By: Troll? (trashbin67.delete@this.gmail.com), January 2, 2015 7:55 am
Room: Moderated Discussions
Linus Torvalds (torvalds.delete@this.linux-foundation.org) on December 8, 2014 1:34 pm wrote:
> Jouni Osmala (josmala.delete@this.cc.hut.fi) on December 8, 2014 1:10 pm wrote:
> >
> > I'm assuming that 90+% of programs already run fast enough and they don't matter for this.
> > Its all about asking question in what use current computers are too slow , and can you parallerize
> > that or are those cases already parallel. And I'm assuming you can parallerize atleast 10%
> > of those times where user waits CPU for long enough to actually notice it.
I won't disagree with the statement, but there is something to be said for parallelizing via specialized cores vice GP ones. We already have things like hardware MPEG decoders and DSPs on mobile devices, GPUs for graphics, and appliance crypto, and the price/perf/power ratios seem to come out well. Why not have a standard block of 20 or 50 or however many ASICs on-die with 4-6 GP-CPUs, and let the OS decide how to parcel out tasks? I don't know where FPGA technology stands, but I'm guessing it loses on power. I won't speculate on how much variety you would need, but individually they wouldn't have to be hugely complex. (except graphics)
> Jouni Osmala (josmala.delete@this.cc.hut.fi) on December 8, 2014 1:10 pm wrote:
> >
> > I'm assuming that 90+% of programs already run fast enough and they don't matter for this.
> > Its all about asking question in what use current computers are too slow , and can you parallerize
> > that or are those cases already parallel. And I'm assuming you can parallerize atleast 10%
> > of those times where user waits CPU for long enough to actually notice it.
I won't disagree with the statement, but there is something to be said for parallelizing via specialized cores vice GP ones. We already have things like hardware MPEG decoders and DSPs on mobile devices, GPUs for graphics, and appliance crypto, and the price/perf/power ratios seem to come out well. Why not have a standard block of 20 or 50 or however many ASICs on-die with 4-6 GP-CPUs, and let the OS decide how to parcel out tasks? I don't know where FPGA technology stands, but I'm guessing it loses on power. I won't speculate on how much variety you would need, but individually they wouldn't have to be hugely complex. (except graphics)