AI Analyst @MilkRoadAI | Finding opportunities across AI, photonics, defense, space, and tech.
$1 to see my portfolio →
Joined June 2026
- Tweets2.9K
- Following260
- Followers36K
- Likes804
Pinned Tweet
$CBRS just reported an absolutely massive quarter, and I think there are a few things investors should be paying close attention to here.
The company reported core revenue of $230.9 million, growing 70% year over year and beating expectations of $215 million. But what's even more impressive is its cloud business, which grew a staggering 265% to $141.9 million. Meanwhile, hardware revenue declined 29%, showing just how quickly the company's revenue mix is shifting toward cloud services. And management is expecting that momentum to accelerate. Cerebras raised its full year revenue guidance to $940 million–$960 million, well above the $887 million analysts were expecting. For Q4 alone, management is guiding for $310 million–$330 million in revenue, compared to estimates of just $271 million. Cerebras now has $26.1 billion in remaining performance obligations, which represents more than 27 times the midpoint of its projected annual revenue.
That's an enormous amount of contracted future business relative to the company's current size, although it will take time and execution to convert those commitments into actual revenue. The infrastructure expansion is equally impressive. Cerebras has 700 MW of data center capacity under contract for delivery by the end of 2027. At the same time, its upcoming CS-4 system is promising up to 30x faster inference, while CS-5 is targeting 10,000 tokens per second per user. And profitability is moving in the right direction. Core operating margins were negative 14% this quarter, but management expects them to improve to between negative 1% and positive 1% in Q4. If they can deliver on that guidance, Cerebras could be approaching operating breakeven much faster than many investors anticipated.
Now, there are still risks. The company remains unprofitable, its GAAP gross margin is only 20%, and building out hundreds of megawatts of infrastructure requires significant capital. But I think the bigger picture is becoming increasingly clear. Demand for high performance AI inference is growing rapidly, and Cerebras is positioning itself as a serious alternative in a market NVIDIA has dominated for years. I'm not saying Cerebras is going to replace NVIDIA, but I do believe the AI compute market is going to be large enough to support multiple major winners. And if Cerebras can execute on its backlog, improve margins, and deliver on its next generation systems, this could become a very interesting company to watch over the next several years.
A Tesla and SpaceX merger could potentially create a $100T company over the long term, and as crazy as that sounds, I don't think it's completely out of the question if Elon Musk executes on his biggest ambitions.
Think about what a combined company would control. You have Tesla's autonomous driving, Optimus robots, energy storage, and manufacturing capabilities alongside SpaceX's Starlink network, rocket launches, Starship, and expanding AI ambitions. Add Terafab into the equation, and you're potentially looking at one company controlling everything from chip manufacturing and AI infrastructure to robotics, transportation, and space.
And a merger could allow Musk to consolidate resources, share technology across his businesses, and accelerate projects that would otherwise take years to develop independently. Now, $100 trillion is obviously an extremely ambitious, decades long scenario that would require massive breakthroughs in AI, robotics, autonomous transportation, and the space economy. It's certainly not what I would value the company at today. But if even a fraction of Musk's long term vision becomes reality, I think the upside could be significantly larger than most investors currently imagine.
🤝 Paid partnership
The Jensen Huang situation just keeps getting worse
There are now claims circulating that the former NVIDIA CEO had a THIRD affair, following reports of another relationship with an NVIDIA employee.
There is the nail in the coffin!
Samsung and SK Hynix are on track for the same HBM capacity in 2027 but they are not going to get the same results.
According to Hana Securities, both companies are projected to reach 260,000 HBM wafers per month in 2027.
Hana expects SK Hynix to ship more HBM from that capacity, about 2.3 billion GB versus about 2 billion GB for Samsung.
But here is the twist, hana actually expects Samsung to make MORE HBM revenue, mostly because of higher HBM4 pricing.
I think this tells you how tight this market really is, pricing power is now just as important as volume, and both Korean giants are set to print money in 2027.
