I sometimes forget that normal people haven't been obsessing about this stuff for going on a decade. As just one example, out of hundreds of posts on NVDA, is a conclusion we reached early on, still hold, and tried to express in the introduction to a 2020 IBD story:
Nvidia (NVDA)
dominates the artificial intelligence market with bleeding-edge
processors now powering the AI revolution. And that lead has pushed
Nvidia stock to record highs.
But challengers are circling Nvidia's fortress. The tech giant looks
to defend itself with a competitive moat by taking steps to make it
harder for competitors and would-be rivals to match, if not surpass,
Nvidia's formidable leadership in AI. A competitive moat serves as a
barrier that protects a company's business stronghold. It can include
proprietary technologies, strategic partnerships, and other corporate
advantages.
Nvidia, which just delivered a beat-and-raise quarterly report,
has built a well-regarded software developer platform and developed a
full-stack hardware solution for AI. It also has forged major alliances
with other tech leaders, such as Amazon and Google, and invested in
trailblazing AI startups. These moves have created a dominant AI
ecosystem.
Defending its moat won't be easy. The Santa Clara, Calif.-based
company must grapple with Big Tech rivals eyeing a piece of the AI
market, and government regulators raising antitrust concerns.
Nvidia Stock: A Powerful Competitive Moat
"Their competitive moat is very powerful, because for the past 15 years they've been investing in software in a way that allows their hardware to outperform regular silicon because of the software optimizations and acceleration libraries that are updated constantly," Rosenblatt Securities analyst Hans Mosesmann told Investor's Business Daily. "They have that advantage over everybody else."Mosesmann rates Nvidia stock as a buy with a price target of 200. It
closed Thursday at 117.59. Since the start of 2023 through Aug. 29,
Nvidia stock has risen 705% on investor enthusiasm for its leadership
in AI.
However, Nvidia stock fell 6% Thursday in the wake of its earnings
release after the prior day's market close. Nvidia earned an adjusted 68
cents a share on sales of $30.04 billion in its second quarter ended
July 28. On a year-over-year basis, Nvidia earnings soared 152% while
sales jumped 122%. It was the company's fifth straight quarter of
triple-digit gains. Yet the chipmaker failed to reach lofty "whisper numbers" with its fiscal Q2 results and outlook.
For the current quarter, Nvidia forecast revenue of $32.5 billion, up 79%.
Nvidia seized the lead in AI by building on a key advantage in its
chip technology. The company's high-performance chips evolved from
devices that run parallel processing applications for video game
graphics. And those processors have since been optimized to run AI
workloads.
The ChatGPT Craze
Nvidia became an even more prominent AI stock with the ChatGPT craze. The OpenAI chatbot and its imitators have sent Nvidia stock soaring over the past two years.Nvidia isn't just a chipmaker now. Through its hardware partners like Super Micro Computer (SMCI) and Dell Technologies (DELL),
it offers a full stack solution of hardware, software and services for
AI computing. No other company comes close to matching its capabilities.
Nvidia CEO's AI Vision
Jensen Huang, Nvidia's co-founder and chief executive, has won praise for the chipmaker's successful pivot to AI.Huang early on recognized the importance of graphics processing units
for high-performance computing and data centers. He funneled profits
from sales of Nvidia's lucrative gaming GPUs into developing AI
technology and the AI market.
"Jensen was ahead of the curve when he came out with CUDA software
and then got deeper into the data center space," CFRA Research analyst
Angelo Zino told IBD.
"He saw the tea leaves back then and invested heavily in AI
capabilities and making sure he was ready for this type of moment when
it came."
It's Nvidia's AI Market
Morningstar analyst Brian Colello argues that Nvidia helped create the artificial intelligence market.In addition to general-purpose AI, Nvidia is developing AI products
for specific industries such as architecture, automotive, engineering,
financial services, health care, life sciences and more.
The semiconductor company "is reaping the rewards of making the early
investments to enable this entire AI ecosystem," Colello told IBD.
