Managing Editor’s Note: The AI boom is well underway…
Maybe you missed out on the initial rush, or maybe you’re simply looking for new avenues to profit from the rise of artificial intelligence.
Well, this may be the second chance a lot of folks have been waiting for. Jeff’s hosting a special strategy session next Wednesday at 8 p.m. ET called AI Supercharged, where he’ll dive into an opportunity at the cross-section of two of today’s biggest tech megatrends – AI and crypto.
There are a handful of “AI coins” that Jeff’s got his eye on to take advantage… and a short amount of time to get ahead of the opportunity.
Jeff will explain everything next Wednesday, September 24, at 8 p.m. ET. You can go here to automatically sign up to join him.
Well, that was unexpected…
The tech community woke up this morning to a surprising deal.
NVIDIA (NVDA) announced that it is making a $5 billion equity investment in Intel (INTC).
In a press conference this afternoon, it was stated that the two companies have entered a product development partnership to design two new collaborations between Intel’s x86 central processing units (CPUs) and NVIDIA’s graphics processing unit (GPU) technology.
Institutional capital didn’t appear to understand the deal at all, sending Intel up 29%, NVIDIA up 3.8%, and AMD (AMD) initially down 3.3%. AMD’s share price recovered on the close, only down 0.48%.
On the surface, this deal makes no sense at all. In fact, the $5 billion investment is completely unnecessary.
So, what’s the deal?
Intel has been floundering for the last decade. It represents one of the worst train wrecks in the history of high tech, right up there with Boeing.
For decades, it held a leadership role. It was the bastion of U.S. semiconductor manufacturing… and now it is years behind the competition. It missed out completely on the largest growth opportunities of the last decade and can’t get a single new product launch right.
NVIDIA, put bluntly, does not need Intel for anything.
So before we get to the “why,” let’s have a look at “what” this deal looks like.
There are two components:
To state the obvious, NVIDIA and Intel have been competitors, though calling Intel “competition” may be generous, considering NVIDIA has been eating Intel’s lunch…
These are not two companies that we’d think of to establish a “historic partnership” (their words, not mine).
Tech companies partner all the time on joint product development without the need for a multi-billion-dollar equity investment. Both companies certainly could have announced a cooperation on products without the $5 billion kicker.
When asked about the $5 billion, NVIDIA CEO Jensen Huang replied that it’s a “great investment” on the press conference call held at 10 AM Pacific to provide an overview of the deal.
I’m absolutely sure that’s not the case. Intel is down a staggering 51% since December 2023, something I predicted on multiple occasions. It isn’t growing, and it’s mired in $51 billion in debt – unprofitable, bleeding cash, and years behind the competition in every product area.
Sure, with a long enough time horizon, tens of billions of investment, and some minor miracles, there’s hope for Intel. And to be clear, I would love for it to succeed. But it is not a good investment now.
But putting the $5 billion aside, it’s the product cooperation announcement that is so interesting, or rather, what I mean to say… It’s not at all what it seems.
Despite Huang’s comments that Intel and NVIDIA had been working on this product collaboration for a year, there were no specifics whatsoever about the to-be-designed x86 CPUs or the joint products that would be announced.
No timeline has been given whatsoever. Neither CEO could even give what manufacturing process the chips would be designed on. It hasn’t been decided yet.
Huang repeated a few times how great the deal would be for both companies:
Huang mentioned that this was a $50 billion market opportunity for the two companies.
But here is what was not said, but reflects the reality of this “partnership…”
The two CEOs emphasized several times that the new partnership was a product partnership, and not about the foundry (i.e., the manufacturing).
Ironically, on the public relations call today, Huang went out of his way to compliment and highlight how and why TSM was such a phenomenal foundry partner, with no mention of Intel’s foundry capabilities. And Intel CEO Lip-Bu Tan went out of his way to avoid admitting that Intel would continue to use TSM for these new x86 CPUs and not Intel’s own foundry. How embarrassing for Intel.
What does this mean? Less margin for Intel. TSM generates about 57% gross margins from manufacturing semiconductors for companies like NVIDIA and Intel.
But here was the craziest part of the call…
Huang also went out of his way to highlight the importance of NVIDIA’s ARM-based CPUs and that the Intel partnership would in no way impact the work that it is doing on ARM-based processors.
Worse yet, he clearly stated that the combined NVIDIA GPU + to-be-designed Intel x86 CPU would be a product offering to NVIDIA customers. Which product combination do you think NVIDIA will prefer to sell to its customers?
The answer is shown above, in that exact order.
AMD has also been eating Intel’s lunch on CPUs. AMD went from almost no market share to commanding more than a 25% market share in data center CPUs. Its products are the best on the market, better than Intel, better than NVIDIA.
And it will probably be 18–24 months before Intel is able to ship a product to work with NVIDIA’s GPUs in the data center. No revenue before that, only burn. And in the meantime, AMD will continue to materially eat up even more market share from Intel.
So why did NVIDIA even bother with the product collaboration? It’s purely a defensive maneuver against AMD’s growth.
Intel still commands a majority market share in data center CPUs, albeit it shrinks every year. And the only space where Intel remains strong, but in a stagnant market, is the PC market. NVIDIA will likely pick up some new revenue in a couple years’ time in PCs, which will amount to less than 3% of overall NVIDIA revenues coming from the low-end PC market.
So, back to my original question… why?
x
For Intel, the answer is simple. The new CEO, Tan, was in need of some good news to share after six months on the job. He’s been eager to turn Intel around since he took the helm, he needed to show some progress, and he needed the money. Tan said on the call that he took the $5 billion to “shore up our balance sheet.”
He has been doing a lot of that lately:
That’s $15.2 billion raised, plus another $5 billion from NVIDIA. For Intel, the deal is about the money.
Tan is smart to do it, too. He needs tens of billions of dollars to fix the problems with Intel’s semiconductor manufacturing business, and that was the whole point of the U.S. government investment of $8.9 billion… to restore U.S. semiconductor manufacturing for advanced semiconductors.
But what about NVIDIA? The related product revenue business will be insignificant for the next two or three years, and the reality is that NVIDIA would get almost all of that revenue anyway from its existing products and future roadmap. So why the $5 billion investment?
That’s easy.
Protection money.
Huang’s brilliant move was to pre-empt the U.S. government. We can’t forget that NVIDIA has roughly a 95% monopoly on high-end GPUs for artificial intelligence. NVIDIA throws its muscle around to keep it that way, and all of its products are manufactured by TSM, not in the U.S.
NVIDIA is now a $4.2 trillion company, the most valuable company in the world, and a monopoly.
Can you hear the antitrust bells ringing??? NVIDIA will generate almost $100 billion in free cash flow this year and 71% gross margins. There are many who believe NVIDIA abuses its position of power…
Which is why Huang’s move to offer $5 billion to “support” the Trump administration’s $8.9 billion investment is such a smart move. Do it proactively, without being asked to do so. Pre-empt claims of anti-trust violations.
Pay the piper and contribute to the national goals of restoring U.S. semiconductor manufacturing. Well done.
NVIDIA will continue to win. And AMD, as far as I’m concerned, is an incredible investment right now.
It will continue to eat into Intel’s market share, has products that are competitive to what NVIDIA offers, has guaranteed manufacturing capacity with TSM, and will be one of the biggest players in GPUs designed for inference – the running of AI applications.
Jeff
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
The Bleeding Edge is the only free newsletter that delivers daily insights and information from the high-tech world as well as topics and trends relevant to investments.
