tag:blogger.com,1999:blog-6264947694886887540.post5240240538862637372..comments2024-03-18T15:18:27.236-07:00Comments on Ken Shirriff's blog: Counting bits in hardware: reverse engineering the silicon in the ARM1 processorKen Shirriffhttp://www.blogger.com/profile/08097301407311055124noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-6264947694886887540.post-31269764923536755442021-01-09T23:47:11.719-08:002021-01-09T23:47:11.719-08:00Great article!
You noted that the ARM1 layout is m...Great article!<br />You noted that the ARM1 layout is much more regular than say the 6502 or Z-80. I wonder how much of that is because they didn't care as much about area, and how much it is because of the constraints of CMOS design compared to NMOS or PMOS?<br /><br />I have done some hobby CMOS layout, just to try to get a feel for it (using Magic VLSI, which is free but a bit archaic), and one thing that has struck me is how much you're constrained by the design rule that requires a minimum distance between NMOS and PMOS transistors (i e between N and P diffusion). Maybe it just wasn't possible to create the same kind of jigsaw puzzle layout in CMOS? It would be interesting to compare e g the layout of the 6502 and the 65c02 to see if they managed to squeeze out some of the air in the 65c02 CMOS design as well.<br />Toivo Henningssonhttps://www.blogger.com/profile/05599157383718975426noreply@blogger.com