ALT Hana Securities charts of Samsung and SK Hynix HBM wafer capacity and shipments
How can two of the biggest banks on Wall Street be THIS far apart on the same stock?
According to Korean media, Nomura has a price target of ₩4.7 million on SK Hynix. Morgan Stanley's target is ₩1.6 million.
Nomura thinks memory demand will keep outpacing supply through at least 2028. Morgan Stanley thinks new capacity will ease the shortage later in 2027.
The estimates for 2028 operating profit are even wilder. They range from ₩94 trillion to ₩691 trillion for SK Hynix, and from ₩122 trillion to ₩939 trillion for Samsung.
I think this is the single most important debate in semiconductors right now. If the bulls are right on supply, these stocks are still very cheap. $MU
ALT English table of Nomura and Morgan Stanley price targets for SK Hynix and Samsung
I actually think the biggest winners from Jensen Huang leaving NVIDIA could end up being the frontier AI labs, particularly OpenAI and Anthropic.
And the reason has a lot to do with DeepSeek, access to advanced AI chips, and the future of open source AI because Alice Huang, the woman at the center of Jensen's scandal, allegedly promised NVIDIA chips to DeepSeek. There are also serious questions surrounding whether advanced NVIDIA chips were being shipped into China through intermediaries to get around US export restrictions. The Information previously reported that DeepSeek had obtained thousands of restricted Blackwell chips through overseas data centers. As we all know Jensen has spent years advocating for broader access to AI computing power, including pushing to allow NVIDIA to sell more chips into China. He's also been one of the biggest supporters of open weight AI, with NVIDIA investing heavily in models, tools, and infrastructure that make it easier for developers to build AI systems without relying entirely on proprietary models.
But with Jensen gone and NVIDIA facing increased scrutiny, I wouldn't be surprised if we see much tighter restrictions on where these chips are shipped, who can purchase them, and even whether Chinese companies can access NVIDIA powered data centers overseas. In fact, Reuters is already reporting that Chinese AI companies are rushing to secure overseas computing capacity because they're worried about additional restrictions. And this could have enormous implications for open source AI. Companies like DeepSeek have shown that powerful open weight models can compete with proprietary systems while putting significant pressure on pricing. But developing those models still requires enormous amounts of computing power, and restricting access to the most advanced chips could make it substantially harder for Chinese labs to keep pace with OpenAI, Anthropic, and Google.
Obviously, open-source AI isn't going to disappear, and China is already accelerating its efforts to develop domestic alternatives through companies like Huawei. But if NVIDIA becomes less aggressive about supporting opencweight AI and the US tightens access to advanced chips, I think the competitive environment could shift significantly in favor of the American frontier labs.
🤝 Paid partnership
I genuinely believe people are still massively underestimating just how big $NBIS could become over the next several years.
According to estimates from William Blair, Nebius is projected to grow its annualized revenue run rate from $1.25 billion in 2025 to an astonishing $71.7 billion by 2030. That's nearly 57x growth in just five years, which is absolutely insane when you consider the amount of infrastructure, capital, and computing capacity required to support that kind of expansion. What's even more interesting is the projected growth trajectory. Analysts are forecasting 540% growth in 2026, followed by 109% in 2027, 79% in 2028, and nearly 70% in 2029. Even by 2030, when Nebius could theoretically be generating over $70 billion in annualized revenue, growth is still expected to exceed 40%.
What makes Nebius even more interesting to me is one of its exposure to ClickHouse, which has been seeing some incredible growth of its own. For those unfamiliar, ClickHouse is a rapidly growing database company specializing in real time analytics, allowing businesses to process enormous amounts of data incredibly quickly. Nebius owns a significant minority stake in the company, which raised $400 million earlier this year at a $15 billion valuation. Interestingly, recent private secondary market pricing implies a valuation of approximately $56.6 billion for ClickHouse, which would value Nebius's 25% ownership stake at roughly $14 billion alone. That's an enormous amount of potential value tied to just one of Nebius's investments.
