Embedded Computing Design

Avnet recently surveyed part of its core audience, and the topic was AI’s Impact on Engineering. It doesn’t get much closer to home than that. 

We know that AI is everywhere. An area of particular concern for the DevTalk crew is how AI fits into today’s higher education programs. With our special guest Kevin Lu, the Associate Dean for Engineering and Science for Undergraduate Studies at Stevens Institute of Technology, we look at two specific areas—how do you ensure that students aren’t “cheating” by making use of AI, and second, how do you ensure that the careers you are getting them ready for are the right careers, in this changing world of AI? It’s a fascinating topic, and one that will evolve regularly.

We’ve all seen the wireless charging devices; you set your phone down on the charger, come back after a while, and your phone is charged. If the charger manufacturer followed the standard set forth by the Wireless Power Consortium (WPC), that scenario would play out. If they didn’t follow the standard, then you may come back to an uncharged device. 

The user interface, aka the HMI, is oftentimes the only interaction a user has with a platform. Hence, it represents your brand, and it needs to look great and work great.

It’s time for an update on Wi-Fi. On this week’s Embedded Executives podcast, we take a look at what’s the latest, what’s coming, what’s real, and what’s not so real, all as it pertains to Wi-Fi. 

There are lots of different ways that IoT devices can communicate. In some instances, that’s a good thing. But in other cases, it’s hindering the growth of IoT devices. And that’s a problem, at least according to Mike Rohrmoser, a Vice President at Digi. 

In this week’s Embedded Executives podcast, Mike explains how protocols are evolving and what needs to occur to ensure interoperability between all devices, including those that may have been in the field for decades.

There’s no moss gathering under the Connectivity Stands Alliance, as the group recently announced the latest rev of the ZigBee spec, version 4.0. This new version enhances the security efforts from the previous generation, and also takes a hard look at smart energy in the latest devices. 

One of the keys, as you’ll hear, is the backward compatibility that’s been maintained. Faisal Bhaiyat, the Co-Chair of ZigBee Working Group, Connectivity Standards Alliance, and also a Senior Software Architect with Silicon Labs, fills me in on the details on this week’s Embedded Executives podcast.

High-performance computing at the Edge involves systems that bring server-class processing closer to where data is generated. Instead of sending private information to the Cloud, devices at the network’s Edge, gateways, sensors, industrial controllers, cameras, and embedded AI platforms, run complex workloads locally. This is beneficial for applications like medical imaging, Edge servers, networking, and enterprise-level LLM (Large Language Model) deployment, and includes markets such as healthcare, industrial automation, smart cities, robotics, automotive, and advanced consumer devices. Its growth is driven by tighter latency requirements, rising data volumes, stronger privacy demands, and the need for more efficient, scalable distributed compute architectures.

You’re likely aware of many of the great things AI can do for you. And I’m guessing you’ve seen many of the nefarious things it can be used for. I was just brought up to speed on one of the latter items, and it’s something you should be aware of as well. 

In my discussion with Marcus Webster, the Founder and President of Webster and Webster Associates, I learned that people are using AI to help them get jobs, and in some cases, they are jobs that they should not be getting, for security reasons. People are submitting fake resumes and even conducting fake video-based job interviews. Honestly, it’s both fascinating and scary. 

Check out this week’s Embedded Executives podcast to learn more.

Most of the RISC-V offerings that I’ve seen have been pushing very high performance, with lots of cores and lots of chips. Upbeat Technology is taking a different tack by looking at very low power levels. 

In my discussion with Jerry Chen, Founder and CEO of Upbeat Technology, he explains the reasons for that approach, and, as you’ll hear, he tells it like it is and explains why the company partnered with SiFive. 

The applications that make sense for the Upbeat offering include always-on IoT, wearables, drones/UAVs, smart audio, predictive maintenance, and edge-AI sensors, basically anything battery-powered. Hear our discussion on this week’s Embedded Executives podcast.

I recently spent some time with Synaptics’ new CEO, Rahul Patel. It was at the company’s Tech Day event held in its San Jose headquarters. Rahul arrived at Synaptics after spending more than a decade at both Qualcomm and Broadcom, so he is no stranger to the semiconductor space. And that’s a good thing, because he joined at a time when the company was already in the midst of a transformation, attempting to become a major player in the Edge AI semiconductor space.

