1) I'd always assumed this but things I've heard recently made me wonder. Are CPUs (just the 0.Xum cores .... not including the pin area) multi-layer like motherboard PCBs are? (I'm assuming that if they are then the # of layers basically depends on crosstalk and heat disipation?)
2) Why don't companies mix manufacturing processes. ie: Why doesn't Intel make a processor based on 0.18um tech and it's cache based on 0.13um tech (since the cache is simpler and that would save on space/heat)?
3) It seems there are 'standard' steppings for die size. 0.25 > 0.18 > 0.13 > 0.09um etc.... What motivates some companies to do half steps like 0.15um? Is it a technology/architectural choice or simply marketing/bragging rights?
4) Does Intel (or AMD or whoever) make any of their older chips on their newer processes? ie: Can I find a 0.18um P MMX? I know this seems silly but that would allow for some kewl embedded/shrunk devices.
Thorin
2) Why don't companies mix manufacturing processes. ie: Why doesn't Intel make a processor based on 0.18um tech and it's cache based on 0.13um tech (since the cache is simpler and that would save on space/heat)?
3) It seems there are 'standard' steppings for die size. 0.25 > 0.18 > 0.13 > 0.09um etc.... What motivates some companies to do half steps like 0.15um? Is it a technology/architectural choice or simply marketing/bragging rights?
4) Does Intel (or AMD or whoever) make any of their older chips on their newer processes? ie: Can I find a 0.18um P MMX? I know this seems silly but that would allow for some kewl embedded/shrunk devices.
Thorin