Pitchmen love to find the hottest product, and then say they’ve used their uncanny skill and their unparalleled industry connections to find the secret little company who will benefit from that hot product. They’ve been doing it for years with Apple, fueled by the cottage industry of “whisperers” inside Chinese manufacturing plants who start the rumors flying, and by the folks who always buy the latest Apple device on the first day and then tear it apart to try to identify the semiconductors and other parts they used.

Sometimes it works a little bit, for a little while… but more often, in my experience, it’s the investing version of a conspiracy theory — employee X worked for this company before and has a new company, now, so he’ll have the inside track as the supplier for the next hot product. From Luke Lango’s breathless spiel about how certain he was that his favorite LiDAR stock would get the contract for the Apple Car (project now canceled), to Ray Blanco’s pitches over the years about his expectation that Elon Musk would ID his little battery supplier at Tesla’s next Battery Day (he didn’t), to James Altucher’s pitch, just a few weeks ago, about the tiny company that was bound to be the supplier for Apple Vision Pro headset (it isn’t). Sometimes the stock went up a bit, largely driven by the attention stirred up by these manic pundits, but it is extremely rare for it to really work out.

Novonix is now back down below where it was when Ray Blanco poiunded the table for that one in 2020 and 2021 as the next Tesla partner (and in a two-fer, his other favorite play at the time, Fisker (FSR) as a “next Tesla” car company, is probably on the brink of declaring bankruptcy). Luke Lango’s “i-Car supplier” Aeva (AEVA) is down 80%, thrown on the same trash heap as all the other LiDAR hopefuls that went public through SPAC mergers in 2021, and Kopin’s (KOPN) earnings call a couple weeks ago threw even more cold water on Altucher’s promise that they would supply critical components for the Apple Vision Pro (he’s still peddling that as a “special report” bonus in this current tease).

Those are just a few examples of the “overpromise” tendencies of pundits who are promoting the “secret” supplier or partner for the “next big thing.” We could have named a dozen others, but you get the idea.  And coincidentally, perhaps, these highfalutin’ promises tend to be used to sell “upgrade” newsletters — the $2,500 or $5,000 subscriptions that make investors believe they really provide the keys to the kingdom, and which are, interestingly enough, almost aways nonrefundable commitments.  They usually offer some kind of credit for these nonrefundable upgrade subscriptions, even a generous-sounding one — but the nice thing about newsletter publishing is that it’s almost infinitely scalable… they don’t care how many emails they send, what they care about is getting the dollars in the door and keeping them. It’s no skin off their nose to give you “double credit” for some other letter, or give you a second year for free if you complain, those “guarantees” cost the publisher nothing at all, they just smooth the disappointment of the person who ponied up $2,495 nonrefundable dollars and is furious that the promised 100X gains didn’t happen.

So what’s the latest pig in a poke being promoted? It’s from James Altucher again, and it’s all about NVIDIA… which, of course, is the hottest stock with the hottest story in the market, as NVIDIA chips are flying out the door at premium prices to power everyone’s favorite new AI experiments.

The first version of this ad, which we covered back in March, was about NVIDIA’s anticipated reveal of its new Blackwell chips (Altucher calls it the “X Chip”) at their developers conference that week, (called GTC). NVIDIA was in the news that week, of course, and a lot of partnerships and breakthroughs were talked about… and lots of folks see continuing catalysts in the AI industry, driven, at least in part, by the increasingly powerful chips coming out of NVIDIA (those Blackwell chips were announced at the conference, with a lot of details unveiled, and they will be hitting the market in the coming months… though the earlier generation of GPU chipsets for AI work are still flying off the shelves, and NVIDIA and their manufacturing partner, Taiwan Semiconductor, remain “capacity constrained”).

And that, of course, has helped NVIDIA reach new highs, particularly following the relief and/or excitement which followed their earnings report last week (when they “beat” the profit estimates and raised their forecasts yet again).   NVDA and it’s place in the market are unprecedented in many ways, not least because we’ve never seen one of the world’s largest companies (by market cap) trade at anywhere near this valuation (35X sales, 40X earnings), but it’s certainly working fantastically well for investors right now, and the growth is still very imprssive.  Even if it might feel like we’re holding on to the tail of a tiger — living in at least a little fear, as NVDA shareholders, that their gross margins will eventually have to come down and their revenue growth slow.

So let’s look at what the investment actually might be, what the odds of success are, and whether we should be expecting a big “win” this time as the Blackwell chips roll out to customers “any day now”, after Altucher failed to deliver his big news on March 21 (the promise at that time was that NVIDIA would mention this secret little pivotal partner company as part of the release of the Blackwell Chip/X Chip… and they didn’t).

