You won’t do anything more with power consumption but I wonder what wattmeter would show, if it’s the real difference comparing to other manufacturers from the wall or it’s only wrong HiveOS value. I didn’t check it, I don’t have the card anymore.
Failed to initialize with these Tweaks.
Any suggestion which parameter to “fine tune” / change?
Rise RC to 38 and RCDRD to 18 and if it runs fine lower one by one every 15-30min.
its new TRM vers. with enhanced gpu detection… 
I rolled back to TRM 0.8.0 and no issues since…might be too high oc on some gpu tho…
Mem 965 is way to high for Hynix, start with 930 and rise every 5mhz
Nice good to see MH like this without falling .
My Vega 56 Samsung running nice and stable as well …
But in Hynix rig i had one problem with one GPU showing “got incorrect temperature 511, ignore”
so i removed it and now mining is stable but still LOW MH but better than nothing.
Back to “got incorrect temperature 511, ignore” - i want to change PCI Express extender for new one … or is there anything else ? Anyone have same issue like me ? Cables from PSU are okay.
Lucky you are, my Hynix are running 925 and 945. Don’t you have invalids?
good point … 
–CL seems to be a problem for my card. Everything below --CL 21 leads to an error.
Anything ideas?
no invalids, i get some on high cre Samsung… but its 1 2 now and then. these are solid
I had this issue when one power cable burned itself and also when I had problem with external fan which made some noise in electricity. You should check your electric devices.
1 Like
You can set CL 20+, this shouldn’t affect your hashrate
Thanks. Are RC and RCDRD the most important values concerning hashrate?
Does any of the Tweaks directly affect power consumption?
Hi, i tryed all the tweaks that we have here in my hynix, and it was the best i got just add it in the tweak oc table. amdmemtweak --REF 65535 --RFC 249
What are the straps on your image?
And REF as well. Yes, power consumption is slightly higher but it’s for sure worth changing.
1 Like
Regarding the Gigabyte Vega 64 (Gaming OC), bios version 113-D0500300-X06 (vega 56):
I tried to increase MHS and decrease wattage - nothing helped. As soon as I modify MEM to a number higher than 845, it crashes. Same with core frequency, anytime I increase above 1000, it crashes.
Straps impossible to add too - tried perhaps 10 of them, all of them made the gpu crap itself. So for now, I’m staying with this setting:
BUT! What I discovered, is that if I set the core voltage to 810, I can at least get the wattage down by 10-15W. (sometimes it drops down to 157W)
I wanted to ask - is this setting healthy? won’t it break the gpu?
1 Like
The lower the better (cooler gpu)
Got me a new Vega 56 today(GPU7)… and its Samsung! starting to get these Samsungs figured out.
Had a 64 bios from the previous owner so flashed the 56 back on to it, runs great at 1020 mem on the normal Samsung strap
amdmemtweak --CL 20 --RC 37 --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 12 --RCDWR 12
5 Likes