Quantcast
Channel: Raspberry Pi Forums
Viewing all articles
Browse latest Browse all 4980

Compute Module • Re: What is performance of the VideoCore VII GPU in the CM5?

$
0
0


How do you get to 16 FLOPs per core per clock?
I assume each core has a NEON SIMD that can do 4 floating point multiply plus accumulate operations. That is 8 FLOPs per core per clock.
On the A76/Neoverse N1 you have 2 128bit NEON SIMDS. It has roughly the same theoretical FP throughtput as Sandy/Ivy Bridge. On the X series cores, starting with X1 you have 4
Can you elaborate on this? I can not find any information that the A76 cores on the BCM2712 chips have 2 NEON SIMDs.
I searched on DuckDuckGo and
The ARM Cortex-A76 features two SIMD NEON execution pipelines. These pipelines are designed to enhance performance for media processing and machine learning tasks.
While not entirely reliability, Duck Assist seems more likely to be correct than CPU Monkey.

Statistics: Posted by ejolson — Sat Dec 21, 2024 5:05 pm



Viewing all articles
Browse latest Browse all 4980

Trending Articles