Post
Topic
Board Mining (Altcoins)
Re: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v9.7 (Windows/Linux)
by
TechPark
on 11/07/2017, 12:01:45 UTC

1200W is definitely enough, it can easily take 1300W out of the wall.
I would suggest you to double check the basics:
- reset BIOS to default
- set PCI-E to Gen2 and disable built in video adapter
- do not use Molex on MB if you have powered risers
- setup default BIOS and settings on all cards.
- install all the latest updates for Windows (run>winver>must be 1703)
- uninstall with DDU and reinstall the drivers for Nvidia recommended by Claymore ("10xx cards in Windows 10 x64: just use latest 372.54 drivers from Nvidia website, note that you must have Win10 Anniversary update")
- setup RAM >4GB
- setup virtual memory > 16GB
- setup exclusions for AntiVirus software
- make sure that you have no more than 2 card per SATA PSU cable
- use only essential commands  in config/bat file
- avoid using temperature or overclocking settings keys inside config/bat file
- run 1 card only
- use MSI AB to setup up minimal settings - decrease Core clock and memory clock
- if stable add a second card
- reinstall the driver
- setup up minimal settings - decrease Core clock and memory clock
- add more cards one by one
- if running stable on minimal settings increase each parameter one by one
- if stable you may try to overclock it
- use GPU-Z and HW info to monitor the hardware

Post screenshots here. We might notice something odd.



I will try this and get back to you, but this is now weird.

Havent changed anything but some oc in afterburner, 75% Power, 85 temp, -400 core, +400 mem and its been running for 16 hours.
Tried this as well yesterday with a Little modifications and its jsut crashed after a while.

Checking in afterburner, it usually crash when the % Power make a down spike or drop, does anyone know why this happens? Do I need to add more voltage or Power to remove the spikes? Because i guess thats what causes the instability.
Also, spikes usually happens when i get a new job in claymore, maybe it should be like that.

Uploaded image of it, usually its the last gpu that fails. http://imgur.com/a/fJTTt

Edit : In my Control panel, Power management is set to optimal Power by default, is it best to have that or maximum performance?

This spikes are down spikes not up spikes.
When the core is not doing much work there it does not take much power you see it as down spike.
If you suspect the power then follow my advices above related to power. leave other tasks for later.