Ready for round two?  Think it’s going to be the same company again?  Let’s see what we can learn…

The spiel is still all about the “X Chip,” and that’s still NVIDIA’s Blackwell chip… but the promise is no longer that we’ll get rich when it’s introduced, since that ship has sailed, but that we’ll get rich as those chips roll out.  From Altucher:

“While NVIDIA is manufacturing the X Chip, I believe it will be an entirely different AI company whose shares soar the most off of the release and roll-out of this new technolog… because when NVIDIA officially announces the X Chip release date, they’re going to shock the world with one major detail that most people have no clue is happening behind the scenes.

“NVIDIA is going to announce that they have partnered with a virtually unknown company to unleash the full potential of these X Chips.

“You may be wondering why mentioning this company’s name will be such an important moment….

“… they are 1,700X smaller than NVIDIA… but they’re NVIDIA’s most important partner on this project.”

“I believe NVIDIA executives are going to mention this company… and when that happens, we’ll see this tiny AI firm’s shares go haywire.”

Here’s the pitch from the order form:

“Join Paradigm Mastermind Group Before Nvidia’s X Chip Rollout

“James Altucher will teach you how to immediately act on this A.I. opportunity – that’s poised to surge 10,000% by 2025…

“The AI Wealth Window is about to transform into an Accelerated Wealth Window.

“And this could be a life-changing financial event for those who act immediately.

“Because a tiny $1 billion AI firm is standing at the center of this all.

“And with their Nvidia partnership…

“This small stock could see its revenue surge to $50 billion or $100 billion by 2025

“I believe that kind of growth would mean up to 10,000% returns.

“And the biggest rise could take place immediately following March 21st – when the X chip is publicly unveiled.”

And Altucher’s colleague Doug Hill calls it a “tiny contractor who will be explicitly involved in this massive A.I. infrastructure project.”

As with Altucher’s overheated pitch about the Apple Vision Pro a couple months ago, this is a sales pitch for the Paradigm Mastermind Group, which is essentially just a combo newsletter from Altucher and his Paradigm Press colleagues Ray Blanco and Jim Rickards. They’re charging $2,495/yr., with what they call a 100% satisfaction guarantee… which, as I noted above, doesn’t mean you get your money back if you’re unsatisified, this is completely non-refundable, but it means you get a “credit” that you can use to buy the other Paradigm Press newsletters. So yes, to be clear, it doesn’t much matter whether or not this stock pick is successful, they have your $2,495, and you’re not getting it back.

What other clues lurk in the presentation?  Well, the further along we get in this pitch… the more clear it becomes that Altucher’s “interview” with Doug Hill is following almost exactly the same script as the ad they did about two months ago… they’ve just excised the references to “March 18” and instead put in some less-specific language about the “rollout” of the X Chip happening “soon” (their “deadline” is now May 28, but that’s the deadline for the newsletter deal, not a specific date for the release of NVIDIA’s new Blackwell chips, or the news from this little company).

And the specific clues about the little company have really not changed at all since the first version of that “AI Wealth Window” presentation….

Here’s how Altucher describes the stock in the ad — again, not many real clues:

“This technology needs to be custom-tailored for the client and then maintained and managed by a company that specializes in this type of A.I. infrastructure

“NVIDIA focuses on the manufacturing and development of AI, so they’ve opted to outsource this process to the company I’m telling you about

“High-performance computing firm that focuses on setting up enterprise AI systems… when somebody buys NVIDIA’s new chips, this company is going to do all the heavy lifting.

“NVIDIA makes these chips, this tiny company installs them. NVIDIA needs them to roll these chips out.”

There are literally hundreds of NVIDIA partners, from global system integrators to distributors to suppliers, with locations around the world… the list that NVIDIA makes public actually numbers over 1,000 partners, and includes most of the world’s large technology contractors and consulting firms, as well as lots of smaller companies.

Many of these are private companies who are well-established in the IT consulting, hardware and distribution businesses, like SoftServe, WWT, and Lambda, or specialized distribution companies that “own” a particular country or region.

The ones who might reasonably fit the limited clues we’ve received are ePlus (PLUS), SMART Global (SGH), mostly thanks to its Penguin business, PC Connection (CNXN), Perficient (PRFT), and maybe even GDS Holdings (GDS) if you want to stretch to China.  So who might it be?

The only other clue we get is…

“… insiders at NVIDIA have declared this is one of the most important partnerships that NVIDIA has… this tiny A.I. contractor is already working with the Department of Defense, Tesla, Meta, and even the Department of Energy… but the X-chip will likely make them one of the busiest and most explosive companies throughout the A.I. sector.

“… this is the company that’s going to install it and set it up everywhere.

