Home / Blog / The Silicon 100: “The building blocks of future wonders”

The Silicon 100: “The building blocks of future wonders”

Of late, we’ve been focusing so much on stories around technology and the pandemic, that I ended up putting EE Times’ Silicon 100: The Class of 2020 on the backburner. But I finally got the opportunity to look through it and found it pretty interesting.

The list of “electronics and semiconductor startups that grabbed [EE Times’] attention during the previous year” is published regularly – this is the 20th edition – and it’s the first one that covers 100 companies (vs. 60). The leap from 60 to 100 companies is a big one, reflecting the fact that investors have a lot of interest in hardware startups (the primary focus here, as opposed to software or services).

In his introductory article, list curator Peter Clarke attributes the willingness to investors to open up their purse strings to hardware companies after snubbing them in favor of potential unicorns (those firms likely to grow quickly to a $1 Billion valuation) in the software and/or services space to the end of Moore’s Law.

Clarke saw Moore’s Law as the reason why investors had earlier focused their hardware funding on “the latest ASIC of SoC that could exploit the next manufacturing process node,” ignoring other electronics activity. As opportunities in this arena decreased, and the “smart money” went into Uber et al.

Many venture capitalist firms stopped investing in hardware altogether. Much of the hardware investment that remained came from strategic investors, such as Intel, Samsung, Qualcomm, and Bosch, or from government-backed regional development boosters who hoped to encourage job creation.

But then, as the end of Moore’s Law came into sight, R&D activity moved out into other areas.

Engineering creativity extended into a slew of re­lated activities, including alternative computational architectures and business models such as open-source hardware, RF communications, sensors, 3D assembly and packaging and chiplet-style manu­facturing, and human-to-machine and machine-to-machine interfaces. Some of these activities required far less funding than the $100 million or so needed to get a leading-edge digital chip out the door, and so they started to gain some traction.

It also became clear that software and networking architectures that had proved powerful in the cloud running on servers could be supported by fresh architectures and ICs to reduce power consump­tion and potentially transfer similar functions to the edge.

The result? An explosion over the past few years of hardware funding – and the explosion of the number of startups that EE Times deems “worth watching.”

Here are the buckets – pretty much all of which Critical Link has our hands in – that the companies worth watching fall into:

  • AI / neural networks / machine learning
  • semiconductor manufacturing
  • materials research (Si, GaN, SiC, etc.)
  • analog ICs
  • SoCs
  • memory ICs
  • FPGAs
  • power ICs
  • energy harvesting
  • signal processing
  • wireless chips
  • automotive electronics
  • IoT technologies
  • sensors
  • MEMS

Of these, “the consensus now is that artificial intelligence, machine learning, 5G communications, and transportation solutions are set to drive the industry onward and upward for the next several years.”

Geographically, the US continues to rule the roost, accounting for 49 companies on the list. The preponderance of these (40 in total) are in California. There’s only one company from my home state of NY, and it’s in the Bronx, not upstate, where Critical Link lives. China, with thirteen companies on the list, ranked second, followed by England with nine. Israel had seven of the startups, Canada five.

After listing all 100 startups (with a thumbnail sketch of what they produce and who has invested in them), they take a deeper dive on ten of the companies.

The Silicon 100 is gated content that you’ll have to register for, but it’s well worth the read. I’m looking forward to the “future wonders” that will be built off all of this technology. I’m pretty sure that at some point, we’ll be looking at a number of these companies as component providers for our boards.