- disable aggressive undervolting
- remove the Soc Frequency limit
- set Memory Clock to 900
Also make sure you’re using TeamRedMiner
1 Like
Ok thx I will. I was using NBMiner, I will change that as well.
I just tested, but GPU dies all the time. (Fixed now)
Using TeamRedMiner.
I will lower cClock and Core Voltage… I just copied some settings I found here from other Vega 56 AMD/Ati, so I thought they would work.
Anyway, I can go higher this time than 160MH/s
EDIT: 172MH/s now
I will try to increase values to see if I can reach 200MH/s or a stable hashrate with low power consumption.
Once you have a stable hashrate, what values should be increased in order to try to obtain a little bit more hashrate?
Be warned that 200+ mhs will require alot of power, like 160w+ (in reality 210w at the wall). It may not even be possible.
Also after some time you may get some invalid shares, which means you need to tinker with mem tweaks. Vega are reknowned for being variable, so you can spend alot of time with mem tweaks.
1 Like
Yeah I think I prefer it to be closer to 180-190MH/s with low power rather than 200MH/s.
Running this new strap on the Hynix last night, 8 hrs so far no invalids or crashes.
amdmemtweak --RC 35 --RP 11 --WR 14 --CWL 8 --FAW 12 --RAS 20 --REF 65535 --RFC 248 --RTP 5 --RRDL 6 --RRDS 3 --WTRL 9 --WTRS 4 --RCDRD 11 --RCDWR 10
Looks like only one card can run stable with those 190 settings, GPU5 got some invalids after about 9 hours on that strap and OCs.
1 Like
Is it possible to bios mod the Vega FE to 56? Just picked up a FE from a friend, highest I can get without any straps is 135mh looking to put it with my other 5 Vega’s and hit 190-200. Any help is greatly appreciated. Love this thread, been perfect getting my rig going.
Is it normal that Vega 56 constantly alternates between 171.3Mh/s and 156.6Mh/s?. It´s not a constant hashrate.
What strap do you use for those AMD/ATI GPUs?. Just in case they work for mine, as I can’t go higher than 171.4Mh/s and it´s not constant all the time. Thx!
EDIT:
Is this one?:
Samsung Straps:
amdmemtweak --CL 20 --RC 36 --RP 11 --WR 14 --CWL 8 --FAW 12 --RAS 20 --REF 65535 --RFC 248 --RTP 5 --RRDL 6 --RRDS 3 --WTRL 9 --WTRS 4 --RCDRD 10 --RCDWR 8
ya that’s basically the same strap. I took off the CL which does not seem to impact anything and was running RTP 4.
amdmemtweak --RC 36 --RP 11 --WR 14 --CWL 8 --FAW 12 --RAS 20 --REF 65535 --RFC 248 --RTP 4 --RRDL 6 --RRDS 3 --WTRL 9 --WTRS 4 --RCDRD 10
I am guessing that card is running effective core much lower than where you have it set, try 1364 on the core and drop mem to 880.
As for the hash dropping thats normal for ERG mining, approximately every two minutes a block is solved and mining stops temporarily so in TRM you see a slight reduction in hashrate.
2 Likes
Oh I see. Ok thx mate I will test with those settings 
EDIT: Nothing, changed those settings and GPU dead:
----------- [2021-09-01 18:04:14] Pool erg.2miners.com received new job. (job_id: 12b7) [2021-09-01 18:04:14] GPU 0: detected DEAD (07:00.0), will execute restart script watchdog.sh [2021-09-01 18:04:14] Watchdog script executor thread executing script 'watchdog.sh’
Argggggg
Maybe it doesn’t work for me because my version is:
Samsung HBM2 · 113-D0500300-101
and yours is -102
Anyway, I’m testing and improving my hashrate:
@ss-silver Hi, I’m using those settings on a Vega56 Asus Strix, and it was running fine for a couple of day. I had to turn off the rig, and putting it back to work it started to reboot every 2/5 minutes.
I’m now trying to put VDD to 814 (it was 812 before) but does somebody have an idea of why it has been stable for a couple of days, and now this strange behavior?
EDIT: the only difference is that this time the monitor is plugged and turned on, before the system was headless, can it be it?
I am not sure I have seen that before on Eth several times, rig runs for weeks then reboot and less hash or needs more power.
I resolved this once by taking off the card and cleaning it, but didnt work every time I tried that.
1 Like
Thank you for the great thread and advice so far! I managed to pick up a few more MH/s on GPU2 (its a gigabyte vega 64 flashed to 56 so I know the power shown is more than the others already.
GPU1, the sapphire nitro+ i found hard to get stable settings, at the moment this is a good as I can get, nice power consumption but I’d like another 15-20MH/s from it.
The Powercolour Red Dragon is a beast though! I might try and reduce the power consumption slightly on it. I have my PSU efficiency set at 90% in the HiveOS to make the power consumption a bit more realistic. For tweaking is it better to remove that factor for a better comparison?
One question guys. Is Vega 64 a good option as well for mining?. Or are they worst than Vega 56 and consume more power with less hashrate?. Thx
you can flash them to 56 models
Yeah but will they perform the same for hashrate / power or worst?. I think that some models have 3 PCIe 8 pin connectors, right?.
i have a vega 56 with 3 power connectors and its my lowest performer at the moment. the gigabyte 64 uses less power when its flashed to 56
1 Like
Mmmm interesting. The ones I found, have 3 power connectors ( Sapphire Radeon Nitro+ Vega 64), but I didn’t know if I should buy them or just wait for Vega 56.
Thanks!
