Intel's biggest mistake and Nvidia's rise
Chip War: The Fight for the World's Most Critical Technology
Data is not the new gold. Chips are.
The internet, cloud, social media, “Big Tech”, and the whole digital world exist because engineers figured out how to control tiny electrons moving through silicon.
In Chip War, Chris Miller covers the story behind the chip revolution, how it stopped the Cold War, and how is it transforming the world today.
So what the hell is a chip?
Most materials either conduct electricity or don’t. A semiconductor is a special material, like silicon, that can do both, based on how you treat it!
A chip is made from a semiconductor material and contains millions or billions of transistors. Each transistor acts like a tiny switch, controlling the flow of electricity.
When a small electrical signal is applied, the transistor allows electricity to flow (a binary “1”), and when it’s removed, the transistor blocks the flow (a binary “0”). By rapidly switching between these states, transistors process data and perform calculations.
The more transistors packed into a chip, the more calculations it can handle simultaneously, which makes devices faster and more powerful.
Who makes all those chips?
Creating a single chip is extremely complex, and involves many different companies working together:
First you need software to design the chip (Cadence, Synopsys, Mentor Graphics).
Then you actually need to design it (Nvidia, AMD, ARM, Intel)
And then manufacture it (TSMC, Samsung, Intel)
But to manufacture, you need specialized equipment (ASML, Applied materials)
And materials (Shin-Etsu Chemical, SUMCO)
And finally package and test it (ASE Group, Amkor Technology)
There are only 3 companies that make chips - TSMC, Intel, and Samsung. TSMC is the only one that focuses only on making chips, and doesn’t compete with its customers. We will get to why it’s super important in a second.
The crazy fact I didn’t know until reading the book is that Nvidia doesn’t manufacture any chips! It only designs them, and TSMC (from Taiwan) actually makes them.
Chip manufacturing isn’t just complex - it’s also the most global process ever, no single country can handle the entire process alone. In recent years, with the US-China ‘Cold War II’, China has tried to not depend on anybody - with no success so far.
Now that we understand the chip-making process, let’s look at how two major players - Intel and Nvidia.
The end of Intel?
As we just discussed, Intel has 2 different businesses:
A chip factory (called a foundry, or fab), which also makes chips for other companies.
The chip design business, which uses the fab to produce them.
This is a great example of how focusing on 2 things makes you lose on both ends.
Intel is worth now $80B (was ~$280B 3 years ago)
AMD (only designs chips) $214B
TSMC (only makes chips) $670B
Nvidia 2.52T - x30 of Intel…
So how did the decline start?
Shortly after the deal to put Intel’s chips in Mac computers, Jobs came back to Otellini [Intel’s CEO] with a new pitch. Would Intel build a chip for Apple’s newest product, a computerized phone?
All cell phones used chips to run their operating systems and manage communication with cell phone networks, but Apple wanted its phone to function like a computer. It would need a powerful computer-style processor as a result.
“They wanted to pay a certain price,” Otellini told journalist Alexis Madrigal after the fact, “and not a nickel more…. I couldn’t see it. It wasn’t one of these things you can make up on volume. And in hindsight, the forecasted cost was wrong and the volume was 100 × what anyone thought.”
Intel turned down the iPhone contract.
Just a few of years after Intel turned down the iPhone contract, Apple was making more money in smartphones than Intel was selling PC processors.
Intel never found a way to win a foothold in mobile devices, which today consume nearly a third of chips sold. It still hasn’t.
After missing the smartphone, Intel also missed the AI:
Over several months in 2017 and 2018, executives at the two companies [Intel and OpenAI] discussed various options, including Intel buying a 15% stake for $1 billion in cash, three of the people said. They also discussed Intel taking an additional 15% stake in OpenAI if it made hardware for the startup at cost price, two people said.
Intel decided against a deal, partly because then-CEO Bob Swan did not think generative AI models would make it to market in the near future and thus repay the chipmaker's investment
And we’ll wrap up with a great quote from Ben Thompson’s article:
The fundamental problem facing the company is encapsulated in that paragraph:
Intel doesn’t have the best manufacturing
Intel doesn’t design the best chips
Intel is out of the game in AI
The rise of Nvidia
From the falling giant to the rising one.
As we discussed, Nvidia focuses only on designing the best chips, letting TSMC focus on making them. This specialization is why Nvidia can stay ahead in AI without owning factories.
Here’s the story of Nvidia’s success in ~100 words:
In 2006, Nvidia realized that high-speed parallel computations could be used not only for computer games and graphics. They released CUDA - a software that lets GPUs be programmed in a standard programming language, without any reference to graphics at all.
Even as Nvidia was churning out top-notch graphics chips, Huang spent tons of money on this software effort (at least $ 10 billion), to let any programmer work with Nvidia’s chips. Huang gave away CUDA for free, but the software only works with Nvidia’s chips.
By making the chips useful beyond the graphics industry, Nvidia discovered a vast new market for parallel processing, from computational chemistry to weather forecasting. At the time, Huang could only dimly perceive the potential growth in what would become the biggest use case for parallel processing: artificial intelligence.
Nvidia has since bet its future on AI. From its founding, Nvidia outsourced its manufacturing, largely to TSMC, and focused relentlessly on designing new generations of GPUs and rolling out regular improvements to its special programming language. As investors bet that data centers will require ever more GPUs, Nvidia has become America’s most valuable semiconductor company
Remember - it was written before ChatGPT was released, and when Nvidia was worth 10% of what it’s worth today!
What the future holds
What will happen when we run out of chips? In 2020 we had a good chance to find out:
During the COVID-19 pandemic, the world talked about a “chip shortage”.
People assumed it was because factories were closed - that’s not true! The global supply chain remained surprisingly resilient - factories continued operating even through lockdowns. The real issue was misjudged demand.
Early in the pandemic, car manufacturers cut their chip orders, expecting a slowdown in sales. Meanwhile, demand for tech products exploded as people needed laptops, gaming consoles, and smartphones to work and entertain themselves at home.
By the time car companies realized their mistake and tried to increase orders, the production capacity was already taken by the tech industry. The result wasn’t a shortage of supply, but a rise in demand!
This meteoric rise continues today, highlighting how crucial chips have become - not just for tech but for every aspect of modern life, from communication to transportation.