Funny enough my strix hynix does better than my msi airbost…lol…
im on the gold coast so hr south of brisy… gona be a mission to move all the rigs here ?
Hi Silver, I have the same card 4 nos, The card detects then goes offline without temp and percentage when i boot it comes back up . Please help.
Using these settings, my vega 56 reference card drops to 33mh… Not sure why? If I flash it to 64, I can get up to 186mh stable at 150v, but 56 seems to drop the hash significantly.
Hello. 580 on Hynix give 63Mh, but 470 on Elpida easy give 66-67Mh (as fact this 470 may give 70Mh, but rig in my bedroom, i don’t want speedup coolers over 50%). All card flashed one click timings in Polaris Editor!
33 MHs is typically the unstable card result. Try the memory timings with different overclock/undervolt settings
1 Like
Next question is… once I find the right BIOS for this “Sapphire Radeon RX VEGA64 Nitro+”, how do I know which overclock and Strap should I use?.
Any of you have a “Sapphire Radeon RX VEGA64 Nitro+” flashed to Vega 56?
EDIT: I used again @Smining570 data for a Sapphire he has and… bingo 
2 Likes
Does anyone have good settings for Gigabyte 64 flash to Gigabyte 56? It is the strangest Vega I have. Unlike other ones, they only way to get hashrate to rise is to lower both clock and voltage. I lowered to about 1300 MHz and 800 mV while holding memory constant 900 before it got unstable and it settled around 175 MH/s at 165 W. Pretty abysmal comparatively. And yes, I verified this odd relation on the pool and it does hash what it claims. Also have 2 and both behave the same.
Is it a reference (1 blower) or aftermarket (2 fans) ?
I have one aftermarket type since recently and also struggling to get it running like the others. At first I updated to the F6 bios, but it was capped at 190W and therefore the hashrate was limited. I’ve found another F6 bios that goes up to 210W but decided to flash the FA1 which goes up to 222W. The card doesn’t reach the cap anymore, but I’m still on summer settings with :
core 1340
vdd 860 (still in process of lowering step by step)
mem 847
soc 860
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
This gets me around 183MH/s for 193W
Credits go to @ss-silver
Don’t use the same settings as you use for 64 on 56.
Can you tell us which settings exactly you have applied? Which card is it? And which bios have you flashed?
They might help for a few °C, but don’t expect some extra household fans to lower temps a lot.
You need to evacuate the hot air outside of your house.
And what lots of us do during summer is running at lower clocks so voltages can get lower resulting in lower wattage meaning lower heat. This off course also means lower hashrate, but the cards might live longer too.
Just read on another forum that the gigabyte cards report the ACTUAL consumption, whereas the other cards report the lower amount. This makes more sense now, will tweak like other Vega’s. It is an aftermarket 2 fan model.
Yes, that can be read on this forum too 
Nevertheless, this 64 to 56 one consumes more energy than his original 56 brothers, whilst for reference cards the 64 to 56 tend to consume less than their 56 versions.
please can you help me? after flashing from 64 to 56 bios why is the GPU detected dead on TRM… I’ve changed the oc and the riser I’m using, but the result is still GPU detected dead…
What is the brand and type of card and what is the new bios you’ve flashed?
And what are your OC’s and straps now?
I’m using a vega 64 samsung amd/ati gpu and flash using a samsung vega 56 saphirre and here’s a photo of the oc and straps
this is my vega 64
and this is my flash gpu 56 saphirre
So a lot of things with Vega could be throwing it off. When all else fails, I try a vega 56 bios that is the same manufacturer 64 that it is (i.e. gigabyte 56 for gigabyte 64). It can and has been a riser for me in the past, but you already tried that. Click on those crashes and show us the text, I have one Vega that crashes repeatedly due to a hardware issue and its apparent from the crash text, but thats unlikely. Also, Ive noticed they hate lower voltage so first things first is try 900mV. Lastly your 930 memory is very high, I usually hit around 915 - 920 if its a reference blower card and 930-940 for better aftermarket cards. Lastly, use ss-silver mem straps, not sure which you are using. Heres my settings
Just pulsing what other miners are doing - Ive noticed that despite getting the pool hash I anticipate, I am only getting paid about 75% of my predicted payout from nanopool for Ergo. Since profits have dipped so much, are others considering going back to ether? I’ve found the profitability may be swapping. Love these cards that we can swap between the two and both are efficient.
@ ss-silver Just for tests sake, I compared your mem strap vs stock hiveOS suggested straps, and found that at least for ether yours provide 1 MH/s more and no power increase. Massive help - thank you!
nice I do switch between ERG and ETH still, I am running ETH right now but ERG block time is back to 2 min so will probably be back on ERG soon.
for ETH I run
amdmemtweak --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
if they get invalids I bump to RC 38, if more invalids RCDRD 13
@Elysium have not decided if I want to move the rigs or just pay to have them hosted in the US while I am out there, but im sure I would try to build another rig out there, if I can find cards…