I’m sweating bullets as I write this, and I sprained my fingers during sparring, but I’m still going to bloody write this thing.
I’m almost done with “Understanding Digital Signal Processing”. I just have the last chapter to trudge through (the last chapter happens to be 200 pages of very niche DSP techniques, but I should be done by mid-July). I got through IIR filters, which involved brushing off my complex analysis knowledge to understand stability conditions. One very interesting thing from the book is quadrature signals; the idea is that you can represent a single complex number via a pair of wires, where the “i” multiplication happens via a 90 degree phase delay of one wire with respect to another. This is a very direct way of representing complex numbers in the real world, which I found neat!
I started reading though John Preskill’s Quantum Information notes. It’s slow going (only got through one and a half chapters), but the contents is very compelling. Two things that stuck out to me are that information erasure is associated with heat dissipation via entropy, and that the information of a quantum computer is encoded in the nonlocal correlations between the qubits.
I made progress on the camera. That jumbled mess is actually a live feed of the camera (I wasn’t checking the valid signal from the camera when transferring pixels to the BRAM buffer). There are a couple of problems with it still: namely the messed up colors and the fact that the image gets split into 4 pieces. Time to dial in on my testbench…
I picked up “Modern Compiler Implementation in ML” once again. This time, I tried taking some notes in parallel with making the toy compiler. I got stuck on setting up the ocaml lexer, but I’ll eventually break through.
Research is chugging along. I’m building a JAX based sandbox for PTA systems. There are some interesting technical challenges (double double precision, dealing with the functional style of JAX etc.) and it’s slow going, but I love it.