Hardware at the Edge: Selecting Optimal Technology for AI Applications

We evaluate four major hardware technologies - CPU, GPU, FPGA, and ASIC-based architectures.

Data processing requirements in edge AI applications are increasing exponentially.

As the software requirements evolve, engineers must also choose hardware technology that can adapt to rapidly evolving demands at the edge. They must also balance required performance, power, and cost targets. We evaluate four major hardware technologies - CPU, GPU, FPGA, and ASIC-based architectures. While some of these technologies are highly flexible, they do not scale to meet performance requirements. Conversely, others deliver high performance and low power consumption but limited capacity to adapt to changing workloads. In this webinar, presented by Colin Alexander, Director of Product Marketing, we discuss how FPGA and eFPGA IP solutions can meet the requirements of several edge applications that were traditionally only done in a data center.

What you’ll learn:

  • Balancing performance, power, and cost targets
  • How to evaluate hardware technology
  • FPGA and eFPGA solutions

 

Register to View Webimar

Country

About the Speaker(s)

Colin Alexander

Colin Alexander – Director of Product Marketing

A resident of Edinburgh, Scotland, Colin Alexander brings over 35 years of technology experience to the Achronix marketing team. As Achronix's Director of Product Marketing, he lends years of multi-disciplinary engineering, marketing, and strategic management skills to connect the Achronix suite of semiconductor solutions with top manufactures in the automotive, networking, telecom, AI and machine learning, and computational storage fields. Colin has a long history of successfully bringing technology products to market.