This tech giant is challenging Nvidia: Is the artificial intelligence (AI) market about to take a sharp turn?

When OpenAI launched its ChatGPT chatbot about a year ago, it set off a new zeitgeist in an unsuspecting world. Artificial intelligence (AI) is still a hot topic in cities, Wall Street, and Silicon Valley. This is especially true for the generative AI technologies that power systems like ChatGPT.

Investors soon noticed. nvidia (NVDA -0.33%) By providing the AI-accelerated hardware that makes all of this possible, we’ve achieved amazing results. Nvidia’s stock price has more than tripled in 2023, driven by physical sales of AI-specific processors and expectations of continued dominance in this hot market segment.

But Nvidia shouldn’t rest on its laurels. They are not the only chip designers on the market, nor are they the only companies interested in lucrative AI opportunities.

The latest challengers to enter the ring and challenge Nvidia’s AI mastery are: Samsung (SSNL.F -28.75%). The South Korean tech giant has formed a partnership. Naver (OTC: NHNC.F) — a giant online entertainment company in the same country — develops both hardware and software that matches or surpasses the best tools available today.

Specifically, Samsung and Naver claim that their upcoming AI chips will be eight times more energy efficient than Nvidia’s H100 accelerator.

This isn’t the same as a direct performance record, but a more power-efficient solution could actually pose a much bigger threat to Nvidia’s throne. Here’s why:

The efficiency advantage of AI computing

In the realm of high-performance AI computing, efficiency is key. Pure performance doesn’t really matter because you can always throw more hardware at a number crunching problem.

The supercomputer that trains the ChatGPT-style AI system is equipped with thousands of Nvidia A100 accelerators, each with nearly 7,000 processing cores. The real challenge is providing enough power to drive this monster and then cool the resulting space heater. The OpenAI/Nvidia system consumes 7.4 megawatts at full power. This is similar to a cruise ship crossing the ocean or a large steel mill.

AI giants are therefore looking for power-hungry solutions that can deliver better results per watt.

Samsung and Naver’s claims of an AI chip that is eight times more energy efficient than Nvidia’s H100 could represent a paradigm shift. In a world increasingly conscious of energy consumption and costs, more efficient chips don’t just mean lower power costs. This means smaller carbon footprints, more compact physical installations, and the ability to deploy more powerful AI systems without prohibitive energy costs.

Challenge Nvidia’s dominance

Nvidia has long established itself as a provider of AI acceleration hardware, which is reflected in its soaring stock price and market position. But with Samsung and Naver entering the field promising groundbreaking, energy-efficient AI chips, Nvidia faces a new kind of competition. It’s no longer a question of who has the fastest chip on the market. It’s about who can deliver results in the most efficient way. And this time, Nvidia may not be the clear winner.

These developments are not just a two-horse race. Companies like AMD and Intel also each have their own AI chip solutions. The AMD Instinct MI300 line offers more memory and lower power requirements than previous generations. Intel’s Gaudi3 solution focuses on faster baseline performance and next-generation networks that tie the processors together. Everyone has a unique master plan.

But I’ve never claimed that these alternatives blow Nvidia’s power efficiency out of the water. Samsung and Naver’s focus on low-power requirements could set a new standard, forcing others to follow suit, but still gives the Korean duo a strong first-mover advantage. As AI technology becomes increasingly integrated into a variety of sectors, from healthcare to finance, the demand for efficient, powerful, and cost-effective AI computing will continue to grow.

What are your future plans?

This has been all the theory so far. Benchmark testing and real-world deployment results will not be available until after 2024, when Nvidia’s competitors launch new AI chips into the mass market. Investors will be left to make informed guesses about how closely each company’s claims align with its bottom line performance, from power consumption and raw-number-crunch performance to next-level connectivity and other potential game-changing ideas.

Can Samsung and Naver’s AI efficiency-focused chips deliver on their promises? How will Nvidia and other competitors respond? Time will tell, but one thing is clear. The AI ​​chip market is evolving, and along with it, the environment of artificial intelligence itself is also evolving. The next few years will be critical in determining the direction of this technology and its impact on the world.

We’re not sure which company(s) will dominate the AI ​​hardware market in the long term, but Samsung has just joined the ranks of potential winners. If you didn’t previously consider Samsung to be a leading chip designer, the company has just joined the elite list.