Data Fest 2020

AI Hardware

AI Hardware is computers and chips designed specifically for machine learning applications. In this section, we will share with you our experience in AI hardware design and highlight industry trends.

AI Hardware is a very hot area, where nanometers of silicone meets the building-scale supercomputer. It's where novel and brave computer architectures make possible fantastic algorithms. Powerful computers made the Deep Learning Revolution possible, it took decades for engineers to develop powerful enough hardware to implement DNN ideas.

Today, we stand at the point where custom AI hardware comes to our homes, laptops, smartphones, and fill the clouds. AI hardware lets us train extremely complex language models, develop safe autonomous vehicles, and makes a smartphone smarter than supercomputers of the nearest past. AI Hardware opens a door to Machine Learning algorithms from papers to life.
 

Track program

RISC-V Based Neural Network Processor, ANN in FPGAs

RISC-V Based Neural Network Processor, ANN in FPGAs

GAP8: Ultra-low power and high-performance AI processor

GAP8: Ultra-low power and high-performance AI processor

Embedded Computer Vision for Autonomous Systems

Embedded Computer Vision for Autonomous Systems

FPGA and ASIC in datacenters

FPGA and ASIC in datacenters

DNN on FPGA

DNN on FPGA

ML Perf, Machine Learning Hardware Benchmark

ML Perf, Machine Learning Hardware Benchmark

Entropy Encoding for CNN Inference

Entropy Encoding for CNN Inference

Huawei's DaVinci AI compute architecture

Huawei's DaVinci AI compute architecture

Cookies help us deliver our services. By using our services, you agree to our use of cookies.

Learn More