Think about what that means. While Nebius shares have been getting absolutely crushed in the public markets, one of its most valuable private investments has potentially appreciated significantly. And with the enormous financing requirements associated with building AI data centers, I think having an asset like ClickHouse could provide Nebius with additional financial flexibility down the road. This is why I remain incredibly bullish on Nebius over the next several years. There are certainly risks, and I don't expect the road ahead to be smooth, but if the company can execute even a meaningful portion of what analysts are forecasting, I believe the long term upside could be substantial. If you want to see how I’m investing around $NBIS and the broader AI infrastructure buildout, check out my full portfolio below.
link.milkroad.com/tx5qip
While the market has been selling, I've been buying.
You can see exactly what I've been buying, selling and holding, along with my full portfolio for just $1. milkroad.com/pro/?utm_medium…
980
I am genuinely torn on the whole Jensen Huang situation right now.
On one side, I really wish Jensen had stayed as CEO of NVIDIA because I genuinely believe there is no better advocate for AI than him. This man has spent decades building NVIDIA into one of the most important companies in the world, and nobody has done a better job of explaining the potential of AI. He’s the face of the AI revolution, and I honestly don’t think there’s anyone who can replace what he brings to the industry. Losing someone like him at such a critical moment is a massive blow, not just for NVIDIA, but for the entire AI ecosystem.
But on the other side, we’re talking about a CEO facing serious questions over his personal conduct and judgment. It’s incredibly disappointing to see someone who has accomplished so much put his reputation and the company he spent decades building in this position. And as much as I admire what Jensen has accomplished, being one of the greatest CEOs of our generation doesn’t mean you should be immune from accountability.
And that’s what makes this situation so difficult for me. I can disagree with someone’s personal decisions while still recognizing just how important they have been to an entire industry. At the end of the day, Jensen’s resignation doesn’t change the demand for AI, the need for computing infrastructure, or NVIDIA’s technological leadership. But I do think the industry just lost its strongest voice at a time when it needed him the most.
I’m genuinely conflicted about this whole situation. I wish he had stayed, but I also understand why his position became increasingly difficult to defend. Curious where everyone else stands on this.
$NBIS just delivered one of its strongest quarters ever, yet the stock is getting crushed. And I think the reason behind the selloff is actually more interesting than the earnings themselves.
First, let’s look at the numbers. Revenue surged 571% year over year to $980.8 million, comfortably beating expectations of $908 million. Adjusted EBITDA came in at $465.8 million versus estimates of $353 million, while revenue grew another 68% from the previous quarter. Annualized recurring revenue reached $6.4 billion, and the company ended September with $7.4 billion in cash.
What’s even more impressive is what’s happening with demand. Nebius has secured more than $12 billion in new contracts since the beginning of Q3, while GPU pricing increased by over 20% sequentially. Revenue contracted per megawatt also climbed from roughly $20 million to more than $25 million, and estimated GPU payback periods improved from 22 months to 19 months. Customers are willing to pay substantially more to secure AI computing capacity, while Nebius is recovering its infrastructure investments faster. That’s a pretty encouraging development for the entire neocloud industry, especially considering how much money these companies are spending on GPUs and data centers.
But here’s where the problems begin.
Despite beating earnings expectations, Nebius lowered its 2026 revenue outlook toward the bottom of its previous $3 billion to $3.4 billion range. Some of the computing capacity intended for Microsoft has been delayed, pushing expected revenue into 2027. The issue appears to involve power infrastructure and permitting problems at facilities operated by third-party partner DataOne. Temporary generators were taken offline before enough replacement fuel cell capacity was available, creating delays in bringing additional computing capacity online.
