Does Nvidia Take the Money... Or Wait to See What's Behind the Curtain?
It seems everybody has an opinion on what Nvidia's (NVDA) latest blowout earnings and guidance means...
But to me, the bigger question boils down to: Is the euphoria surrounding Nvidia?really?signaling the dawn on the Internet part deux? Or is it a sign of the market's last hurrah before a blow-off top?
As somebody who raised a red flag or two (or three) in the mid-to-late 1990s – based on historic metrics, logic, and common sense... all suggesting that today's bubble should burst?any day?– don't look at me for an answer. Trying to get this one right is like trying to guess the number of gumballs in a giant jar...?
What I?do?know is that artificial intelligence ("AI") is real... so real that it was here long before ChatGPT arrived on the scene.
Nvidia, in fact, has been talking about AI since at least 2011... back then, for the way it enhanced the realistic nature of video games.
Now, AI has become a quiet force in everybody's everyday life... mostly without fanfare. It's what makes the camera on your phone so good, and your Ring doorbell capable of determining whether it just saw a person or a dog. It's why you know how much traffic you're about to face, what the best detour will be, and why you can get an alert that "light rain" will start in five minutes.
The rollout of ChatGPT changed the narrative in a huge way...
It was obvious in Nvidia's most recent earnings report, which was so good that it sparked a whodunit of sorts about the?real?meaning of such strong demand so early in the game.
It's one thing for Nvidia CEO Jensen Huang to declare as he did during a speech over the weekend in Hong Kong, "We have reached the tipping point of a new computing era."
It's another to know when it will actually tip... or something simpler, like whether the current wave of orders was a one-time event.
领英推荐
While the company didn't come right out and say so, it's hard to believe that most if not all of last quarter's demand came from anyone other than?the current kingpins of generative AI: Microsoft (MSFT), Alphabet (GOOGL), Meta (META), and Amazon (AMZN)... each jockeying to make sure they don't get left behind.
Such so-called "pull-forward" of demand is fine, as long as there's a greater amount of demand right behind it.
And that's the wildcard here...
As one friend who is no neophyte when it comes to tech companies put it in a note to friends...
Four companies, but primarily two, can create the perception of something much broader going on if you look at the results of their much smaller suppliers.
Interestingly, this just may be a big one-time lift, that could roll back over once whatever they are doing is complete, or just be hard to anniversary and grow from. The bulk of the orders may have already been placed.
Importantly, he adds...
It is impossible to know. Does not mean that the suppliers can't get another bounce up (especially if the channel reloads on their other stuff), or that it can't sustain at these levels for a while. It just means nobody knows...?including NVDA.
That's the important point here: Even Nvidia doesn't how all of this will shake out.
That's where this gets even more interesting... (To read the rest please click here.)
I help Risk & Resilience Managers build unique knowledge to become a top 1% Resilience Engineer, with innovative but proven Resilience Strategies | Master Risk, Resilience, Antifragility & Complexity
1 年Herb Greenberg intersting read. The semiconductor industry is definitely an interesting one. But it has a long-run way with upfront investments and is cyclical. If you are interested, I just started a series delving into the shortages 2020-2022. Here is the link: https://www.dhirubhai.net/posts/marco-felsberger-resilience-and-risk-expert_resilience-shortage-procurement-activity-7069583281481990144-Kqaf?utm_source=share&utm_medium=member_desktop
Lead Product Designer @ Makai Labs
1 年NVDA moving into the CPU market utilizing ARM architecture (as opposed to traditional x86) is definitely an interesting move that I was not expecting.. wonder how AMD and INTC will answer
Product Design and Development | Emerging Tech | A.I., NLP and Machine Learning | Researcher | Startups
1 年The AI boom has zero to do with Generative AI. That's a toy at best that writes poems and makes/edits photographs. The last thing our human brains need is another "Cambrian" explosion of data. We can't even figure out what to watch on Netflix tonight! MD Anderson is in a meeting right now talking about how to use neural networks to cure cancer. They are tipping their hat to Hinton for backpropagation, OpenAI for adding more inputs and hidden layers and NVIDIA for making gaming cards -- but that's the extent of their appreciation. I think NVIDIA is way overpriced and many startups will come up with new designs for mass parallelization. But the idea of 3-4 companies being the only tensor core buyers in town is a nonstarter, IMO. Cloud is for ongoing operations. AWS offers an 8 GPU instance. Anderson needs 160 GPUs, if not more. And they need them for just the training. Most large caps will never need a UI (chat window) for their models. Curing cancer is not Generative AI. But it uses GPUs, NNs with a lot of inputs, and NNs with a lot of layers. FYI Jason McMunn
首席財務官-国际的 |獨立董事 |審計委員會主席 |私募股權 |上市公司 |風險管理 |複雜訴訟
1 年Apparently, they will "TAKE THE MONEY"? (source not verified; I'm busy building infra to house this stuff)
In Action
1 年Past 4Q revenues of AMD + NVDA + INTC is constant suggesting market share take over for NVDA. Even then NVDA is less than half of annual rev with AMD and INTC combined. I think wall street does not understand semis. Or AI is just clouding it’s brains. Or just some hype creators…. On a different note Elon just entered a viper pit and praising the critters. Wait till bitten? No idea