better predictors, better at discarding results without needing to flush everything.
-
-
Replying to @stephentyrone @RichFelker and
i.e. finer grained tracking of everything. Power/area/complexity cost is worth it.
2 replies 0 retweets 1 like -
Replying to @stephentyrone @rygorous and
Why can't we just use that area for 100x as many super-dumb cores?
5 replies 0 retweets 0 likes -
Replying to @RichFelker @stephentyrone and
you can, if you are willing to sacrifice cache coherency: https://www.extremetech.com/extreme/230458-meet-the-new-worlds-fastest-supercomputer-chinas-taihulight …
1 reply 0 retweets 0 likes -
Replying to @SebAaltonen @stephentyrone and
You only need coherency at atomics, & that can be done via global flush on any atomic
1 reply 0 retweets 0 likes -
Replying to @RichFelker @SebAaltonen and
Performance would be pretty terrible though, I think. :-P
1 reply 0 retweets 0 likes -
Replying to @RichFelker @SebAaltonen and
yeah. one can make synchronization and inter-core communication be special cases.
1 reply 0 retweets 1 like -
Replying to @cr88192 @RichFelker and
you could also have non-coherent pages (default) and coherent pages for communication.
1 reply 0 retweets 0 likes -
Replying to @SebAaltonen @cr88192 and
That needs a new programming model. HW imposing that = destined to fail miserably.
2 replies 0 retweets 0 likes -
Replying to @RichFelker @SebAaltonen and
my idle thinking here was probably a "make this page sync" opcode or similar.
2 replies 0 retweets 0 likes
That's a radically different programming model. Locks don't work to synchronize.
-
-
Replying to @RichFelker @SebAaltonen and
yeah, it would be more copy-in / copy-out, or explicit push/pull/sync...
0 replies 0 retweets 0 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.