Munich, New York It was an argumentative feat that Nvidia founder and CEO Jensen Huang had to pull off on Wednesday night. On the one hand, Huang had to explain the failure of the billion-euro purchase of the British chip developer Arm – a sensitive setback. On the other hand, he had to provide a stock market-ready outlook on the future and present an armless growth strategy. He succeeded in the trick. The upcoming Metaverse, the connection between the off- and online world, which Facebook boss Mark Zuckerberg had proclaimed, played a decisive role.
The figures for the past year initially gave argumentative momentum. These turned out excellent. The worldwide demand for chips drove the semiconductor specialist’s business to a record high. Sales rose in the fourth quarter, which ended in January, by 53 percent to a good $7.6 billion. The bottom line was a profit of three billion dollars, more than twice as much as in the same period last year. For the past financial year, Nvidia reported sales of $26.9 billion (up 61 percent) and net income of $9.7 billion (up 125 percent).
Huang addressed the failed takeover of British chip developer Arm early on. “We did our best, but the headwind was too strong,” he said in the analysts’ desk, referring to the resistance from US, European and, above all, British supervisors. The failure of the deal burdened the quarterly result with 1.3 billion dollars: The arm’s owner, the Japanese technology investor Softbank, is allowed to keep this advance.
But Huang didn’t dwell on the bankruptcy for too long. Rather, his group is well positioned for the future even without an arm – which is mainly due to the upcoming Metaverse, which Nvidia calls “Omniverse”. On this in-house platform, creative people, designers, researchers and engineers should work together virtually with physically correct simulations in real time.
Top jobs of the day
Find the best jobs now and
be notified by email.
“Today’s Internet is 2D,” Huang explained. “We are building the Omniverse to make it 3D and connect it to the real world. This is the actual new threshold.”
3D applications belong to the future, and not just for video games or films, “but for all kinds of digital applications.” With the in-house Omniverse engine, companies could build a digital factory, control a fleet of autonomous cars or industrial robots. Employees could then monitor production from home, even on the other side of the world. “We’re still at the very beginning here,” says Huang.
At first glance, such visions may sound a bit esoteric. In fact, Nvidia is in pole position to capitalize on the upcoming Metaverse, no matter what that might look like. 3D and AI calculations are becoming more important in many sectors of the economy, and Nvidia supplies the hardware for this: powerful graphics processors.
It all started with the development of graphics cards for video games, says Jonathan Curtis, who manages the almost eleven billion dollar Franklin Technology Fund at the US fund house Franklin Templeton. And Nvidia has remained true to its core product to this day. “It’s paying off now. We’re in a golden era for GPUs, and that’s Nvidia’s bread and butter.”
The processors known as GPUs are crucial in gaming, but have long been indispensable in data centers and for AI applications, i.e. for calculations such as those needed for autonomous driving. A small but interesting business is also that of processors for cryptocurrency mining.
In the gaming sector, Nvidia sales rose 37 percent to $3.4 billion last quarter. The data center business is almost as big today, having jumped 72 percent to $3.2 billion.
An important customer is the Facebook group Meta. He wants to use Nvidia technology in his planned supercomputer, whose performance should overshadow that of all previous high-performance computers. “Nvidia is growing strongly in its core business, in gaming chips, but also in other areas,” says fund manager Curtis. “Nvidia is making itself less dependent on the volatile gaming market. That’s a good thing.”
Continued growth story
Investors are rewarding the strategy, valuing Nvidia at almost 70 times annual earnings. Big competitor Intel has a price-to-earnings ratio of one to ten. But can it always go on like this?
The share price of the most valuable chip manufacturer in the world has collapsed by around a fifth since the fall. However, experts believe that this is at best a correction at a high level. The growth story is far from over, says Peter Fintl, chip expert at the consulting firm Capgemini.
New customers such as the car industry, which wants to master autonomous driving in the coming years and needs chips for this, are optimistic. In order to get hold of the coveted hardware, car manufacturers such as Mercedes-Benz have embarked on a previously unknown deal: Nvidia will receive a permanent share of their earnings.
For CEO Huang, this is the future. “If we can dock a GPU to enable AI, robotics, autonomous cars, cloud computing, super computing, then we want to bring the full spectrum of Nvidia’s power,” he said. For customers, this means one thing above all in real terms: Nvidia wants to be paid for its power.
“The Omniverse business model is based on contributions,” Huang explained frankly. Nvidia relies entirely on license models and monthly service subscriptions. “We receive income over the entire life cycle of a car. And you can imagine what will happen in ten years. There are great opportunities there.”