And this is exactly why investors are concerned. Nebius has the customers, the contracts, and the demand, but it’s having trouble delivering the infrastructure quickly enough to recognize that revenue. When you’re growing this aggressively, even relatively small construction or power-related setbacks can have a substantial impact on financial projections. To make matters worse, the company increased its 2026 capital expenditure guidance to $27 billion, compared with its previous range of $20 billion to $25 billion. So Nebius is now expecting to spend more money while generating less near-term revenue than initially anticipated. That’s not exactly the combination investors want to see.
Looking ahead, management expects revenue to reach $12 billion to $13 billion in 2027, representing roughly four times its anticipated 2026 revenue. However, getting there will require another $40 billion to $45 billion in capital expenditures. That’s an extraordinary amount of money for a company currently generating less than $1 billion in quarterly revenue.
One encouraging development is how Nebius plans to fund this expansion. Customer prepayments covered the associated capital expenditures for its Q3 deals, compared with approximately 50% to 60% in Q2. Management also expects customer prepayments and financing secured against existing contracts to cover around 75% to 85% of its 2027 spending.
This could substantially reduce the amount of additional capital Nebius needs to raise through traditional financing, although it doesn’t eliminate debt obligations or the risk of shareholder dilution. Management is also targeting pre-tax breakeven by the end of 2027, but investors should remember that accounting profitability and positive free cash flow are two very different things when a company is spending tens of billions of dollars annually.
Personally, I think there’s a lot to like about these earnings, particularly the acceleration in revenue, improving GPU economics, and the willingness of customers to commit substantial capital upfront. But I also think the market is sending an important message to the entire AI infrastructure industry. Signing multibillion-dollar contracts is one thing. Actually delivering the computing capacity on time, managing the financing, and turning those contracts into profitable revenue is something entirely different.
Nebius has demonstrated that demand for its infrastructure is incredibly strong. Now it needs to prove that it can execute on its aggressive expansion plans without constantly running into delays and escalating costs. For long-term investors, I don’t think these results fundamentally change the AI infrastructure demand story. But they do show why execution, financing, and access to reliable power are becoming some of the biggest risks facing the neocloud industry.
I genuinely believe people are still massively underestimating just how big $NBIS could become over the next several years.
According to estimates from William Blair, Nebius is projected to grow its annualized revenue run rate from $1.25 billion in 2025 to an astonishing $71.7 billion by 2030. That's nearly 57x growth in just five years, which is absolutely insane when you consider the amount of infrastructure, capital, and computing capacity required to support that kind of expansion. What's even more interesting is the projected growth trajectory. Analysts are forecasting 540% growth in 2026, followed by 109% in 2027, 79% in 2028, and nearly 70% in 2029. Even by 2030, when Nebius could theoretically be generating over $70 billion in annualized revenue, growth is still expected to exceed 40%.
What makes Nebius even more interesting to me is one of its exposure to ClickHouse, which has been seeing some incredible growth of its own. For those unfamiliar, ClickHouse is a rapidly growing database company specializing in real time analytics, allowing businesses to process enormous amounts of data incredibly quickly. Nebius owns a significant minority stake in the company, which raised $400 million earlier this year at a $15 billion valuation. Interestingly, recent private secondary market pricing implies a valuation of approximately $56.6 billion for ClickHouse, which would value Nebius's 25% ownership stake at roughly $14 billion alone. That's an enormous amount of potential value tied to just one of Nebius's investments.
Think about what that means. While Nebius shares have been getting absolutely crushed in the public markets, one of its most valuable private investments has potentially appreciated significantly. And with the enormous financing requirements associated with building AI data centers, I think having an asset like ClickHouse could provide Nebius with additional financial flexibility down the road. This is why I remain incredibly bullish on Nebius over the next several years. There are certainly risks, and I don't expect the road ahead to be smooth, but if the company can execute even a meaningful portion of what analysts are forecasting, I believe the long term upside could be substantial. If you want to see how I’m investing around $NBIS and the broader AI infrastructure buildout, check out my full portfolio below.
link.milkroad.com/tx5qip
While the market has been selling, I've been buying.
