Last week, Microsoft and Apple’s exclusive two-trillion-dollar club embraced its newest member – Nvidia.
Nvidia reported a blowout quarter. Its revenue more than tripled over the past year, hitting about $22 billion. Its earnings per share shot up by 765%.
It took Nvidia about two decades to hit a market cap of $1 trillion.
It took just eight months to bag that second trillion.
That’s because demand for chips continues to exceed supply. And that’s because semiconductor chips are essential to the artificial intelligence (AI) revolution.
As Nvidia CEO Jensen Huang said: “Demand is surging worldwide across companies, industries, and nations.”
This is a dream scenario for chip designers and suppliers like Nvidia. And the government has catalyzed some of that demand.
Let’s dig into it some more.
The Government Angle
Now, there’s an abundance of dysfunctionality in Washington. No surprise there.
But I’ve spent more than two decades in hundreds of private meetings on the Hill. I’ve spoken with Congress, the Senate, the Securities and Exchange Commission (SEC), and the Fed.
And I’ve never seen such limited communication and efficiency.
Last week, I was speaking to Congressman Bill Foster, and he concurred.
Case in point: He was in his home office in Illinois. That’s because Congress was on a 12-day winter break despite a budget crisis looming.
Disfunction aside, Foster is part of a bipartisan AI committee created by the White House’s AI initiative last fall.
He and I discussed U.S. infrastructure needs – from upgrading power grids to AI-based cybersecurity.
That’s because AI technology is essential for nuclear and other energy efficiency, cyber-fraud prevention, and national security.
There was one bipartisan act that opened the path for domestic chip production: the CHIPS and Science Act.
The CHIPS Act became law on August 9, 2022. It allocates $52 billion to the domestic chip industry.
It might also not surprise you that almost all this money has yet to be deployed.
I’ll explain why that’s important for you and your money in a moment. First, let me unpack what the CHIPS Act is about.
The America Fund
The CHIPS Act’s core priority was to set up something called the America Fund. That fund would stimulate domestic semiconductor production.
The U.S. is known for its ability to design chips but not manufacture them.
Nvidia headquarters is in California. However, its largest manufacturing partner is TSMC, which is in Taiwan.
Taiwan manufactures 22% of global chips. South Korea produces 28% of them. China and Japan manufacture 28% in total.
The U.S. wants to change that, and the place to start is with AI hardware production.
AI applications rely on two types of chips.
The first type is Graphics Processing Units (GPUs). GPUs excel in what we call “parallel processing.” That means they can handle complex math computations and process large data sets at the same time.
Nvidia focuses on designing GPUs. Those are the chips that enable AI processes to run quickly.
The second type of chips are Central Processing Units (CPUs). CPUs focus on quick “sequential processing.” That means they handle tasks or instructions in a specific order.
Both GPUs and CPUs are essential for the optimal performance of AI applications.
And that’s why the CHIPS Act is significant.
CHIPS 2.0?
The CHIPS Act would facilitate more development for both types of CHIPS in the U.S.
See, the U.S. wants to claw back some of this production as a matter of national security.
That’s why, shadowed by the Nvidia news last week, the White House authorized $1.5 billion for U.S. chip manufacturing firm GlobalFoundries.
That money would help GlobalFoundries expand its domestic production capabilities in New York.
It’s part of the $39 billion from the CHIPS Act allocated toward domestic production. Plus, New York will invest another $575 million alongside that federal money.
And I’m hearing talk of another marquee act brewing from my D.C. sources.
Last week, at an Intel foundry event, U.S. Secretary of Commerce Gina Raimondo said more than the CHIPS Act was needed for the U.S. to secure a lead position in the semiconductor supply chain arena.
But what she said after that was more interesting to me. She said we need to…
…diversify our semiconductor supply chains and have much more manufacturing in the United States, particularly leading-edge chips, which will be essential for AI.
And she said…
I suspect there will have to be, what do you call it, CHIPS Two or some sort of continued investment if we want to lead the world.
AI and Semiconductors’ Co-Dependence
The global semiconductor market has been growing steadily over the past few years.
It had a market size of $528 billion in 2021. It is projected to grow to $1.4 trillion by 2029 to keep up with the AI revolution.
One way AI and semiconductors are integrated is through neural networks. Neural networks are a platform for machine learning programs.
They can detect patterns in data – including in human behavior – from voice recognition to financial decisions to food purchases.
Another way AI and semiconductors work together is through edge computing. Edge computing is a decentralized programming platform.
AI algorithms can use edge computing to run on devices closer to their data sources. That means faster response times.
Many technologies can benefit from edge computing power.
These span smartphones, Internet-of-Things devices, autonomous vehicles, and streaming devices – especially in areas with limited internet connectivity.
They all rely on AI hardware and software advancements.
And the AI revolution is only in its first inning.
So, if you feel that you’ve missed the boat on Nvidia, don’t worry. You can still take advantage of any further upside on Nvidia and other companies forging forward in AI.
A simple way to do that is through the Artificial Intelligence and Technology ETF (AIQ). It covers a broad array of large companies in the AI hardware space.
Regards,
 |
Nomi Prins
Editor, Inside Wall Street with Nomi Prins