Hi all,
I am building a small rig which currently has 2 570 and 1 1060. I face an issue with the 570 ones which I suspect that has to do with the lack of connection to a monitor.
The first gpu is a 570 (Elpida) and when I connect using teamvier (or connect a monitor) everything seems fine as I do get a 28 mh/s after having modified the bios. If I do disconnect from teamvier then automatically the hashrate drops to 24 mh/s. The 1060 is not affected by teamvier or the monitor.
Today I installed a second 570 (hynix) which is the second gpu (stock 19) following exactly the same steps (same drivers, same patch on drivers and then a new bios. In fact I tried two bios one with 1950 and one with 2200 on the memory but there was no difference. I get always 24mh/s and i suspect that it has to do with the lack of monitor ( I suppose the teamviewer only uses the first 570).
I know there are some dummy dvi plugs which I will try. Is there any other reason that I may have this big difference between the two 570 and how can I solve the issue?
ps: In all videos in youtube I see rigs without any dummy plug.
What drivers? Are you using compute workload?
Hi
at this moment i use 4 rx580 nitro+ 8Gb (Micron)
with bios RX580MicronModded_Undervolt_v1 and driver Beta for [Suspicious link removed]pute on windows 10 64b
I mine with claymore V10 with
EthDcrMiner64.exe -epool eu1.ethermine.org:4444 -ewal wallet.MainRig -epsw x -esm 0 -estale 0 -asm 1 -dcri 9 -cclock 1250 -cvddc 950 -mclock 2250 -mvddc 950 -tstop 85 -tt 65 -fanmin 25 -fanmax 100
But my system is instable.
I would like to test this bios mod:
https://mining-bios.eu/product/mining-bios-sapphire-rx-580-nitro-8gb-special-edition-micron-performance-timings/?currency=EURwith driver Beta for [Suspicious link removed]pute
What is the best for my cards or what do you think about this ?
TY
Too high memory clocks...