Implementing algorithms in embedded processors

Article By : Majeed Ahmad

Cloud platforms like AWS and Azure offer efficient implementation of algorithms in embedded processors.

EDN and EE Times are conducting regular reviews of our editorial coverage, seeking to fill the gap between our offerings and our readers’ appetite for information. Our brain trust for this project is the EDN Editorial Advisory Board, a panel of industry luminaries—CTOs, executives, and university professors —who will help us understand where the electronics industry is heading, and unearth the knowledge the engineering and business communities need to keep abreast of during this time of rapid changes.

logo for the EDN editorial advisory board

Cloud computing is becoming the center of gravity of design and development activities, and it’s now covering almost every aspect of embedded system design. So, it’ll be interesting to watch how semiconductor companies leverage the cloud technologies, noted Jeff Bier, president of BDTI and founder of the Embedded Vision Alliance.

During an interview with the EDN Editorial Advisory Board, Bier said that algorithms are now made available in the cloud using open-source frameworks such as TensorFlow and PyTorch. Next, a variety of toolsets in the cloud enable designers to embed these algorithms in processors. “In the future, it’s going to be critical which processor has the easiest path from perfecting algorithm in the cloud to getting it implemented in the embedded processor,” he said.

Bier: Cloud platforms now cover almost every aspect of electronic design.

In short, whose cloud has the embedded implementation button? The cloud platforms like AWS can come up with claims, saying, “give me a chip for efficient implementation of algorithms for the targeted processor.” That’s a lot easier and faster than writing a code yourself.

The cloud platform suppliers like Amazon and Microsoft have the incredible scale to solve the design problems. Bier also noted that there are plenty of EDA tools in the cloud. Not surprisingly, therefore, design engineers are gravitating toward chips that support such cloud platforms.

Intel’s DevCloud is a case in point. All development boards and tools are connected to Intel servers, and all engineers need is an account to carry out design work from home or the office without buying anything. A variety of new tools and new ways may well emerge in the coming months and years.

Read more from this interview with Jeff Bier, president of BDTI and founder of the Embedded Vision Alliance, at our sister publication EE Times.

This article was originally published on EDN.

Majeed Ahmad, Editor-in-Chief of EDN, has covered the electronics design industry for more than two decades.

Related articles:

Leave a comment