Embedded Computing Design

Seemingly, if you've read any news report that touches on the technology sector, there’s some mention of artificial intelligence. But too often, we hear people talking about AI as an application, and that’s wrong, just like the IoT is not an application.

Welcome to Part II of this two-part blog. In Part I, I explained why AI is the enabler to an application, and not an application unto itself. This is a fairly common misconception, especially as AI is such a new technology.

Speech recognition on an FPGA? That doesn’t sound like the most effective path, but Bill Jenkins, of Achronix had a different opinion. Hear his take on why the FPGA is the right way to go for this application on this week’s Embedded Executives podcast.

We say all the time that any system can be hacked and this week’s Embedded Executives podcast guest is someone who can get into your system.
 

Having a standard that everyone can design to makes markets take off. Open source, both hardware and software, is nothing new. But when paired with AI, it is relatively new.
 

I’ve said it before and I’ll say it again—John Deere is a technology leader that sells agriculture equipment, not the other way around.

Gallium nitride technology has been in the news quite a lot over past couple of years. It’s being touted as a far superior alternative to some of the competing technologies, particularly those that are silicon-based.
 

Whether you’re aware of it or not. AI is pervading many of our “traditional” applications, such as HVAC.

If you took a general poll in the embedded space and asked, “What’s currently the hottest application,” a majority of the responses would be AI. However, that response would be inaccurate because AI (or its subset machine learning) is not an application. It may be the means to some other application, but on its own, it’s a technology.

Big announcement from the Biden administration—the U.S. Cyber Trust Mark. In its simplest form, it’s a logo/sticker that goes onto any compliant IoT device. In its most complex form, designers need to understand what this standard is all about.

Adding touch capability to your display can be a tricky endeavor, especially if that display is aimed at an industrial application, where the conditions may be less than perfect. In this video, you’ll see how Infineon addresses the issue with a product that works just as well in consumer products as industrial.

AI is here and it’s everywhere. However, AI is not the end game. AI is used to reach the end game, whatever that happens to be.

Smart door locks are fairly common today, but they could be significantly smarter (and lower power). 

Compute Express Link, aka CXL, is a low-latency, cache-coherent interconnect that connects processors, memory, and other accelerator devices. It’s built on top of PCI Express and over the past couple of years, it’s been emerging as the winner in the interface wars.

NXP’s NHS52Sx4 family of ultra-low-power MCUs is aimed squarely at battery powered, miniaturized wearable medical devices thanks to its high level of integration and versatile battery support.

AI seems to be omni-present. It’s in the news, it’s in our classrooms, and it’s certainly in our engineering labs. But are we spending too much time discussing the technology that’s the current “new thing,” or is it really the greatest technology of our lifetime?

Energy harvesting has been with us for quite some time, but it has not achieved the lofty goals that industry insiders predicted. 

It took a little explaining for me to understand this one, but now it makes sense, a lot of sense.

It really is possible to make memory that’s geared toward a particular application, if you’re still of the belief that AI is an application. I learned that fact recently, thanks to the folks at Micron Technology, who recently announced that they are delivering qualification samples of the company’s Universal Flash Storage (UFS) 4.0 mobile solution.

Saying that non-volatile memory (NVM) is ubiquitous is an understatement. From our consumer devices to industrial and medical/healthcare applications, even in our cars, NVM is omnipresent. It’s used to store the operating system, the program files, our pictures, videos, and other data. In environments with high levels of radiation, such as aerospace and medical devices, selecting the right NVM can be a key consideration since radiation can impact the operation of many technologies.