Porter Stansberry has a gift for weaving fascinating themes from financial history into a stock pick that seems inevitable. Or maybe its his copywriters following his well-worn path these days, I don’t know, but I always find the “free intro” part of his Complete Investor articles interesting… whether I end up liking the stock or not.

And this was a doozie, you can check it out here but the story starts with Nikola Tesla trying to design a death ray, with designs that were later stolen by the Nazis, which leads to atom splitting technologies and cyclotrons that sprouted up around the world in the second half of the 20th century, which was part of the founding of CERN, which is where we finally get to the idea we’re talking about today:

“After mastering the Cyclotron and the Betatron, particle scientists Touschek and Wideroe dug in at CERN to build the next level of particle accelerator: the super-powerful 1950s Synchrotron, capable of speeding up particles to billions of electrovolts. A Who’s Who of international eggheads soon joined them.

“By the 1970s, CERN boasted a bewildering lineup of particle reactors – including a Synchro-Cyclotron, Intersecting Storage Rings, and a Super Proton Synchrotron – had upped the energy needs of those machines by 1,000-fold, and had commandeered a string of power stations to run its ever more complex atom smashers.

“At that point, one of CERN’s research physicists got fed up with this increasing energy inefficiency in the particle-science field. He decided to invent a method of power conversion that would take tiny amounts of power and turn them into large amounts, so he took a step back from CERN and started his own electronics company.

“Which brings us to this month’s recommendation…

“In a twist that would make Nikola Tesla proud, this former particle physicist has now become a key player in our biggest modern-day arms race: he’s built a weapon of mass production for the artificial intelligence (“AI”) industry.”

OK, so it’s a long path to get there, but that path direct your thinking where the copywriter wants it to go — you want to know the story, since it has some narrative grip, so you consider signing up for the newsletter… and once you buy the newsletter subscription, you are predisposed to believe that the company’s success is all-but-inevitable. If your brain is like a lot of others, it is great at two things: Following linear arguments that embrace its ability to recognize patterns, and rationalizing decisions to protect itself from feeling stupid.

The story goes on to highlight the “Quiet Lineage Behind Modern Computing,” which credits three 1960s-era semiconductor pioneers with their breakthroughs that laid the foundation for computers – Robert Noyce and his integrated circuit, and Bob Widlar and Bob Pease’s analog circuits… and goes on:

“These three men share a pattern that we believe is the most underappreciated wealth-creation pattern in technology investing.

“They didn’t just discover something – they built the companies that turned the discovery into core infrastructure

“They were personally the named inventors on the foundational IP, rather than the CEOs who hired the inventors – the patents have their names on them….

“When the world finally caught up to what Noyce, Widlar, and Pease had built, it was already too late to compete with them. The IP estate had been laid down and the companies had been built. The moat was 20 or 30 years deep before anyone outside the industry noticed.”

And, of course, that leads to the “missing pioneer” story:

“We believe there is one more name that belongs in this lineage. He is alive, he is still working, and his life’s work is on the verge of becoming the most critical infrastructure layer of the AI era.

“He is still the founder, CEO, chairman, and largest shareholder of the company we will be recommending this month.”

So that’s the research scientist who got frustrated with the power inefficiency at CERN and went on to start his own thing… and that means Complete Investor is talking about Patrizio Vinciarelli, and the company he founded in 1981, (and still leads at 79 years of age), Vicor (VICR).

Here’s how Vicor describes its market:

“The applications in which our products are used are typically in the higher-performance, higher-power segments of the market segments we serve. Our products are sold worldwide to customers ranging from small to large global original equipment manufacturers (OEMs), their contract manufacturers (CM’s) and internationally to distributors who support their supply chains. Our Advanced Products serve customers concentrated in the High Performance Computing, Aerospace and Defense, Industrial and Automotive markets. Applications range from powering advanced processors in server motherboards and artificial intelligence accelerator cards and systems, server rack power conversion, radar systems, industrial automation, instrumentation, test equipment, solid state lighting, telecommunications and networking infrastructure, and mild hybrid and pure electric vehicle (EV) automobiles. Our Brick Products serve customers in aerospace and defense electronics, industrial automation, industrial equipment, instrumentation and test equipment, and transportation (notably in rail and heavy equipment applications).”

Vicor was a pretty sleep analog semiconductor company for decades, rising in value but never posting exciting growth or attracting much attention… until it got a spurt of activity right around COVID, driven by rising demand for their high-density power modules in both data centers and, to a lesser degree, electric vehicles. And now, of course, the data center boom and the focus on power electronics in data centers has caused anything even remotely related to boom, including Vicor, so the stock has shot higher over the past year, up 6X or so (it has moved from around $100 to over $300 just this year, though has now dipped back down to $275 or so).

So why might we consider it now? Well, I can’t say that I have a great understanding of electrical engineering or power management, but they seem to be the leader in 48V power conversion and management modules, with solid customer relationships in hot areas like space, EVs, and high-performance computing that could continue to generate strong growth.

And, more importantly, this is a manufacturer, they’ve been making power modules here in Massachusetts since the 1980s… but they’re also leaning more on their intellectual property in recent years, and are now, thanks both to their own hardware manufacturing constraints and their legal wins, pursuing a two-pronged business model, both building their own high-end integrated power control circuits and licensing their technology. That might set the company up for some good positive surprises in the years to come, because new licensing deals will generate revenues at close to a 100% gross margin, so those revenues can drop right down to the bottom line as earnings.

