Nice, I was unaware of those
Nice, I was unaware of those pages... I like the thermal distribution layout, the thermal dissipation piece could maybe use a bit more refinement; Im not a big fan of the peltier but if you were to go down that road you may want to use some copper standoffs & thermal grease to give thermal continuity from in front of the peltier to behind it?
As I see the cost delta between the beta & gamma is fairly pronounced, from what Im reading theres at least 3 FPGA's on the gamma?
Not being hugely familiar with FPGA's, could some of this workflow be pushed off to something GPGPU'ish? I ask since nvidia's newest flavor of arm (X1) has 256 gpu cores, is prebaked into its own embedded board, and there are several people making carrier boards (E.G. auvidea, connecttech, etc.) that are capable of ingesting video, that may help keep the price & mass lower?
