I have an update regarding this issue:

- First I told you it happened on my 480 ASUS DUAL 4GB that was having an OCL thread hang, then at restart the GPU would stay in a lower performance mode, albeit having the clocks applied at miner start.

The problem is that this issue with non-rounded milivolt values with the miner makes the GPU have an OCL thread hang AND reset it's clocks. Then with the clocks reset when you open the miner back after the restart changing the default clocks/voltages makes the GPU enter in this lower performance mode.

With this ASUS DUAL 480 4GB, 1050core 1755mem and custom strap it hashes 25,35mh/s normally, 21,9mh/s on this lower performance mode. Restarting the PC makes it go back to the higher hashing value.

My specs:

Windows 10 x64 Enterprise - BUILD 10240
Driver - AMD blockchain Beta driver driver, 23 aug (windows driver version 22.19.659.0)
GPUs:
1x ASUS DUAL 480 4GB
1x Sapphire 470 Nitro+ 4GB
3x MSI 750 Armor OC 4GB

They all exhibit the changing clock/voltage -> lower performance until reboot issue that IMO is ADL related (as it happened always for me with Polaris GPUs, not only with mining, general use and gaming too, you can even see how the bandwidth is cut with AIDA or other bw gpu bench when you alter settings at least on windows with polaris GPUs).

The x25mv x50mv x75mv x00mv only voltage issue is a deduction of mine after seeing the gpu instant hang when using other voltages than these ones on the ASUS DUAL 480 4GB.

First post in this forum, just to say good job Claymore on the 10.0 version, it seems to take really well the clock and voltage options now if paired with AMD GPUs and the blockchain driver.

I would like to ask Claymore 2 questions:

- I have probably have run into a little issue in regards to core voltage settings with the miner. I have one GPU that when set to non-rounded voltage values (as the GPU only recognizes 6.25mv steps, this means all the voltage values that arent x25mv, x50mv, x75mv and x00mv) it misheaves pretty quickly and has OCL thread hang just a minute after starting mining (DAG is already loaded and no EDC errors are shown. If I use rounded values, it seems to take them better and if a OCL thread hang happens, it happens after several hours and probably blamed to memory clock instability.

- As you know, the clock and voltage control API from AMD has a mechanism built in that in the end reduces performance very time you touch any core/mem voltage/clock values if they were changed AFTER the GPU entered, even if for just a little moment, in maximum core and memory Pstate. I have seen when doing the OC/UV settings with your miner that the settings I put are enforced just before the DAG loads, but still with soft monitoring the GPUs enter for a brief period in max clock/voltages before that, causing that any clock/volt setting enforced by the miner, if it happens to change an existing value (this doesnt happen at all if the GPU had these values since last reboot), it enters in this "reduced performance mode" and thus hashrate is lower than you should. Is there any chance that the voltage/clock options are enforced first of all, so this behaviour doesnt happen? Because every time I have a OCL thead/hang, I have to do 2 reboots, the first one automatic from the miner and the second one to have the clock/volt dialed in a first time so the next reboot the GPUs dont enter in this "lower performance mode". i know i could just bios flash all the OC/UV/fan settings, but im still at finding the best OC/UV settings for each gpu, and I would like to nail them via driver changes rather than BIOS modding yet.

All in all very satisfied with this new version.

Thanks in advance!