

How to make your CPU as fast as a GPU - Advances in Sparsity w/ Nir Shavit
Oct 23, 2022
Chapters
Transcript
Episode notes
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Introduction
00:00 • 3min
Neural Magic - A New Way to Optimize GPU Performance on CPUs
03:00 • 2min
Is There Something More to Sparse Than to Dense Neural Networks?
05:11 • 3min
Is the Brain Sparse Because It Has the Limitations of the Cell Phone Power?
08:33 • 3min
Are We Getting Rid of the Flops?
12:00 • 3min
The Big Game of Efficiency and Neural Networks, Right?
14:43 • 2min
The Harm to Layers
16:51 • 3min
How to Be More Accurate When You Have a Sparse Model
20:11 • 3min
The Limit That Keeps Us From Building the Sparse Model
22:48 • 3min
Why GPUs Are So Unsuited for Sparse Models?
25:40 • 3min
Neural Magic Is a Way of Running Deeper Neural Networks
28:24 • 4min
Is There a Type of Sparse Model for Neural Network Execution?
32:33 • 4min
Combined CPU and GPU Architecture for Machine Learning
36:45 • 4min
Graph Neural Networks
40:55 • 2min
Is There a Connection Between Machine Learning and Neural Networks?
43:08 • 2min
The Exciting New Thing in Neural Magic
45:25 • 2min
How Easy Is It to Get Started With TensorFlow?
47:02 • 3min