Whether you’re
using a Mac or a PC, it’s important to maximize
your deep learning performance. I’m David Shaw, and in
this episode of AI News we’ll look at Apple machine
learning frameworks or Core ML, and metal performance shaders,
or MPS, on Intel processor graphics. [MUSIC PLAYING] Core ML, available
on Apple devices, is the main framework for
accelerating domain-specific ML inference capabilities such
as image analysis, object detection, and natural
language processing. With Core ML, it takes advantage
of Intel processors and Intel processor graphics to build and
run machine learning workloads right on your device. This removes the dependency on
network connectivity, security, and even privacy concerns. Core ML is built on top
of low-level frameworks such as MPS, the Intel
processor graphics, and basic neural network
subroutines Intel processors. They’re highly tuned and
optimized for Intel hardware. MPS is the main building
block for Core ML to run machine learning
workloads on GPUs. As an application developer, you
can write your own application to use the MPS API directly to
target underlying GPU devices. Check out this article
to see the benefits of using Core ML and MPS
API on Mac OS platforms. You’ll learn how to
take full advantage of the underlying Intel
processor graphics architecture. In addition, read the
performance case study. It shows the techniques used
in achieving high hardware efficiency, all using
the highly optimized MPS primitives for Intel
processor graphics on Apple Mac OS platforms. See the links provided,
and I’ll see you next week. [MUSIC PLAYING]

Apple* Machine Learning on Intel® Processor Graphics | AI News | Intel Software
Tagged on:                                                                                                                                 

One thought on “Apple* Machine Learning on Intel® Processor Graphics | AI News | Intel Software

  • December 6, 2018 at 3:01 pm

    Great …thanks for your information


Leave a Reply

Your email address will not be published. Required fields are marked *