Ohhhh, thats bad news!

Hello,

I am preparing to switch my 4GB rigs to Linux based OS, but have this situation as described in subject.

9x470 GB on Win10 gives me 262MH/s @ 1050W on the wall. HiveOS with same OC, and downvolting gives less with much much higher consumption - 259MH/s @ whooping 1350W and cards are very hot. What I am doing wrong? Maybe there is any tricks to lower consumption? 

Hi,

On all rigs miner runs perfectly, but on one I can't start it. It just stops in "Setup miner ..." and cursor is blinking. What can I do to make it working?

But they are quite good on cryptonight about 5000 H/s

Yep, it is 30 GPUs. I don't want to waste my time selling it one by one. I already have such amount of GPUs. I just want to lower my cost of farm, because previous ones where bought in 2017 December 2018 January when high prices kicked in... Speaking about el. consumption I will squeeze easily 6GPU rig to 700 Watts with 180MH/s. So 3500W in total, i don't know how eowonderful could make them run with only 3000W.

kerney666 any solution to my issue?

0.11 USD/kWh

Friend of mine is selling rigs with RX574 in total of 900 MH/s for $3700. I think it is quite good price. Would You buy it during this "mining is dead" time?

Tahnk You. It was running perfectly on 18.8.1 drivers with the same clocks/voltages. But only 6 GPUs were seen. So I don't think that it is clocks issue.

Hi, thanks for good product. Had newer drivers, but in this setup only 6GPU from 8GPUS were recognised. Installed 18.6.1 drivers - now get this error. Please help

Thank you for the help. One question: is there way to set constant fan speed (i.e. 50%) in OverdriveNTool? Cause those spining and stoping down fans, makes me crazy ...

Hello, everynone!

I have few 4x Powercolor RX 570 Red Devil rigs. Each card produces ±29MH/s.

Windows 10 Pro, AMD Crimson-ReLive-Beta-Blockchain-Workloads-Aug23 drivers (22.19.659.0), BIOS moded 1500 straps, to 1625,1750, 2000MHz, Claymore miner 10.3

But have very very annoying problem:

I want to lower power consumption. As I understand best way is to lower GPU core clock. If I lower it on MSI Afterburner to 0.8V, GPU-z shows almost 0.92V. In such scenario one card uses about 95W (GPUZ), and about 125W on the wall. Seems quite good result, but if I touch any slider (i.e. change fans peed) on Afterburner, voltage goes back to 1.1+V and power rises to ±150W (GPUZ). Voltage increases back also when Claymore miner restarts. It is freaking annoying. Because i need to monitor if voltage is not changed back and I have overconsumption and additional cost and worthless heat.

Tried to use Claymore settings via .bat for voltage regulation - it won't work also

With Spahhire Trixx i can lower voltage as supposed, but I can't find how to do it on all cards.

Is it possible to disable that variable voltage, so that I always had constant 0.9V and forgot that routine maintenance of checking if voltage is the same. Plus, I think I can lower consumption to 85W (GPUZish) and have stable system.

P.S. I lowered memory voltage to 0.9V via BIOS editor. But somewhere on forums read, that if you mess with something more than straps, than variable voltage action kicks in action. Is this true?

My system screenshot: