Yeah, now I just gotta get it to go fast Well, and get it working with cgminer. Since I'm driving the parts with a micro controller I need to write a new cgminer driver to drive that…

Before going faster I need to get some cooling in there though. I noticed on your bring up page you mention keeping the temps below 50C. Is this a hard limit?

I was just doing a longer test where I kept all four chips completely busy with your test vector and verified that the correct nonces came out. Running at 240Mhz a couple of the chips did get a little hot. The two running at 0.8V stayed well below 50C but the two at 0.84V got up to about 61C.

No errors during the 12 minutes I ran the test. Theoretical hash rate is 32 * 240Mhz = 7.68GH/s and actual turned out to be 7.62GH/s.

But, how bad is it to run the chips that hot? Is that a might get bad results now and then or a might damage chips?

Thanks!