By: sylt (no.delete@this.thanks.com), November 17, 2010 2:26 pm
Room: Moderated Discussions
someone (someone@somewhere.com) on 11/17/10 wrote:
---------------------------
>
>It burns a lot of power using hundreds of thousands of
>logic transistors re-discovering something about a scrap
>of code every single loop iteration or subroutine call for
>every execution of all copies of program that one compiler
>need only to discern once. Everything can't be discovered
>at compile time but it is stupid to ignore that which can
>be and then passed on using a suitable equipped ISA.
The question is if you need to discover and handle the "hard" OoO cases in hardware anyway to get competitive performance, won't you get most/all of the simple ones for free? Then there may be very little benefit of from having the extra compiler generated information?
---------------------------
>
>It burns a lot of power using hundreds of thousands of
>logic transistors re-discovering something about a scrap
>of code every single loop iteration or subroutine call for
>every execution of all copies of program that one compiler
>need only to discern once. Everything can't be discovered
>at compile time but it is stupid to ignore that which can
>be and then passed on using a suitable equipped ISA.
The question is if you need to discover and handle the "hard" OoO cases in hardware anyway to get competitive performance, won't you get most/all of the simple ones for free? Then there may be very little benefit of from having the extra compiler generated information?