"Nvidia has one of the widest and deepest competitive moats in all of
technology."
Nvidia Stock: Challengers Circle The Moat
But Nvidia's top rivals in the tech industry are circling the moat.Advanced Micro Devices (AMD),
a major Nvidia competitor in graphics chips, has emerged as its biggest
challenger in AI. It also has a software stack for GPU programming
called ROCm, which competes with Nvidia's CUDA.
"As for AMD, there will be a place for a second general-purpose GPU
company," Colello said. "And AMD is doing the right things investing in
hardware and investing in the software ecosystem to get second place.
And a distant second place in AI doesn't sound all that appealing except
AMD will still likely be far, far ahead of third place, and it's second
place in a massive and fast-growing market."
On Aug. 12, AMD completed its $665 million acquisition of Silo AI,
the largest private AI lab in Europe. The deal will further AMD's
commitment to deliver end-to-end AI solutions based on open standards,
AMD said.
On Aug. 19, AMD announced plans to buy computer hardware design firm ZT Systems
for $4.9 billion. Based in Secaucus, N.J., ZT Systems is one of the
largest private providers of AI training and inference infrastructure in
the world, serving some of the largest cloud-computing companies.
However, AMD intends to sell ZT's manufacturing business and keep its
systems design business.
Meanwhile, AI chip startups like Cerebras Systems, Groq and SambaNova
Systems are developing semiconductors to take on Nvidia head on.
Nvidia Stock: A Platform Battle
The AI market could end up looking like the mobile phone market with Apple (AAPL) vs. Android, Rosenblatt's Mosesmann said."So, you have one guy that controls the entire stack and is capturing
most of the value," he said. "And the rest of the industry is focused
on an open architecture and open software."
Those other players in the rest of the AI market include AMD and custom processor makers Broadcom (AVGO) and Marvell Technology (MRVL). Hyperscale cloud service providers like Amazon (AMZN), Microsoft (MSFT) and Alphabet's (GOOGL) Google also have developed their own application-specific integrated circuits, or ASICs, for AI applications.
A Multivendor Market
Meanwhile, PC and server chip leader Intel (INTC) continues to struggle with its AI processor efforts. A possible partnership between Intel and Japan's SoftBank to develop an AI chip to compete with Nvidia recently broke down, the Financial Times reported.Moses Guttmann, co-founder and chief executive of open-source
platform ClearML, believes the AI market will eventually move to a
multivendor approach to artificial intelligence.
Berkeley, Calif.-based ClearML makes an open-source platform for
running AI workloads across different processors and systems. It
provides a software orchestration layer to support hybrid AI computing.
"It makes sense to have an external software layer that is
vendor-agnostic, silicon-agnostic and (computer hardware)-agnostic to
make sure that the market has more options once other silicon vendors
are back in the game," Guttmann told IBD. Nvidia controls an estimated
90% of the AI GPU market.
The Need For Competition
For now, most AI applications are running on Nvidia GPUs, but AMD has a competitive offering with its latest Instinct accelerators, Guttmann said.
"We always want to have competition," he said. "It's good for innovation and it's good for the market and getting the best bang for your buck."Nvidia Stock: A Sticky Ecosystem
Nvidia built a very "sticky" end-to-end system of software, hardware and services, Colello said. And once customers have chosen Nvidia, they're likely to stick with it, he said.
"The race for AI makes it even less likely customers will switch, because why would you slow down your AI development?" Colello said. "Cost savings are nice, but there is a real time-to-market trade-off by seeking those cost savings by going somewhere else."
The CUDA Factor
Nvidia is producing best-of-breed graphics processing units, or GPUs, for data centers. However, the company's competitive moat is mostly on the software side, Zino said.
And that starts with Nvidia's proprietary CUDA software for accelerated computing with GPUs. The CUDA platform includes compilers, libraries and developer tools to help programmers accelerate their applications. More than 5 million developers worldwide use CUDA....
