Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Intel / AMD have a 4-wide decoder, of which can be a 6-wide decoder if executing out of the uOp cache.

Apple just went with an 8-wide decoder, surprising a bunch of people. There's not much difference between 4-wide and 8-wide, aside from Apple deciding that such a wide single-core unit was worthwhile.



> Intel / AMD have a 4-wide decoder, of which can be a 6-wide decoder if executing out of the uOp cache.

uOps are already decoded by definition


That depends on how many execution units it has to play with, although I guess the bottleneck at that point could end up being the length of a predictable flow.

Not that Apple will tell us how many units it has


https://www.anandtech.com/show/16226/apple-silicon-m1-a14-de...

It's actually well known how wide the M1 is.


It's still guesswork how many ports it actually has.

You can measure the ROB pretty accurately but the execution units have to be used to be measured.


You can only measure accurately if you actually know what the details look like.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: