News for hackers. Stuff that matters.
Re:Using the most unhinged AVX-512 instruction to make faste... by fuhsnn on Sunday January 26, 2025 @04:18PM (#42834228)
Do they cover anything Sapphire Rapids Xeon's don't? I thought they share the same arch (Golden Cove).
Re:Using the most unhinged AVX-512 instruction to make faste... by suzumer on Sunday January 26, 2025 @04:31PM (#42834355)
According to this [1] wikipedia article, the only feature Sapphire Rapids doesn't support is VP2INTERSECT.

[1]:https://en.wikipedia.org/wiki/Advanced_Vector_Extensions

Re:Using the most unhinged AVX-512 instruction to make faste... by nextaccountic on Sunday January 26, 2025 @04:52PM (#42834534)
Re:Using the most unhinged AVX-512 instruction to make faste... by hogwarts2025 on Sunday January 26, 2025 @05:24PM (#42834838)
Or Zen 5. :-p
Re:Using the most unhinged AVX-512 instruction to make faste... by janwas on Monday January 27, 2025 @04:12AM (#42838956)
Note that the article mentions using both outputs of the instruction, whereas the emulation is only able to compute one output efficiently.
Re:Using the most unhinged AVX-512 instruction to make faste... by iamnotagenius on Sunday January 26, 2025 @04:29PM (#42834332)
Yes, you are right; I meant "consumer grade cpu".