The CPU gap
So everything is fine in the Nvidia kingdom? Huang remained vague on one point: how he wants to close Nvidia’s gaps in the CPU sector, now that the arm deal has failed. They will “now re-examine how we can use the freed-up capital.”
In order to understand the challenge, a look at the structure of the chip world is necessary: Nvidia has a market share of around 80 percent for graphics cards. The two US competitors AMD and Intel dominate the rest. All earn well, albeit in different segments.
There are roughly two types of PC chips: main processors (CPUs) and graphics processing units (GPUs). CPUs can perform more complex arithmetic operations, but are slower. GPUs are suitable for simpler arithmetic calculations, but with billions of transistors on a chip, they can carry out a particularly large number of operations in parallel. This is particularly important in the AI area, where large amounts of data have to be evaluated.
“Nvidia has conquered the GPU market but doesn’t offer CPUs. AMD offers GPUs and CPUs, but is lagging behind in terms of market share. And Intel has gone through a zigzag course,” explains Glenn O’Donnell, chip expert at US analysis firm Forrester. Intel is the market leader in CPUs. The group has entered and exited GPUs twice in the last ten years. At the moment, Intel mainly offers integrated graphics cards for laptops and now wants more.
To attack Intel, Nvidia needs CPU expertise, says O’Donnell, and that was the real rationale behind the ARM takeover plan. “Arm has done an excellent job of designing CPUs. Their main advantage is the low energy consumption.” The latter is particularly relevant for smartphones in order to extend the battery life. Almost all smartphone chips today are based on Arm technology.
Example iPhone: “Apple used to be a big customer of Intel. Not anymore,” says O’Donnell. With its new M1 chip, Apple relies on ARM technology, which the British license. “Nvidia wanted to secure exactly this technology. If the acquisition had gone through, Nvidia would have become Apple’s most important supplier.”
If Huang had started the takeover attempt a year earlier, he might have gotten away with it, says Capgemini’s Fintl. But the political pressure fueled by competitors in America, the EU and Great Britain, where poor is considered a “national treasure”, has simply become too great. Why did Huang still pursue the takeover plans? “It was just a shot in the dark,” says Curtis. “It was worth a try. That would have been the deal of the century.”
from the dream. Fintl believes that what is a setback for Nvidia is an advantage for the chip industry, at least in the short term: If Arm remains independent, access to basic technologies would be unhindered for all sides.
And Nvidia itself is still extremely highly valued at $663 billion. Intel only reaches around 200 billion dollars, with sales a good four times as much as Nvidia. The striking difference should also be due to Nvidia’s charismatic boss.
“If a problem is easy to solve, then it doesn’t interest me,” Huang once said during a visit to Munich. This is perhaps also due to the fact that he was not exactly born with success. Huang was born in Taiwan, his parents emigrated to Thailand when he was young. When the political situation there deteriorated, they sent the boy to his aunt and uncle in the USA. He spends his youth in Oregon, studies electrical engineering and does his master’s degree at Stanford.
On his 30th birthday, he founds Nvidia with Curtis Priem and Chris Malachowsky. First, the company is called “Next Version”, when it is to be entered in the commercial register, the founders come up with the Latin “Invidia” (envy). The rapid rise followed with chips for computer games, and in 1999 the married family man listed the company on the stock exchange.
Like few other managers, the 58-year-old has managed to open up new business areas. “Nvidia is one of the big drivers of the tech world,” says Curtis. “Jensen Huang did a lot of things right,” agrees Forrester expert O’Donnell. “Huang is an electrical engineer. And technology groups that are managed by a technician instead of a numbers man are usually better off.” The negative example is Intel, where a former CFO was in charge for a while. The new CEO Pat Gelsinger is again a technician.
Unlike Intel, Nvidia does not invest a cent in chip factories. Nevertheless, the group has access to the most modern production processes: Nvidia uses the contract manufacturer TSMC. A clever decision: The world’s largest so-called foundry is technologically far ahead. Rival Intel, on the other hand, has to spend many billions to bring its own factories up to date. That squeezes the margin and ties up capacity – which makes Nvidia significantly more profitable.
If Huang has his way, the rise should continue, including in the Metaverse. “We’re still in the early days of AI adoption. It’s unbelievable what’s still possible there,” he announced on Wednesday. For the current quarter, Nvidia announced a further increase in sales to $8.1 billion. The analysts had expected around 7.3 billion dollars. But investor hearts do not beat indefinitely for Nvidia either. In any case, Wall Street commented coolly on Thursday night: Nvidia shares fell by 2.6 percent in after-hours trading.
More: Nvidia fails to acquire Arm.