One of my colleagues wrote a first impressions blog post last week. It's from our company's perspective, but is a solid overview of the product and intended capabilities, from the POV of an AI developer or data scientist.
> There you’ll see the 10 Cortex-X925 (“performance”) cores listed with a peak clock rate of 4 GHz, along with the 10 Cortex-A725 (“efficiency”) cores listed with a peak clock rate of 2.8 GHz
> If you start Python and ask it how many CPU cores you have, it will count both kinds of cores and report 20
> Note that because of the speed difference between the cores, you will want to ensure there is some form of dynamic scheduling in your application that can load balance between the different core types.
Sounds like a new type of hell where I now not only need to manage the threads themselves, but also take into account what type of core they run on, and Python straight up report them as the same.
< The CPU memory is the same as the GPU memory and is much larger than any other discrete GPU available in a desktop. That means much larger datasets and bigger models can be run locally than would be possible otherwise.
Isin't this the same architecture that the Mx from Apple implements from a memory perspective?
> If you start Python and ask it how many CPU cores you have, it will count both kinds of cores and report 20
> Note that because of the speed difference between the cores, you will want to ensure there is some form of dynamic scheduling in your application that can load balance between the different core types.
Sounds like a new type of hell where I now not only need to manage the threads themselves, but also take into account what type of core they run on, and Python straight up report them as the same.