Quote from: akadamson on May 05, 2018, 02:14:48 PM

Patrike,

I'm still getting about 50% on automatically changing OC between algos when I'm profit mining.

Sometimes it does, sometimes it doesn't.

Here's how I basically have it setup.

I created a
1080Ti - Normal ( this is for most everything and is used as the stop OC )
1080Ti - x16r
1080ti - Phi

1070ti - Normal
1070ti - x16r
1070ti - Phi

Then I have these groups
am1r - normal
   - 1080ti - Normal
   - devices 0-5
am1r - Phi
   - 1080ti - Phi
   - devices 0-5

am2r - normal
   - 1070ti - Normal
   - devices 0-5
am2r - Phi
   - 1070ti - phi
   - devices 0-5

I had to do the groups because there is a non-gpu video driver and it would crash when the AB tried to update it

For a phi algo in the profite profile
I would set the am(1/2)r - phi as start and am(1/2)r - normal as the stop

it's a similar thing for x16r

But what is happening is that if it switches to Phi and sets that OC, then the next switch is to x16r, it never sets the am(1/2)r - x16r group and so the miner runs on the phi OC. it's true the other way as well.

And also about 50% of the time, it never switches back to normal on a stop.

I looked at the log and I see no reference to the start or the stop there, but occasionally I need a enumeration of the GPU setups/groups.

anyone else doing this and have ideas/suggestions, etc. I'm running the latest AB 4.5.0 with the updated AB remote so I could change this to card names instead of devices, but I suspect the same would occur.

UPDATE:
it appears that the first miner (am1r) gets set *mostly* right, but the second one (am2r) does not, could there need to be a programmable delay betweeen the OC command changes from miner a to miner b?

I think this needs to be investigated via the Remote Agent log file on the computers where you run the mining. It's possible to search for "Executing GPU clocking command" to see what commands Awesome Miner sends to MSI Afterburner.