“This behind-the-scenes company is valued at just $1 billion… incredibly small… massive opportunity to scale”

“One of the easiest things you can bank on in AI right now.”

From what I can tell, only one of those companies has been publicly connected to the Departments of Energy and Defense, and has worked with both Tesla and Meta Platforms in some specific way that made it into a press release… and it is still right around a billion dollars in market cap, and is a preferred partner for NVIDIA (in fact, in a recent year they got two “partner of the year” awards in different areas). That’s SMART Global Holdings (SGH).

So yes, exactly the same “secret back door partner” Altucher said would be named in NVIDIA CEO Jensen Huang’s announcement of their new chip back in March.  It wasn’t mentioned, but one hates to waste an ad… so they’re trying again, hoping that it will be mentioned “in the press release” when the B100 actually starts rolling out to customers.  I wouldn’t hold my breath, but one never knows… and investors do sometimes get nutty about companies that are mentioned in the same breath with NVIDIA (speculators even drove up the share prices of the few companies that NVIDIA actually owns shares of a few months ago, just because NVDA’s small portfolio of investments in partner companies and startups finally got big enough that they had to disclose it in a 13F in February for the first time… though those partnerships were mostly quite old and none were recent “buys” by NVDA).

Smart Global is not a big company, but it does have some programs that specifically put NVIDIA hardware to use to help people build AI projects, including essentially building servers which use NVIDIA chips as well as offering consulting and platform solutions in high performance computing in general. They have a market cap of about a billion dollars, and also sales of about a billion dollars a year of late.

They don’t stand out in any particularly way when you look at their numbers, revenue has been roughly flat over the past five years, earnings were flat until they fell late last year, and the few analysts who follow the stock think they’ll have adjusted earnings of about $1.21 per share this year and $2.03 next year (the expectations didn’t change much after SGH’s last earnings report).  I’m not settling on this as my guess because I love the idea, just because I think it’s the best match for the very limited clues we received.

If there’s going to be an exciting surge of business for Smart Global, it will probably be within their Penguin Solutions business, which they say “designs, builds, deploys, and manages AI and accelerated computing infrastructures at scale” and has “deployed over 50,000 GPUs in partnership with leaders in AI.”

The company today is an interesting little conglomerate that has been involved in several different parts of the semiconductor industry over the past 25 years or so, and has been bought and sold a few times, including a couple IPOs along the way. They had been primarily in the memory chip business, including owning some manufacturing capacity in Brazil (which they have now mostly sold), but have shifted focus over the past five years, starting when they acquired Penguin Computing. They later also acquired the LED business from Cree, so they still have a bunch of different businesses, but they want to lean more on services revenue, and most of that is in their “intelligent platform solutions” group, including Penguin. There’s a good presentation about that Penguin business here, from last October.

Excitement to come? I don’t know. I’d be shocked if Penguin or Smart Global was mentioned when in any official “first shipment” press release from NVIDIA, and it certainly wasn’t mentioned when Jensen Huang officially unveiled the B100 in March, but I guess one never knows… and the CEO of Smart Global is one of the many partners who participated in a panel session at NVIDIA’s GTC Conference, and they did, of course, also have a booth in the exhibit hall.  They are a real partner… just not necessarily a uniquely important partner that’s more important than hundreds of other resellers, contractors and consultants who work with NVIDIA customers.

Why might it not be Smart Global? Well, lots of reasons… we can’t really claim any certainty in this match, but some of what Altucher implies doesn’t necessarily fit, like…

“They’ve inked a massive deal with NVIDIA to be the AI giant’s middleman”

They do sort of work as a middleman, I guess, and they manage tens of thousands of GPUs, but I don’t know that they have particularly “inked a massive deal” that stands out among NVIDIA’s many, many distributors and system integrators. They’re not likely to buy nearly as many GPUs as Dell or HP or the other big server makers, and they won’t be first in line to buy them like Microsoft and Amazon and Alphabet might be, but sure, I guess they’re one of many middlemen that are helping mostly smaller companies build AI hardware capacity using NVIDIA products.

So nothing definitive for you today, dear friends, but one interesting little company to research while you wait for the Blackwell chips to roll out… assuming you don’t have something better to do with your limited time on this earth.

We should take projections with a grain of salt, but Altucher says they could quickly clear into a $50 billion market cap, maybe even $100 billion in 2025, which would be that 100X investment he teases.

That’s poppycock, of course.  There’s nothing all that scalable about a service provider that helps people set up AI projects (and does lot of other technology contractor work)… but that doesn’t mean the stock won’t go up, it just means we should keep those “100X in a year” daydreams out of our heads.

Yes, the future belongs to the optimists… but there has to be a limit.

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