Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by ajross on Monday November 16, 2020 @03:19PM (#25116510)

I love these die shot RE walkthroughs. This is such a weird choice though; almost no one used this chip for production hardware. The WD 17xx series was the king of floppy controllers in the 70's and early 80's (well, if you ignore Woz's masterpiece, which was discrete logic).

Seems like if you're going to reverse an obscure chip, something with a more exotic application would have been more fun?

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by scarybeast on Monday November 16, 2020 @03:31PM (#25116632)

We definitely wanted fun! I guess there are two ways to slice and dice it: 1) fun because the application is exotic / iconic, or 2) fun because the chip is exotic. This is definitely a case of the latter: an exotic chip with interesting history.

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by bane on Tuesday November 17, 2020 @09:24AM (#25124020)

The next goal should be trying to write a demo for it and enter it into Demosplash 2021.

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by rbanffy on Monday November 16, 2020 @05:02PM (#25117579)

> Woz's masterpiece, which was discrete logic

It had the downside of requiring precise timing. This is an issue if you wanted a faster (or even a slightly improved) 6502 in the machine.

Oh well...

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by ajross on Tuesday November 17, 2020 @08:40AM (#25123572)

None of the Apple II's competitors exploited that, though. EVERYTHING in that world of software was built around tight hardware assumptions. Atari, VIC, Commodore, none of them ever shipped a faster CPU. That needed to wait for later architectures.

No, the Disk II was absolutely a masterpiece. Go back to the linked article, think about everything this very complicated chip did, and recognize that Woz made a card that did exactly the same thing out of 6 chips you could buy at Radio Shack and two tiny ROMs.

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by rbanffy on Tuesday November 17, 2020 @07:00PM (#25131534)

I am familiar with the design. At the time, the only computers that were less timing dependent were the S-100 and CP/M boxes where you could swap out a CPU for a faster one if memory speed permitted.

I don't consider it a mistake - nobody at the time was realizing that we all would, eventually, move our software to faster computers every couple years. Woz is brilliant, but he couldn't have foreseen it and, as you point out, almost nobody did.

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by pvg on Monday November 16, 2020 @10:40PM (#25120051)

This made me realize that Zip chips were a lot more complicated than kid-me thought:

https://en.wikipedia.org/wiki/Apple_II_accelerators#Zip_Chip...

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by rbanffy on Tuesday November 17, 2020 @05:48AM (#25122219)

Those things were amazing engineering. Someone's probably making something like it for the retrocomputing crowd.

With current tech it should be trivial to keep all of the RAM in the module and just asynchronously push some writes out (video and IO) at Apple II speeds.

[ Parent | on HN ]

Re:Reverse engineering a forgotten 1970s Intel dual core bea... by nradov on Tuesday November 17, 2020 @02:14AM (#25121012)

Back then no one really expected microcomputer hardware architectures to have a long lifespan or allow for incremental CPU upgrades every year.

[ Parent | on HN ]