You can see exactly what I've been buying, selling and holding, along with my full portfolio for just $1. milkroad.com/pro/?utm_medium…
3
2,317
Get ready for an absolute TON of AI FUD over the next two days, especially today, following Jensen Huang's departure.
Yes, we lost arguably the strongest advocate for AI and one of the greatest CEOs of our generation, but that doesn't mean the AI revolution suddenly comes to an end.
Nothing has fundamentally changed. The demand for compute is still exploding, hyperscalers are still committing hundreds of billions of dollars to infrastructure, and the race to build out AI capacity is showing no signs of slowing down.
Jensen played an enormous role in getting us here, and replacing his leadership won't be easy but the entire AI industry is much bigger than any one individual.
Don't confuse leadership uncertainty with a collapse in the AI investment thesis.
Imagine if Jensen Huang actually joins Elon Musk at SpaceX to help build Terafab.
We're talking about the man who turned NVIDIA into the most important company in the AI revolution teaming up with Elon to build one of the most ambitious semiconductor manufacturing projects in history.
Jensen understands AI chips better than almost anyone, and Elon has the ambition and resources to push manufacturing to an entirely different level.
If these two actually team up, the implications for the entire semiconductor industry could be absolutely massive.
🤝 Paid partnership
Remember the report that $MU was freezing its $4.5 billion memory fab in Taiwan over a labor dispute? It looks like that scare is fading.
According to Commercial Times, Taoyuan's mayor said he spoke directly with Micron's top leadership and was told construction of the new fab will keep moving forward.
He said Micron told him Taiwan, and Taoyuan in particular, is one of its most important locations in the world.
Micron has also put what it calls its best bonus plan ever on the table for workers, and the next round of talks with the union is set for November 17.
This matters because Fab 11E was meant to start producing in 2028, right when the memory shortage is expected to be at its worst.
I think any delay in new memory capacity is bullish for prices, but Micron keeping this fab on schedule is even more bullish for Micron itself.
ALT Micron memory fab construction image
China's massive memory chip expansion could become one of the biggest threats to the global DRAM market over the next few years.
Charles Kau, widely known as the Godfather of DRAM, recently weighed in on China's memory ambitions, and some of his comments are particularly interesting. According to Kau, China's CXMT currently has 10 semiconductor fabs under construction across Beijing, Shanghai, and other locations, with capacity expected to ramp up over the next two years. That's a massive amount of additional supply potentially entering a market that is already experiencing tremendous demand from AI infrastructure, data centers, and high bandwidth memory.
However, there's an important distinction here. Building more fabs doesn't automatically mean China can compete with Samsung, SK Hynix, and Micron at the technological frontier. Kau specifically pointed to the enormous technical challenges associated with advancing DRAM technology, including capacitor scaling and leakage issues. Even established manufacturers are still working through these limitations. Interestingly, he also suggested that TSMC entering the DRAM market is unlikely, arguing that being incredibly good at manufacturing advanced logic chips doesn't necessarily translate into having the expertise required to manufacture competitive DRAM. Meanwhile, China's progress in NAND flash is an entirely different story. YMTC has already established itself as a serious technological competitor, demonstrating how quickly Chinese manufacturers can close the gap when they successfully overcome manufacturing challenges.
My biggest takeaway is that the long term memory story is becoming increasingly complicated. I'm still extremely bullish on memory demand over the next several years, particularly with the amount of AI infrastructure being built. But there's a difference between being bullish on demand and assuming memory prices will continue rising indefinitely. If China brings enough additional capacity online, we could eventually see significant pricing pressure in conventional DRAM, even while advanced products like HBM remain supply constrained. Still bullish on memory but I’m watching China’s capacity expansion closely and if you want to see how I’m investing around the memory cycle while managing that risk, check out my portfolio below.
link.milkroad.com/tx5qip
BREAKING: NVIDIA has officially confirmed that Jensen Huang has resigned as CEO, effective immediately.
