The D765 was a D765B, probably a shrunk version, hence the different die size.
Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
11–20 of 40 posts
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#12Earlier quoted context omitted.
If this in the context of the BBC Micro, Acorn had already used the 8271 for floppy support with previous “System” and “Atom” models[1], so it’s probably chosen through inertia from a previous design choice. As for the older systems, does anyone have a date for the WD 1771 (the more common alternative)? The data sheet shown on the Wikipedia page is dated April 1979, and if that’s the release date then it would have b…
Google Books shows the WD1771 available and in use in 1977 with more products using it in 1978. (I find Google Books very useful for this sort of archaeology. Change the time range to "1977-1980", search for "wd1771", and then sort by date. Of course this isn't thorough, but it gives a good overview for little effort.)
Off topic but I wish Google allowed doing the same for web results, to search their index the exact way it was some number of years ago so that it would only show results that had actually been indexed at that point in time.
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#13The floppy drive controller chip costing and being better than the main CPU reminds me of some of my arduino projects.
For example I was driving a GSM Modem (over serial) which contained a comparatively advanced ARM SoC driving it.. from an AT Mega 8-bit micro that I was programming. It was great educationally, but kindof hilarious :)
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#14Is there a tech available currently to decap and photo chips at home? For example reverse engineering old PAL chips?
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#15The decision to use this chip (vs the WD) in the BBC Micro made me wonder if there were other obvious-in-hindsight bad/weird/hackish choices in classic computer design? The PC-AT's use of the keyboard controller to control A20 and CPU reset comes to mind. Any others?
Several video game consoles from back in the day (the Magnavox Odyssey 2 and Entex Adventurevision) used the MCS-48 microcontroller mentioned in this write up instead of more conventional CPUs of the time like the 6502 or Z80, which seems a bit strange.
Having dabbled in both MCS-48 and 6502 assembly I can tell you it was probably not done to make software development easier. :) I suspect maybe the hardware was just easier to design around a microcontroller? ...but there could have been other reasons...
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#16A lot of special-purpose ICs are actually general-purpose processors with a mask ROM (or sometimes EEPROM, with interesting consequences), since writing the "firmware" for different functionality is easier than doing a whole "hard-coded" chip design --- the various USB-to-X adapters are one common example of this.
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#17A lot of special-purpose ICs are actually general-purpose processors with a mask ROM (or sometimes EEPROM, with interesting consequences), since writing the "firmware" for different functionality is easier than doing a whole "hard-coded" chip design --- the various USB-to-X adapters are one common example of this.
A large number of Realtek products are some form of MIPS core (or at least used to be).
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#18My guess is that, at some point, a manager at Acorn asked which disk controller would the engineers recommend and they, obviously joking, said "8271". When they realized what they had done, it was too late.
If this in the context of the BBC Micro, Acorn had already used the 8271 for floppy support with previous “System” and “Atom” models[1], so it’s probably chosen through inertia from a previous design choice. As for the older systems, does anyone have a date for the WD 1771 (the more common alternative)? The data sheet shown on the Wikipedia page is dated April 1979, and if that’s the release date then it would have b…
Very believable that there was institutional inertia for using the chip that they already understood well instead of a brand new chip just appearing on the market. The new chip might be better but you could lose a month or three of development time integrating it, assuming you don't run across an unexpected bug.
But at the same time it's no surprise that later models switched over to the cheaper and more technically capable chip.
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#19The floppy drive controller chip costing and being better than the main CPU reminds me of some of my arduino projects. For example I was driving a GSM Modem (over serial) which contained a comparatively advanced ARM SoC driving it.. from an AT Mega 8-bit micro that I was programming. It was great educationally, but kindof hilarious :)
The Commodore 64's disk drive was more powerful than the CPU, containing the same 6502 as well as a pair of 6522s. Amusingly some programs would upload mathematical calculations to the disk drives to speed up processing.
Re: Reverse engineering a forgotten 1970s Intel dual core beast: 8271, a new ISA
#20The floppy drive controller chip costing and being better than the main CPU reminds me of some of my arduino projects. For example I was driving a GSM Modem (over serial) which contained a comparatively advanced ARM SoC driving it.. from an AT Mega 8-bit micro that I was programming. It was great educationally, but kindof hilarious :)
The Commodore 64's disk drive was more powerful than the CPU, containing the same 6502 as well as a pair of 6522s. Amusingly some programs would upload mathematical calculations to the disk drives to speed up processing. https://www.lemon64.com/forum/viewtopic.php?t=27911
Same thing with early laser printers. The original Apple Laserwriter had a faster clocked 68000 compared to the original 128k/512k Mac