We wanted to send out samples this week but we ran out of chips too fast. It is crucial to estimate the yield of the chip production [number of bad chips] before we assemble mining boards with more than 1 chip. There is a huge difference between 1% broken chips and 3% broken chips if for example 16 chips are placed on 1 board. To assess the error rate we had to treat 1000 chips using the same procedure. This left only a very small amount of chips for independent testers. These were selected by bitfury in advance and will receive few chips in coming days (we start sending tomorrow).

We will have more chips in 2 weeks though and we are ready to sent samples after that. However we can not send the whole production lot as samples to all requests we get. We have to introduce restrictions / barriers for requests otherwise we spend our time packaging small packages and send them around the world. To cover our handling expenses and shipping costs we will provide 5 chips in exchange for 1 BTC, AND we will not ship more than 5 chips to 1 address. Users who present interesting results will receive more chips as part of a separate collaboration deal. We are mostly interested in:
1. getting high hash rates from the chips
2. producing low cost mining devices
3. operating the chips in daisy chain (string) to cut on power regulator costs. [this will be explained later by bitfury]

I will write more about the chip sample program in the next week.