NVIDIA CFO Colette Kress has been appointed interim CEO, while the company begins searching for a permanent replacement, considering both internal and external candidates. Huang has also stepped down from NVIDIA's board.
This is an enormous leadership change for the AI industry. Jensen wasn't just NVIDIA's CEO. He co-founded the company, helped build its CUDA ecosystem, and played a central role in its transformation into the world's dominant AI chipmaker.
The biggest question now is whether NVIDIA can maintain its technological leadership, customer relationships, and execution without the person who helped build all three.
The AI infrastructure demand hasn't suddenly disappeared, but NVIDIA has just lost one of its greatest competitive advantages: Jensen Huang himself.
According to Korean media, SK Hynix just pushed its HBM4 production yield above 70%.
This is a BIG deal because HBM4 is the hardest memory chip ever made because it stacks multiple DRAM dies on top of each other and doubles the interface to 2,048 bits versus HBM3E.
Going from 60% to 70% yield means about 16.7% more good chips from the exact same production input and that is basically free capacity without building a single new fab.
SK Hynix already said on its October earnings call that HBM4 yields beat its plan and that it raised internal targets and it also expects HBM4 to pass HBM3E in both shipments and revenue this quarter.
Samsung is reportedly near 80%, but each company measures yield differently, so I would not read too much into the head to head.
I think the bigger takeaway is simple and the supply of the most important AI memory chip is about to get more efficient right as demand keeps exploding, and that is VERY bullish for $SKHY, Samsung and $MU margins.
ALT SK Hynix HBM4 memory chip product photo
Everyone keeps asking when the memory cycle ends and one of Korea's biggest brokers just said not anytime soon.
According to Meritz Securities, DRAM prices could keep rising all the way through the fourth quarter of 2027.
It expects global chip equipment spending to hit a record of about $200 billion in 2027, with combined capex at Samsung and SK Hynix up 28%.
The key argument is simple. When memory got expensive in the PC and smartphone era, people stopped buying. In the AI era, hyperscalers keep buying anyway because they cannot afford to fall behind.
The risk is 2028, when all this new capacity comes online but for now, I think memory pricing power is MUCH stickier than the market believes. $MU
ALT Weekly StockCharts chart of Micron Technology MU
Some $NVDA board members are reportedly backing CFO Colette Kress as a potential interim CEO if Jensen Huang is forced to step down, according to Semafor.
For those unfamiliar with Kress, she’s been NVIDIA’s CFO since 2013 and has more than three decades of experience in the technology industry. Before NVIDIA, she spent 13 years at Microsoft, including leading finance for its Server and Tools division, and later held a senior finance leadership position at Cisco. She also previously worked at Texas Instruments.
What’s particularly interesting is that Kress has been at NVIDIA throughout its transformation from a gaming GPU company into the dominant force in AI computing. She’s helped oversee the company’s financial strategy during one of the largest growth periods in technology history and has extensive experience communicating with Wall Street.
I can understand why some directors would favor her as an interim replacement. She knows the business, understands NVIDIA’s customers and financial operations, and would likely provide some stability during a very uncertain period. But I still believe replacing Jensen would be an enormous mistake. Kress is an incredibly experienced executive, but managing NVIDIA’s finances and leading the company’s technological vision are two completely different responsibilities.
Personally, I’d much rather see Jensen remain CEO, or at the very least retain a major leadership role as chairman, than watch NVIDIA lose the person who helped build its entire AI ecosystem.
According to Taiwanese media, $NVDA may be about to cut chip orders, and I think that could be a HUGE opening for $AMD and the ASIC camp.
Supply chain sources told the Commercial Times that Nvidia wants to slow its chip orders after tighter export reviews and a pileup of H200 processors in Hong Kong. $TSM and ASE would be first in line.
If Nvidia frees up foundry capacity, ASIC firms say customers will take every wafer. Analysts say $AMD, $AVGO, MediaTek and Alchip would be among the biggest winners as cloud companies spread their chip buying around.
