Is there somewhere with more info on these sticks just as how to assess if they are heat damaged
I would imagine heat damage will cause one of 2 things:
1) reduced hash rate. These run very precisely (ive got 335 Mhash and 334.9 Mhash respectively), and heat damage to the asic chip would result in fewer (if any) hashes.
2) increased error rate: much more rejected shares or hardware errors.
I find with a small heatsink to add slight cooling bonus, my HW errors sits around 1.3%. Without any cooling, and the 2 devices horizontal above eachother and creep up to 1.5-2.3% error rate. this alone makes a little bit of forced cooling a bonus for hardware lifespan and ROI
thought i would throw some more info on top of my prior post. I have sold (at profit!) one of my two devices, which results in the lone unit having better airflow, no longer a tigh-squeezed device directly above, and a second small heatsink added. With this change, the device's hash rate has seemingly crept from 335.0 Mhash/s --> 335.4 Mhash/s and the error rate has dropped to less than 1%
some gain may come from CGMiner controlling half as many devices, but it would seem that lower temperatures are the primary cause. I'll try to throw in an update once there is a greater sample count if the HW errors settle to a constant %