Hacker News Reader

Comment by cjfd | Hacker News Reader

I considered trying to do a simple CPU design from logic gates too. But I ended up wondering about some of the performance characteristics. Maybe some people who are knowledgeable are reading this. What I am wondering about is the switching speed of logic gates as compared to the signal speed in the electric connections for a realistic CPU. I.e., how many logic gate lengths (assume logic gates to be square) does an electric signal travel in an electric connection in the time that is needed for a logic gate to invert its output. Another one that seems relevant is how much spacing electric connections need compared to the size of a logic gate.

View on HN

does this answer your question? https://monster6502.com/

"The MOnSter 6502 runs at about 1/20th the speed of the original, thanks to the much larger capacitance of the design. The maximum reliable clock rate is around 50 kHz. The primary limit to the clock speed is the gate capacitance of the MOSFETs that we are using, which is much larger than the capacitance of the MOSFETs on an original 6502 die."

"We are working towards a public launch, currently planned for Mid 2023"

Hmmm...

If you want to go a step further, here's one built from discrete transistors: https://www.megaprocessor.com/

even just building logic gates out of transistors is half the battle. for that, the referenced site also has https://www.megaprocessor.com/stepping-stones.html

This machine is now at the Computer History Museum in Cambridge. I saw it there last year. Highly recommended.

https://www.computinghistory.org.uk/

I'm just here to point out that your question is nonsensical. You're asking for propagation delay of the wires vs switching time of the gates, but you don't want the propagation delay to be measured by the time it takes for a signal to be propagated or the switching time of the not gate, so you are instead asking the total propagation delay to be normalized by the switching time of a not gate and somehow calculate that back into a wire length equivalent normalized by the longest dimension of a not gate.

None of these things make much sense.

Also as an approximate answer to your question: The simplest circuit without a flip flop aka an oscillator can easily reach hundreds of GHz using silicon germanium based semiconductor manufacturing

The performance characteristics won't be great, but it would still be a fun project. In fact one of the early 32-bit RISC processors, PA-RISC, was originally shipped as a set of four large boards covered in discrete 74-series logic. Which actually includes a large ALU chip (without which the design would probably span 8 boards instead...)

The pa-risc ts1, more here https://www.openpa.net/pa-risc_processor_pa-early.html and https://computermuseum.informatik.uni-stuttgart.de/dev_en/hp...

Reply on news.ycombinator.com