MediaTek is targeting about $16 BILLION from its first AI accelerator in 2027 and has already signed a second customer while Alchip has also landed a second ASIC order from a major United States customer.
I don't think people realize how ready this camp already is.
$AMD demand is still running ahead of supply, Instinct MI455X is shipping in volume, and Lisa Su said the first Helios racks are now live.
Nvidia has not confirmed any cuts but if even some capacity gets released, I think AMD and the ASIC names are waiting to take it.
ALT AMD CEO Lisa Su
Anthropic is now using Claude to help design its OWN AI chips.
According to a report from The Information, Anthropic is pitching this to IPO investors as a new strength of its models.
The company has a chip team of just 15 people, and that team says Claude is already speeding up its engineering and design work.
The chips would be used to train and run Claude, and Anthropic has not announced any plans to sell the design tools to anyone else.
This matters because compute is by far Anthropic's BIGGEST cost and its amended IPO filing lists about $895 billion of commitments, and nearly all of it is for computing and data centers.
Broadcom is the largest name on that list with about $161 billion of equipment leases, followed by Google at about $121 billion and Amazon at $110 billion.
Right now Anthropic runs Claude on Google TPUs, Amazon Trainium, Nvidia GPUs and AMD GPUs, and it treats all of them as one shared pool of compute.
A custom chip would give Anthropic more control over that bill, and using its own AI to design it could make the chip cheaper and faster to build.
If a team of 15 people can do the work of a much bigger chip team with Claude's help, that is a big signal for the entire semiconductor design industry.
$AVGO $GOOGL $AMZN $NVDA $SNPS $CDNS
While the market has been selling, I've been buying.
You can see exactly what I've been buying, selling and holding, along with my full portfolio for just $1. milkroad.com/pro/?utm_medium…
1
2
1,337
NVIDIA’s board is reportedly meeting right now to discuss potentially replacing CEO Jensen Huang.
President Trump has reportedly said he will not interfere with the board’s decision, leaving NVIDIA’s future leadership entirely up to the company.
If Jensen actually gets pushed out, I think this would be one of the biggest mistakes NVIDIA could make. He’s not just another CEO as he spent decades building NVIDIA into the company that is now leading the entire AI revolution.
And with AI infrastructure spending expected to reach trillions of dollars over the coming years, losing the person responsible for positioning NVIDIA at the center of that growth would create enormous uncertainty for investors.
Personally, I think Jensen should remain involved with NVIDIA, even if that means transitioning into a chairman role.
If NVIDIA’s board actually decides to replace Jensen Huang as CEO, I think the smartest move would be to make him chairman rather than push him out of the company entirely.
Jensen has spent decades building NVIDIA into one of the most important companies in the world, and I genuinely believe the AI revolution would not be where it is today without him.
Making him chairman would allow NVIDIA to bring in new leadership while keeping Jensen involved in the company’s long term strategy and direction.
The last thing NVIDIA needs right now is to lose the person who helped build its entire AI empire.
Melvin retweeted
If Jensen Huang ever resigns as CEO of NVIDIA, I will seriously consider dumping my entire $NVDA position.
I genuinely believe Jensen is one of the most irreplaceable CEOs in the world, and I don't think people fully appreciate just how much of NVIDIA's success comes down to his leadership.
We're talking about someone who saw the AI revolution coming long before almost anyone else and spent decades positioning NVIDIA to become the company powering it.
His vision, relationships with hyperscalers, understanding of the semiconductor industry, and ability to execute are incredibly difficult to replicate.
I genuinely believe the AI revolution would look completely different without Jensen Huang. And while NVIDIA would obviously continue operating without him, I would have a very hard time maintaining the same level of conviction in the company.
For me, investing in NVIDIA has always been just as much about betting on Jensen Huang as it has been about betting on AI.
🤝 Paid partnership
