The 8008 was the (slower) single-chip implementation of the Datapoint 2200 terminal's processor. The DP 2200 was a serial terminal capable of running code on its own as a standalone machine, conceived in 1969 (originally as a drop-in replacement for the IBM 129 key punch), announced in 1970, and eventually introduced in 1971. Datapoint had commissioned the chip, but rejected the result, as it was slower than the discrete logic design, with rights for the chip-implementation remaining at Intel. This remarkable terminal has been with us ever since.
One said that computer hardware should have hardware support for high level languages. That school led to the VAX, iAPX432, and 80X86. The other school said that computer hardware should be simple so that it could go faster, and that the compiler should be smart enough to map high level languages to simple hardware.
Intel was firmly in the first camp. It flirted with the second repeatedly: see the late '80s i860/i960. Itanium was the biggest bet placed: it was a RISC-like architecture with the added complexity of multiple dispatch encoded in it's Very Long Instruction Word architecture. The compiler could rearrange instruction order to keep hardware busy, AND (to some extent) instruction contents.
Itanium seemed like a reasonable bet, and Intel (and HP) had the clout to convince others. It was a factor in the decline of the Alpha and MIPS server market, and even Sun hedged its bets by porting Solaris.
The MC68000 ISA was meant to run C code.
> The iAPX 432 failure could best be summed up by a meeting of Intel marketroids and Tandem engineers who wanted to use the chip in their next generation machine, after the slides, the senior engineer asked: “How long does it take to execute a procedure call?”
> The presenter looked it up. “Two hundred and fifty microseconds.”
> Tom immediately walked out, followed by the majority of the Tandem software department. The presenter was poleaxed. “What did I say?”
Found in comments to https://www.youtube.com/watch?v=FvmTSpJU-Xc , but I saw it somewhere else, too, I just don't recall where. Doesn't anyone know?
IHVs had little interest in I2O as it would reduce Intel's costs for swapping out vendors and there was no demonstrable performance improvement. The latter was at least in part because the I2O infrastructure was immature in comparison to the IHV drivers. It eventually formed the model for I/O over Infiniband (where it did make sense).
[0] It's not clear if I2O was the original application for the 960 or it was a pivot for an otherwise homeless processor.
For some reason, the BiiN project was terminated in 1988 and Siemens was not interested any more in it.
On the other hand Intel decided to not scrap the results of that project and they introduced the 80960 series based on the architecture formerly known as BiiN.
The commercial name 80960 was derived from their previous 8096 series of 16-bit microcontrollers, so 80960 was initially presented as higher-performance 32-bit replacement for the 8096 series, which was used in various embedded computers.
One interesting feature of BiiN was that it was the first monolithic CPU with an atomic fetch-and-add instruction (first used in 1981 in the NYU Ultracomputer project).
The 80960 inherited the atomic fetch-and-add from BiiN and then Intel added it to 80486, under the XADD mnemonic, together with the atomic compare-and-swap taken from IBM 370 and Motorola 68020 (CMPXCHG).
The applications for which 80960 was best known, like I2O and laser printers, happened significantly later than its initial introduction.
You mean it was biinned.
I developed an i960RP design back in the late '90s for an MPEG-2 encoder PCI card. The encoder chips were also PCI, so the PCI bridge on the i960RP made for a nice design where all the PCI stuff was handled in one fell swoop.
https://en.wikipedia.org/wiki/Intel_i960
The I2O project wasn't until the mid 1990's
In 1984 it was just the start of a research project named BiiN, of Intel and Siemens, which lasted 4 years, until 1988, when Siemens dropped out.
The series name, 80960, did not appear before the launch from 1988.
The reason it never went mainstream is easy enough to explain: expensive chips and (for the hypercube) a very different programming model.
I also believe that MMX was basically lifted from the i860, although again, i'm not sure.
It's funny, I tend to think of Intel as the boring chip company, due to the success and legacy that we like to moan about from ia-32, but they have actually made a fair few plays to try and move things on from there, fighting against the market which just wanted ia-32 compatible but faster processors.
Intel, for better or worse, are a victim of their own success. On the plus side, their success gave them lots of money to throw at the problem of making faster x86 CPUs. It seems, though, that Intel is gradually running out of luck, with AMD and now Apple introducing strong competitors, and Intel's advantage in fabrication eroding. So CPU-/ISA-wise, things could get very interesting in the foreseeable future.
I managed to miss the magic that is the 88k thankfully. Who actually used it? Linotype? Tek workstations maybe?
Applied made high-end in-circuit emulators. My team worked on software for execution trace disassemblers. We had wide and deep memories and could record something like 16,384 bus cycles in emulator (trace) memory.
Once we had some bus cycles to analyze, we’d sort out what the processor had done when it ran (an execution trace) and show it. It was a great tool for answering questions like “How did I get here?” when your embedded code jumped off into the weeds.
The i960CA was relatively simple to work with. The i960MX was a lot more complicated, as I recall.
I remember talking to Vic Poor when I used to work with him and calculators didn't seem to figure into the whole deal. It's probably worth a quick read through Datapoint history.
They had made the 8008 and 8080, but those were awkward and ungainly chips meant to power calculators, of all things. iAPX 432 was a clean sheet 32 bit design designed for high level languages and real computers, kind of like mini Lisp machines, complete with native machine support for garbage collection. But it was taking a while to get out the door, and performance wasn't so great, so they hacked together a quick upgrade to the 8080... the 8088, which was used in the original IBM PC.
History takes over from there.