[repelsteeltje]

It's easy to forget how much innovation divergence was happening in the early 70s. Up till the late nineties we speculated that ISAs other than x86 (spec. RISC designs) would win in the end.

Imagine these 8 bitters were mostly hardwired, with less than a million transistors.

[pjc50]

> ISAs other than x86 (spec. RISC designs) would win in the end

They're correct: ARM won the mobile space, won Apple, and is very gradually seeping into wider availability for other operating systems.

Never mind the majority of raw FLOPS these days are almost certainly going through GPU architectures.

[ido]

Way less than a million! I believe usually in the order of thousands.

[vidarh]

Indeed. The 6502 had 3510 enhancement transistors and 1018 depletion transistors for a total of 4528...

[rzzzt]

With the layout fitting entirely on a large sheet of paper: https://archive.archaeology.org/1107/features/mos_technology...

[iberator]

Amazing article!

[adrian_b]

The 8-bit monolithic CPUs of the seventies all had a few thousands transistors.

The first generation of true 16-bit CPUs, i.e. Intel 8086, Motorola MC68000 and Zilog Z8000, had almost an order of magnitude more transistors, i.e. in the range of 15000 transistors to 50000 transistors.

The first true 32-bit CPUs, like the National 32000 series, Motorola MC68020 and Intel 80386, had a few hundred thousand transistors.

By the end of the eighties, the second generation of 32-bit CPUs reached 1 million transistors.

[childintime]

Gemini assures me one can make a 32b CPU in around 9000 transistors or 300 LUTs, it'll run at 30 cycle accurate MIPS from quad SPI memory.

Same size as a 6809, low-end 486 performance. Who cares ;)

[watersb]

The Motorola 68000, a great CPU with 32-bit operands, was initially implemented with 68,000 transistors.

The model number was decided long before the transistor-level design was finalized.

[jecel]

I think there were actually 68000 transistor positions. In the ROMs and PLAs not every potential transistor is populated but the missing ones were counted as well. But the number of actual transistors is only slightly smaller so it doesn't really matter.

[adrian_b]

I do not remember the exact number of transistors in MC 68000, but I think that it was less than 40000, so not just slightly smaller. In any case, it had more than twice as many transistors as Zilog Z8000, which was super-optimized for size (a very bad decision of Zilog, which lead to a too long time-to-market and to many initial bugs), and slightly more than 4/3 times as many transistors as Intel 8086.

The 68000 transistors number claimed by the Motorola marketing was close to what you get by dividing the die area to the area of one transistor, so it did not correspond to actual transistor positions.

The MC68000 die had large areas occupied with microprogram ROMs, and there as you say only a part of the array of transistors are active, depending on the stored bits. Nonetheless, a significant part of the die was occupied with random logic, where all the physical transistors are used and a part of the area does not have any transistors.

[shrubble]

It depends on how you want to treat the internal (not seen by user or compiler) decoding that both AMD and Intel are doing; some would argue that these CPUs are RISC chips that decode x86 and x64 instructions into their own RISC implementation.

[userbinator]

A 486 already had over a million transistors. These are in the thousands.