On 4/26/2012 10:45 PM, Gregg Levine wrote:
Hello! I recognize the processor. The NS32K was a processor built by the (late) National Semiconductor, that was loosely based on a design created by WE (Western Electric)/Lucent/AT&T at about the same time. But not the firm who wrote that port.
Can you provide a reference, because it's the first time I've ever heard anyone assert this? Last I checked, the National Semi folks seemed pretty clear the NS32000 was not-so-loosely based on the VAX. They were also developed contemporaneously, both being released in the late 70's (1978?), so "based on" doesn't work from the timeline.
Interestingly enough the firm, (The phone company) originally built that processor for the computers that spent their days running the switches that presented us with the means of calling our friends and families. In fact there was a whole series of them leading up to that design. Sadly all of these designs never went very far.
Dunno about that. They made it 4 production generations (BELLMAC-32A, we32000, we32100 & we32200), had a modestly successful commercial run with the 3B series until all-the-world went RISC, and they're still running in telco switch gear around the world. In some ways, that was better than the NS32000, which only made it 3-ish generations before NS gave up and re-purposed it as an embedded processor, and I'd bet a beer that more 3B2s were sold than all the NS32xxx based machines ever built.
And I am not sure what AT&T did to convince UNIX to work under those means and methods. I think there was a discussion on the TUHS list a while back which described what did happen, and then pointed to the Wiki pages on it.
Not sure what was so hard about it (though I haven't looked at my WE32000 databooks in probably 20 years)...rather conventional 32-bit processor, decent MMU, designed to support C language development. Have to track down that article.
KJ