It’s not the first time this has occurred, but it is the latest—a previously commercial real-time operating system is now being offered as open source. In this case, it’s SMX from Micro Digital. 

If you’re wondering, like I was, where the company would choose to go this route, check out this week’s Embedded Executives podcast with Ralph More, the Founder and President of Micro Digital.

There are lots of semiconductor vendors that do a great job of providing AI hardware. But unfortunately, for many of them, their offerings end there, requiring developers to source software and tools from multiple sources. It’s no secret that designing a system with AI is complex enough. Having to go to different vendors for different pieces just adds multiple layers of complexity. 

Every IoT device should comply with the Connectivity Standard Alliance (CSA) Matter standard, but we know that’s not the case. For some developers, doing that is a no-brainer. However, there are some developers who struggle with the process. For that latter group, the CSA recently released a Matter-compliant platform certification that should help them get to market far more quickly. 

To explain how that works, I spoke to Jon Harros, who leads the Alliance Certification and Testing team at the CSA, on this week’s Embedded Executives podcast. Jon goes through the need for the program, how it’s implemented, and what developers need to know.

Synaptics and Google have joined forces to hasten the development of low-power, AI-driven edge devices—marking a potentially significant milestone for open, efficient, and scalable compute architectures. The collaboration brings together Google Research’s new Coral machine-learning accelerator architecture and Synaptics’ deep expertise in low-power silicon design, resulting in the company’s SL2610 platform.

Compilers and ADAS aren’t often used in the same sentence. For developers, that anomaly can represent a design challenge, as compilers in general are not designed for this particular application, which also happens to be at the top of the safety and security checklist. 

To understand how these compilers should be tested and implemented, I spoke to Sjoerd van der Zwaan, the Chief Product Officer at Solid Sands, on this week’s Embedded Executives podcast. 

Sjoerd explains how and why compilers for ADAS are different from general compilers, and why you need to run your compiler through an appropriate series of tests to ensure compatibility.

Every once in a while, we hit a home run when it comes to bringing guests onto the DevTalk podcast, and today we did just that. 

We managed to wrangle in Hidetoshi Shibata, who has been in his current CEO role at Renesas for about six years. Toshi was very forthcoming with us, as we went through his fairly colorful past, and more importantly, the decisions he has had to make as the CEO of one of the most powerful companies in our space. 

AI at the endpoint gets yet another boost from EMASS, a fabless semiconductor company that recently emerged from stealth mode. 

Having your processing at the endpoint presents a few significant challenges, namely the amount of processing that can be made available and the ability to run from a battery. EMASS has come up with a way to do this a little differently from what’s been done previously. They can attach their RISC-V processing element to any sensor, making the device an AI accelerator. It’s a little difficult for me to put into words here, so check out this week’s Embedded Executives podcast, where I spent some time with Mark Goranson, EMASS’s CEO, who explains it in detail.

No matter what your end device is, be it a light bulb or a driver’s license, it needs to be secure. I don’t think we need to go through the reasons why it needs to be secure. How you do that is becoming increasingly more difficult. And you may be looking at protecting millions of devices, where each one needs its own unique identity. 

To get a handle on how you would do this, at least where you would begin this daunting task, I spoke to two experts: Patrick Carrier, Manager for Secure Solutions at Infineon, and Xin Qiu, Head of Security Solutions at CommScope. 

On this week’s Embedded Executives podcast, they lay out the need and then the process a developer would follow.

AI is a technology that generally requires a significant amount of power and high compute performance. Hence, it’s typically relegated to the Cloud or to an Edge-based computer.

Performing AI at the endpoint has not been feasible for those reasons. With ambiq’s Sub-threshold Power Optimized technology (SPOT), performing AI at the endpoint is a reality. And the doors/applications that it opens are extremely cool. 

Hear more about it in this week’s Embedded Executives podcast, where I spoke to Ambiq’s Vice President of AI, Carlos Morales. Note that Ambiq just recently released the latest version of neuralSPOT, which now includes key performance enhancements specific to the company’s Apollo5 family, amongst other features, some of which are specific to AI development.