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

“The next-generation cores feature improved branch prediction, with wider decode and execution engines for the performance cores, and a deeper execution engine for the efficiency cores. And both types of cores also feature enhanced, next-generation ML accelerators.”


I wonder to what degree this also implies, "More opportunities for speculative execution attacks".


I thought the same thing when I read it, but considering the attacks were known when doing this improvement, I hope that it was under consideration during the design.


Fat decode pipelines have historically been a major reason for their performance lead. I'm all for improvements in that area.


I wish there was more detail about the Neural Engine updates


Agreed. I think they just doubled the number of cores and called it a day, but who knows.




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

Search: