skip to main content

Introduction

The field programmable gate array (FPGA) is a highly flexible alternative to burning algorithms directly onto a microprocessor to create an application-specific integrated circuit (ASIC).

Highlights

  • Xilinx is re-architecting its FPGAs for AI.

Features and Benefits

  • Understand the best role for FPGAs for AI between training and inference modes.
  • Learn about Xilinx's new Adaptive Compute Acceleration Platform, ACAP.

Key questions answered

  • How is Xilinx addressing the needs of general software developers to program FPGAs?
  • How is Xilinx supporting machine learning on its FPGAs?

Table of contents

Ovum view

  • Summary
  • Xilinx is re-architecting its FPGAs for AI
  • FPGAs for AI system inference mode
  • Xilinx’s new platform: Adaptive Compute Acceleration Platform (ACAP)
  • Example FPGA user: Ryft
  • FPGAs on the cloud and FPGA middleware
  • A little FPGA history
  • CPU architecture and its bottleneck
  • FPGA is a dataflow architecture

Appendix

  • Author