starchaser137 Posted April 23, 2021 Share Posted April 23, 2021 I have been learning some things about GPUs lately and I have learnt that the basic difference between a CPU and a GPU is that a GPU has MUCH more cores than a CPU, but those cores are not as flexible as those of a CPU, and all of them work in parallel. So what does Apple mean when they say that their M1 SoC has a "8-core GPU" ? Wouldn't that be too underpowered? Then how does it perform so well? Or is it ( as I am guessing ) that they don't literally mean an octa-core GPU, and mean that a "8-block GPU" or something? Link to comment Share on other sites More sharing options...
MigL Posted April 30, 2021 Share Posted April 30, 2021 Each GPU core contains 128 Arithmetic Logic Units, ALUs, for a total of 1024 ALUs. Fabbed by TSMC in a 5 nm process, they have 4 hi-performance CPU cores, and 4 lo-performance CPU cores, for a ( claimed ) world's best CPU performance per watt. 1 Link to comment Share on other sites More sharing options...
Doctor Derp Posted June 23, 2021 Share Posted June 23, 2021 (edited) CPUs are optimized for a blend of integer and floating point operations. This is the best format for generating an end user experience and abstracts like multi tasking through integers. While including added precision for math based ops with floating point. GPUs are optimized exclusively for floating point ops. Which are useful for plotting 2D or 3D generated environments with added mathematical precision. You got it right. CPU and GPU cores are not backwards compatible. The architecture is different enough that 8 GPU cores does not translate to the same meaning as 8 CPU cores. Edited June 23, 2021 by Doctor Derp Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now