What does the instruction set look like on that? Is there another example out of this type of chip?
https://drive.google.com/file/d/0Bx4hafXDDq2EMzRNcy1vSUxtcEk...
It's mainly moving memory around, matrix multiplication, convolution, and applying activation functions (sigmoid, tanh, relu, etc.). Very simple, high-level stuff. This has the handy side-effect of making timing very predictable, which makes the latency a lot more deterministic.