Everything is about to happen. But until that happens, you shouldn't wait for it to make money. You have to see its success to invest in it. If not, what's the point? There. Until May 2023, this was a bearish case for all the big nuisances, including artificial intelligence, generative AI, accelerated computing, and large-scale language models (LLM). At that time, Nvidia and CEO Jensen Huang said he “missed out” on $4 billion. That's right, he reported $4 billion more in revenue than they expected. I say “nostalgic” because Jensen likes to tell stories that way. When Jensen told the same joke, his wife was a little confused as to how it could be a “miss”, but I know he gets a laugh out of me every time. But I like this metaphor. Because until that quarter, you had to be in on the “joke” so to speak, and you had to be in on how Nvidia's chips were being used by all the Titans. “Hyperscaler”, this is a term given to me by another favorite of mine, Jayshree Ural. She is the CEO of Arista Networks, the preferred network solution Titans, which includes the club names Amazon, Alphabet, Microsoft, and Meta Platforms. No one laughs about Jayshree's company either. Because Jayshree's white box solution beats Cisco Systems when it comes to the internet plumbing that connects Nvidia chips to Titans. Nvidia had just shed its identity as a gaming chip company and wasn't known as an acceleration chip company or a company at the core of generative AI with chips that train and give it inference power. It's a wonder of the reasoning behind all the systems that allow you to type questions into ChatGPT (OpenAI, backed by Microsoft), Claude (Anthropic, backed by Amazon), and Gemini (Google, by Alphabet). It's an enhancement to the AI assistant you're currently using, but when Microsoft introduces it to your personal computer, it's set to be the biggest overhaul in PC history. That's why we own Best Buy, even though it won't push up the stock price until late. autumn. But don't get ahead of yourself. This time last year, NVIDIA snuck into the US $1 trillion market cap club in its May quarter. It took 24 years. It took 270 days for the stock market value to reach $1 trillion. Now it's knocking on his 3 trillion door. At the top of this scale, there are only club names Apple and Microsoft. But beyond Nvidia's $3 trillion milestone lies an earnings hurdle Wednesday night. It has more going on than Nvidia because so many people have yet to see what the company's chips can do. At the moment, we're not sure what ads you see when you order something, how robots move, small but productive changes in call centers, and how much better they are than humans. It appears that Nvidia is behind his request to retailers who can answer. Machines will become doctors and give us emotional answers. In short, this is the genesis behind something between a parlor game (Teach me a haiku about my love of baseball on ChatGPT) and something that could potentially make people more productive. Wednesday feels like a very looming day because, frankly, everything seems like such small potatoes. Last year it was about “beat”. Expectations are so high this year that everyone is worried that Barr has finally caught up with the analysts who were valuing it high that it couldn't deliver the magnitude it did last year. Beat and raise seems impossible, but even if it were possible, it wouldn't be next time. So we've moved from a stage of “it's about to happen” to a stage of “we'd better see more success.” Because not only isn't it worth $3 trillion in market cap right now, it might not even be worth $2 trillion if the Titans were all together. With their chips in hand, the zeitgeist of Nvidia's orbit is so hotly contested that Advanced Micro Devices (AMD) is about to explode onto the scene, making them the ultimate “don't trade, own” believers. Even I dread that moment right after Wednesday when the closing bell rings at 4pm ET. What can Nvidia do about Encore? Isn't that what everyone is thinking? But what if the answer is that we'll actually see Blackwell's first results, this time a master machine that's much faster than Nvidia's current version? Until this Wednesday, most people know that Nvidia's 286 and 386 chip, and I believe that Intel is like the 1990s when the 486 chip continued to exceed expectations. Because each new version was so powerful. And at the same time, Intel at the time was able to master a transition that CEO Andy Grove didn't think anyone could master. This transition includes months that had a tail end on the old chip and a front end on the new chip. Why buy a 386 when a 486 is right around the corner? For some reason, the transition made no difference to my bottom line. Thirty years later, how it was actually removed remains a mystery. What does Nvidia actually have to do for Encore? Simple. there is nothing. You have to come up with a completely new shot. When I scream for an encore at an opera, I just wish for one more aria. Blackwell is more of a supercomputer, but not Arya. This is a new way of computing, and it happens when it's much faster than the high-speed computing that Nvidia currently offers. Unlike Intel's 386 at the time, the 486 won't become obsolete. Old things are even better when layered with old things. But the new one is so innovative and so