GPU Driver Crashing for one of the Cards, can I mine at half-rate

Hi,

Rig (3x R9 390) mining for about 7 months. Out of the blue, Ethminer (Win 7) starts crashing within mins of miner running. No OC, GPU COre 1000MHz, GPU Mem 1500MHz. Driver ver 15.12. Using 1000W Gold-rated PSU. Generally running a little high temp (due to ambient temp), GPU @ 68-76C, VRM Temp 70-95C.

Tried a few tests, figured that one of the GPU cards might already be faulty. My guess: Tested with only the 'faulty' card in place, crash occurs when temperature/load starts rising. Interestingly, Underclocking the single GPU to 500MHz, test with FurMark -> No Crash.

Q: I have now shifted to Claymore. Is it possible to run one of the cards at 500MHz, while the rest are operating @1000MHz?
Using Radeon Global Overdrive to tweak the faulty card to 500MHz, maintaining the other cards at 1000MHz. GPU Clock for 'faulty' card ramps up to 1000MHz once EthDcrMiner64.exe starts. Could be linked the the setting below...

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

Thank you!
Kay