I’m David Shaw, and
this is AI News. Deep learning is used in
image and video processing, natural language processing,
personalized recommender systems, and
reinforcement learning. The types of workloads
and algorithms are rapidly expanding,
and Intel has been advancing both hardware
and software to accelerate deep learning workloads. The hard work has paid off. Intel has achieved leadership
performance of over 7,800 images per second on ResNet-50. This was done with the help of
the latest generation of Intel Xeon scalable processors. This outperforms images
per second on NVIDIA GPUs. This is a significant
milestone for customers who have Intel Xeon
scalable processors widely available in their
cloud and data centers. Read the article provided
in the links that shows this achievement,
and follow the steps to reproduce it yourself. So what could you make with
this kind of deep learning? Oh, and by the way, AI
News is going bi-weekly, so I’ll see you in two weeks. [INTEL AUDIO TRADEMARK]

Intel® Performance on ResNet-50 Deep Learning Inference | AI News | Intel Software
Tagged on:                                                                                                     

Leave a Reply

Your email address will not be published.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