The old Cyrix stuff. Last one of those I used was National Semiconductor's Geode CPU on a PC/104 board. It seemed to work fairly well, there's still a metric shitton of the boards I chose out there in the now defunct company's product.
I ran one of those for years as my "mini server" until the SBC revolution started with the Pi.
The old Cyrix stuff. Last one of those I used was National Semiconductor's Geode CPU on a PC/104 board.
Lol! That was one of the parts I supported! My test program ran on that same device. They had a 640 pin grid array ceramic package. They weren't as fast as the M2s but were a complete system on a chip.
Nice. I'll say they've been nothing but reliable. These things go into places You Do Not Want To Be, and go "Yep, I'm good with this."
The core clock speeds ranged from , I worked with every version over a 7 year period 1997-2004. They weren't the fastest processors around but it was a complete system in one neat package. The wafer sort test solution execution time on an was around 2.1s and the compiled test solutions were 2.2g to 2.5g In size, mostly functional test patterns. The wafer sort probe card was a sight to behold. IIRC it was around 32 layers with a lot of ground planes. Final test was also on the LTX Delta STE and test time was slighty faster, around 2.0s. Each 0.35u process 8" wafer produced around 32 testable die. These were big chips, at least they were the biggest I ever worked with.
(post is archived)