News for hackers. Stuff that matters.
Re:Contrasting Intel AMX and Apple AMX by smoldesu on Monday September 05, 2022 @05:25PM (#32729674)
Hasn't Intel been shipping scalar instructions for more than a decade?
Re:Contrasting Intel AMX and Apple AMX by colejohnson66 on Monday September 05, 2022 @07:49PM (#32730842)
You’re probably thinking of AVX, Advanced Vector Extensions. That’s a family of vector extensions that has been out for just over a decade (2011). This is about AMX, Advanced Matrix Extensions. AMX is, as the name implies, built around 2D INT8/BF16 matrices (not “1D” FP16/32/64 vectors). It’s not on silicon available to the public.
Re:Contrasting Intel AMX and Apple AMX by xattt on Tuesday September 06, 2022 @02:11AM (#32733530)
Also not to be be confused with VMX (AltiVec/Velocity Engine), which predates AVX by another decade or so.

To keep the record straight, we have: 1. AMX, 2. AVX, and 3. VMX.

Re:Contrasting Intel AMX and Apple AMX by colejohnson66 on Tuesday September 06, 2022 @10:53AM (#32737883)
There’s also Intel VMX, the Virtual Machine Extensions ;) Sometimes it’s referred to as VT-x.
Re:Contrasting Intel AMX and Apple AMX by timdorr on Monday September 05, 2022 @05:51PM (#32729915)
Intel's AMX specifically hasn't yet shipped, but will be coming with their new Sapphire Rapids Xeon chips this year.
Re:Contrasting Intel AMX and Apple AMX by mhh__ on Monday September 05, 2022 @08:23PM (#32731074)
*Next year (potentially).

Sapphire rapids might well end up only shipping in Aurora and then being replaced immediately by it's successor.