Note: In a previous post today, I offhandedly said "I can actually make my own CPUs, not good ones". On rereading, I realize this may overstate my ability, as it implies I can make mediocre CPUs. In fact I can make fricking terrible CPUs, and only in HDLs. The only time I have actually created a circuit which executes a stored program was in ABEL, an HDL so antiquated we called the device it ran on a "CPLD".

I aspire to someday make a mediocre softcore CPU. That would make me very, very excited

However, I *can* make fricking terrible CPUs. That's not bragging. It's just a factual statement about a thing I can do poorly
Also if someday I *do* successfully participate in a tiny tapeout I assure you I *will* brag about it. I will be insufferable
@mcc
I can't even do that, so I salute you!
@TerryHancock I hope someday the Amaranth tutorial device thing comes to fruition so everybody can see how easy this stuff really is. If you can write Haskell you can write SystemVerilog

@mcc
That's good news, actually! I thought Intel has a monopoly on that talent!

Ha ha, only serious.

@brouhaha i bet i could divide a number if i really needed to
@mcc this post is giving me flashbacks to the one time i had to use VHDL 😡
@d6 it's nicer than verilog imo. vhdl
@mcc not in my hands! πŸ˜…

@mcc i happen to have come across this 3D-printed vacuum-fluidic / pneumatic thing today. The state can't be read by the thing itself so it isn't capable of being made into a CPU (without another innovation) but it feels like it needs to be set here

https://youtu.be/E1BLGpE5zH0

Air Powered Segment Display: 3D Printed Microfluidic RAM?

YouTube
@mcc ABEL was the first HDL I used, but it was for an FPGA though (Xilinx). My next project was a mix of FPGA (SRAM) and CPLD (OTP). Ah, the joys of last millennium digital design.