It might well be that investors are already overpaying for that potential growth, that’s a judgement call — Vicor has gotten very popular with investors, who have noticed that Cerebras (CBRS), for example, last week’s hot IPO, relies on Vicor power management for its massive “wafer scale engine” chips. Or it might be that Vicor is at the center of an explosion of demand, will see its IP royalties explode just as their own capacity increases also enable them to sell more hardware, and the stock will continue this wild run for years.

This is still a relatively small company, though it’s hard for me to adjust to this world where a $12 billion market cap is “small,” and analysts expect earnings growth to be extreme over the next couple years, as their boost in capacity to reach their own $1 billion/year revenue target, drives efficiency and pushes earnings from about $2.50/share last year to something over $6 in 2028 (their “goal” of $1 billion in revenue implies an operating income target of ~$350 million, which is not the same as “earnings” but mean roughly $7/share in operating income). So you can’t say that $275 is “cheap,” that’s still roughly 40X 2028 earnings… but it’s also true that nothing in the semiconductor space is cheap right now.

And there’s at least some possibility that after being lulled a bit by decades of pretty mediocre earnings growth from what has largely been a typically cyclical analog semiconductor company, analysts are maybe understating the pace of growth and the boost from the royalties business over the next few years. They seem to be pretty early in this transition, and in their capacity expansion, and generated revenue growth of about 20% over the past year… and there is a catalyst coming next year, as they are hoping for another legal win for their intellectual property. Here’s how they put it on the call:

“2026 is a year of great opportunity for Vicor. We expect Q2 revenues of nearly $126 million, and 2026 revenues of nearly $570 million. This guidance is based on conservative assumptions about our licensing practice, specifically that we will not enter into new licensing agreements until our second ITC case gets to its final determination in 2027. Additional exclusion orders further restricting importation of infringing computing systems will provide motivation to close new licensing deals on the right terms. Along with revenue growth in 2026, we expect margin expansion.”

And they also reinforced how important Cerebras is to their computing business:

“Our lead computing customer is continuing a steep production ramp of its wafer-scale engine with best-in-class AI inference performance. Wafer-scale engines and future embedded multi-die and CoWoS packages for AI chiplet solutions are uniquely enabled by vertical power delivery.”

The call also had a lot of commentary about their growth ambitions for a second fab, but also reinforced their assertion that they have room to reach $1-1.5 billion in annual production from their existing fab as they look for locations to expand more aggressively… and as they de-emphasize some of their less-advanced products (like their “bricks” line) to make more room for advanced stuff.

So how does this end up working out? I can’t say that I understand the business well enough to have a high level of certainty, but I did listen to a couple conference calls and review their filings and their expansion plans, and the level of optimism is pretty contagious as they double down to take advantage of the spiking demand for high-power computing solutions which happen to mesh well with their historic (and ongoing) innovations in power management.

Orders are ramping up pretty quickly, most clearly evidenced by their 12-month backlog growing by 70% in the past quarter while revenue grew just 5% sequentially. Their revenue guidance for the year is for roughly $570 billion in sales, so that is the number analysts have stuck to, but it’s very likely to be too low at this point, and the growth from there to the stated $1-1.5 billion in annual capacity should be rapid, given what they’re saying about their end-market demand if would be surprising if they’re not over $1 billion within a year or two (the highest analyst forecast has them hitting $1 billion in 2028).

The semiconductor business makes me nervous right now, with the extreme enthusiasm that has driven so many stocks to pretty goofy valuations, but a supplier who provides power modules for high-end computing chips should be in a pretty sweet spot for the next couple years, and that could be enhanced if they continue to win their legal battles in the ITC and are able to increase the royalties/licensing fees and block imports of infringing products.

So might be a little too sucked in to the story, after Porter’s pitch that Patrizio Vinciarelli should be mentioned in the same breath with Robert Noyce… and perhaps more importantly, after listening to Vinciarelli’s confidence and optimism on their earnings calls, I decided to nibble on a few shares. That’s relatively easy for me to do, since I don’t have a ton of very speculative growth investment in the portfolio right now, and no real exposure to semiconductors other than NVIDIA, but I should note, even as I tell you I bought a few shares, that if substantial improvement from licensing and expansion doesn’t come through over the next 18 months or so, this is firmly in the “nutty” camp when it comes to valuation.

With the likelihood that analysts are underestimating the growth potential at least a little bit, I can justify paying roughly 40X what I think they’ll report in 2027 earnings for a 30% revenue growth story, particularly if it could lead to earnings growth above that level if they do indeed hit the licensing revenue surge that they clearly expect to come in 2027 (right now, earnings growth forecasts are at only 32%/year, so analysts are holding off on predicting that margins will improve). Risky, but I think worth a nibble. It would have been nice to learn about this company last year, when things were a little less established, but I didn’t… and we don’t get to go back in time, we can only judge the opportunities in front of us.

Disclosure: Of the companies mentioned above, I own shares of NVIDIA and Vicor. I will not trade in any covered stock for at least three days after publication, per Stock Gumshoe’s trading rules.