expensive that Nvidia's revenue projections could make it worthy of $3 trillion status. Analysts will be puzzled if each Blackwell costs $40,000 and NVIDIA can produce huge numbers of Blackwells, and Blackwells are much smarter than, say, the H100, but yet another huge new Will recognize the beat and raise the scenario from this problem. -It's a very small company. I think this is a much bigger problem than what Intel faced. Not because the transition was difficult — Mr. Jensen was convinced the company's H100 was still valuable — but because the mortals who decide stock prices can't understand what's going on. Blackwell can do it. For example, what does it mean if a Blackwell-led platform (not a chip) can capture video? What does it mean if it can speed up research from years to weeks? Will it help cure ALS, as the team EverythingALS, which I introduced on Friday's “Mad Money,” hopes to? When I say help cure, I mean… The idea is to study all the genetic material that can be harvested and be able to modify it so that people don't die from ALS at first, and then when it becomes chronic, they can be cured. What is Merck's path forward with the cancer drug Keytruda? Is that what you call speed? Is that something that can be made with this machine? I have two tasks for Wednesday. We need to see the results of the current iteration, it has to be something dramatically special, and we have to be able to recognize the consequences of what Blackwell can do. It's the same one-two punch that Apple challenged the iPhone, if not dwarfed by the Vision Pro. Well, that can be difficult. With that in mind, let's introduce Nvidia's “don't replace what's yours” case. First, in the next iteration, the robot must be able to imitate everything that is currently moving, including the movements made by humans. In short, the best example is a robot bartender who, thanks to training and reasoning, can take drink orders, serve food better than humans, and virtually empathize with patrons. Secondly, you have to think that such creations are not tricks, but things that you can immediately use to flip burgers or sell pensions. Third, we need to stay away from anything AMD makes, or anything Amazon, Google, or Microsoft makes. So you can see that these are not competitors, but just enhancements to its own systems that allow Nvidia to make its chips work within those customers' ecosystems. Too many questions to ask? Here's a look at what Jensen is doing to exceed expectations. No, he doesn't care about his stock as much as he cares about Grove. The best business book of all time, Only the Paranoid Survive, makes all other books seem ridiculously sleepy. But he doesn't care about hitting, otherwise he wouldn't have been so proud of what happened last May. What if Jensen can't pull it off? Here's Wednesday's virtue. Through NVIDIA CFO Collette Kress, he may be able to talk about what Blackwell could do in the coming years as well as chips known as supercomputers. Perhaps after explaining what Nvidia can do that's very different from the current iteration, you can imagine what they're capable of. It's a tall order, but NVIDIA is a company that routinely wins big-ticket orders, and we never knew that until May of last year. Therefore, we should expect much more than a simple beat-and-raise, to be realistic and down-to-earth in his one earnings report. And I'm sure we can understand that, or at least understand such a vision of the future. animal. This could cause the stock price to rise further. Is there anything less? We're going to be sold, and we're going to have some big stock sales, probably General Mills, Union Pacific, club stocks like Eli Lilly and GE Vernova, which could move us as far away from technology as possible. Everything else becomes a safe haven from the tech sector, inflated by the value of Nvidia, and really nothing, including Microsoft, Alphabet, and Apple (my other thought is “Own it, Please do not trade stocks.” Sales, Earnings, and Forecast Lines to Tape Within 30 Minutes of Wednesday's Close on Wall Street (For a complete list of Jim Cramer Charitable Trust stocks, see here.) CNBC Investing Subscribers As. Club with Jim Cramer, receive trade alerts before Jim makes a trade. If Jim talks about stocks on his CNBC TV, Jim will send you a trade alert and then he will wait 45 minutes before buying or selling stocks in the charitable trust's portfolio. After issuing a trade alert, wait 72 hours before executing the trade. The above investment club information is subject to our Terms of Use and Privacy Policy, along with our disclaimer. No fiduciary duties or obligations exist or arise from your receipt of information provided in connection with the Investment Club. No specific results or benefits are guaranteed.
Nvidia founder and CEO Jensen Huang displays products on stage at the annual Nvidia GTC conference at the SAP Center on March 18, 2024 in San Jose, California.
Josh Edelson | AFP | Getty Images
Everything is about to happen. But until that happens, you shouldn't wait for it to make money. You have to see its success to invest in it. If not, what's the point?
There. Until May 2023, this was a bearish case for all the big nuisances, including artificial intelligence, generative AI, accelerated computing, and large-scale language models (LLM).