You can tell the software bias in the HN community because almost every book recommended is inappropriate if you want to understand the image of the chip shown

You can tell the software bias in the HN community because almost every book recommended is inappropriate if you want to understand the image of the chip shown in the link. Even looking at "Art of Electronics", it's too high level, and my personal opinion is that it's actually a terrible text book (though a good reference book) -- I sometimes wonder if people who recommend it have actually read it themselves (much like people who recommend Knuth). Instead, what you will want to look at is the topic of VLSI, if are in fact interested in how to read the chip image (though again, you'd need to be able to "zoom" in). I mentioned earlier that often the pre-requisites to this are basically things like circuit analysis, microelectronic circuits, and semiconductor physics, but the reality is that if you want to just hobby with VLSI then you only need a very small amount of this background knowledge, because the software tools for VLSI abstract many of these concepts out into things like "METAL1" and "METAL2", and the tools knows about the underlying physics rules, to where you don't actually have to worry about deriving the spacing rules for metals and such (that is, given an x nm chip process the minimal spacing between poly should be y nm). Indeed, in day to day usage, I think about custom chip design more as constructing "LEGO" blocks + water pipes rather than their underlying physical properties, and at this point I know them better by their highlight color and honestly have completely forgotten what the metals themselves are actually are. Now unfortunately, unlike Computer Science, it's hard to find good books that can serve as a textbook rather than a reference tome. I am not sure why this is the case. It's also really hard to find a good book that is positioned correctly. Most are either way too extreme on the Physics side, or too quickly switch to VHDL or high-level chip languages, assuming that the software will generate the chip "image" for you. Another problem is that VLSI chip design is hugely proprietary; the open source tools are decades away from incredibly expensive tools like Mentor, Cadence, Synopsys, etc -- and you will get a nice throwback to Motif and X11 since usability does not seem to be a strong suit of these tools! So, unfortunately my recommendation is to simply "search for VLSI" or maybe "VHDL" if you want to start more on the software side as a starting keyword for books and pick something that is the appropriate mix of what you are looking for.