23 Nov
2006
23 Nov
'06
12:01 p.m.
On Thu, 23 Nov 2006, arcarlini at iee.org wrote:
I always assumed that the 486SX was a 486DX that
failed the FPU tests
but all the other bits worked. I guess there were some links blown so it
could be identified as such. If that was how things were done then it
would surely have helped bring costs down (by effectively upping the
yield of sellable parts).
Actually, the early 486SX dies were fully working 486DX parts. They just
had a different pinout and no bond wires were installed for the FPU. Intel
did ship later 486SX chips that were made from dies with defective FPUs.
-Toth