention of the amdmemtweak, but still no dice! It still won’t automatically load in HiveOS
Yes. Got it lined out in trm. Running solid at 55.1 actually. Thanks a ton for the write up
Are you stopping the miner then applying then restarting?
Did you ever figure out a fix to this? I also have a Gigabyte Vega64 (flashed to 56) and I am having issues getting MH/s up and wattage down. I can get about 45MH/s with OC before applying the straps, but I think every strap timing I’ve tried at 1000clock 950mem crashes the card. On top of that, I can’t get power consumption below like 160-170w
I figured it out, thanks :).
1- I have removed overclocks from all other cards,
2- applied the tweak ( amdmemtweak --1 --RAS 26 --RCDRD 12 --RC 38 --RRDS 3 --RRDL 4 --REF 21000 )
3- applied settings to the vega card flashed with 56 ROM.
4- The max I could get was 52.3 MH but I lowered it to 51.7 because i was getting rejected shares.
5- Applied the OC to the other cards.
6- And voila looooool.
7- I even got a couple of more MH on the other cards
Thank you for your Help
Hello Guys,
I have 8 Vega 64 (Sapphire Radeon RX Vega 64 8176 MB · AMD/ATI Samsung HBM2 · 113-D0500100-104)
I am trying to flash them to Vega 56 in order to get 55MH as other miners here are getting.
But I am not able to find matching Vega 56 Bios to flash.
Though from VGA Bios Collection | TechPowerUp
I flashed with all 3 bios one by one
Miner only works with stock settings. Once I do OC then miner does not start mining nor gives any error. If i clear OC settings then after rebooting it start mining at 38MH with default settings.
I am not able to do anything except going back to vega 64 stock bios.
Any advice would help and appriciated.
Hello. I just wanted to say big thanks to all those who contributed to this thread so far. 
I have 2 x 6 vega 56 rigs, running for months at 40MH… in one hour now all are running at 50MH approx. except one which dies when I apply the straps.
And with less power consumption than before, running cooler.
I will have to dig a bit deeper to get the 5MH from that last card, and also try to push for the 54MH I saw from @cr0001 and @Rhox
Maybe it has something to do with some other settings… I can’t yet see why my hiveos settings are a little different than the screenshots I see
this is how the whole configuration looks like on my side
I will have to check it in more detail.
thanks again to everyone
Roberto
Hello guys,
Thanks for this thread.
Unfortunatly, I haven’t been able to reproduce those perfs.
I have a 6x VEGA56 rig (Sapphire Pulse version, 5x Samsung HBM2 + 1x Hynix HBM2), all cards flashed at the latest BIOS version (same model, the Sapphire Pulse).
I am using HiveOS 0.6-191@210201, with OpenCL 19.30. The OC settings are (to begin with) :
- Core : 1100
- VDD : 870
- Mem : 1025
To OC my cards, i created the following template and assigned it to the worker (select the worker > Overclooking template > choose the right template > Apply.
First, when I apply the OC, although there is no error, I feel like the OC is not really applied. When running a “amd-info”, the mem states are still very low and the core states very high, which is giving (I guess) poor hashrate and a very high consumption :
- mem state : 800
- core state : 1590
Those states do not change a lot if the miner is running (TRM). For example when the miner is running:
The mem state is still 800 and the core state is a bit lower but not as expected.
==> Is it normal or am I misinterpreting something?
Then, I also applied the following straps :
amdmemtweak –CL 20 --RAS 26 --RCDRD 12 --RCDWR 12 --RC 38 --RP 12 --RRDS 3 --RRDL 5 --RTP 6 --FAW 12 --CWL 8 --WTRS 4 --WTRL 9 --WR 13 --REF 24000 --RFC 239
==> I guess this is too aggressive as the miner doesn’t start, and for other straps it crashes after 10sec
So I tried many different straps, from this thread and others, but without really better results. Even when I get some a bit “stable” like this :
amdmemtweak --CL 19 --RAS 30 --RCDRD 16 --RCDWR 6 --RC 44 --RP 13 --RRDS 5 --RRDL 5 --RTP 4 --FAW 18 --CWL 6 --WTRS 4 --WTRL 9 --WR 15 --WRRD 1 --RDWR 18 --REF 17000 --RFC 248
The mining perfs are very bad. i mean, with a wattmeter i got something arround 240 W/card (although the dashboard show 165W/card), with a poor performance of 36MH/s after stabilization.
Does someone could help me solve this please? I guess there are 2 problems :
One is the OC itself as it seems not applied
The other is finding the right straps.
Cheers,
You say you have Vega 56 that’s the issue. You need Vega 64 flashed as 56 for this tutorial
Hello, do you know how to clear the amdmemtweak parameters to clear this conf? I deleted the overclocking parameters but that does not delete the amdmemtweak parameters, we should be able to clear this conf … Thank you very much in advance
Once you restart the memtweak is cleared. You have to apply it each time you restart your rig. In my case, if I apply the tweak after the rig starts it freezes, so I apply the tweak while the rig is starting. Once I see the loading network stage. This how i got it to work.
It was what i understood but teamreader seems to keep in emmory the settingd it is weird
perhaps it is beacause of this script i found in the forum in a topic
i think this script make a memory setting
wget https://github.com/mostnonameuser/HiveTweaks/releases/download/1.1/timings.tar.gz && mkdir /home/user/timings && tar -xvzf timings.tar.gz -C /home/user/timings && cd /home/user/timings && ./init