Future cpus have less performance potential?

[stlck0]

Hi all,

As the cpu manufacturing process (130nm->90nm->65-nm->45nm->32nm->22nm) move on, would the benefits (lower power usage, lower cost) become less and less? So the future cpus have less performance potential?

 

[1Tanker]

Hi all,

As the cpu manufacturing process (130nm->90nm->65-nm->45nm->32nm->22nm) move on, would the benefits (lower power usage, lower cost) become less and less? So the future cpus have less performance potential?

IMO only, i would have to think that would be the case. As in most things, there's the "law of diminishing returns." i.e A car that has 1000 h.p will not be twice as fast as a car with 500 h.p. One thing with CPU's that could keep performance increasing is new instruction sets(i.e- SSE3, SSE4, etc.) as these aren't dependant on node size. As i said, this is only my opinion, and i'm likely wrong, and a few guys on here may have greater insights into this good question.

 

[djkrypplephite]

just because they're smaller, doesn't mean they'll perform worse or better, it's a matter of architecture. then we'll have 16nm, 11nm, 8nm, 6/5nm, 4nm, 2/3nm, 2nm, 1nm, and then we'll just have "really freakin small"

 

[Heyyou27]

Hi all,

As the cpu manufacturing process (130nm->90nm->65-nm->45nm->32nm->22nm) move on, would the benefits (lower power usage, lower cost) become less and less? So the future cpus have less performance potential?

It doesn't work that way.