Nvidia’s dominance in AI chips holds back funding for startups

Sept 11 (Reuters) – Nvidia’s ( NVDA.O ) dominance in making computer chips for artificial intelligence has cooled venture funding for rivals, investors said, as the number of U.S. deals fell 80% in the quarter from a year ago.

The Santa Clara, California company dominates the market for chips that work with large-scale language data. Generative AI models become increasingly smarter through exposure to more data, a process called training.

As Nvidia has become stronger in this area, it has become more difficult for companies trying to create competing chips. Viewing these startups as risky bets, venture financiers are reluctant to commit large amounts of new cash. It can cost more than $500 million to advance a chip design into a working prototype, so the pullback has quickly threatened the startup’s prospects.

“Nvidia’s continued dominance makes a good point about how difficult it is to break into this market,” said Greg Reichow, partner at Eclipse Ventures. “This has resulted in a backlash in investment in these companies, or at least in many of them.”

US chip startups have raised $881.4 million as of the end of August, according to Pitchbook data. That compares to $1.79 billion for the first three quarters of 2022. The number of deals has dropped from 23 to four at the end of August.

Nvidia declined to comment.

AI chip startup Mythic, which has raised a total of $160 million, ran out of cash last year and nearly had to halt operations, technology website The Register reported. But it managed to bring in a relatively modest $13 million investment a few months later in March.

Nvidia has contributed “indirectly” to the overall AI chip fundraising woes, as investors “want huge returns, with big investments in home run only type investments,” Mythic CEO Dave Rick said.

Difficult economic conditions have contributed to the downturn in the cyclical semiconductor industry, Rick said.

Rivos, a secretive startup working on chip designs for data servers, has recently had trouble raising funds, two sources familiar with the company’s situation said.

A Rivos spokeswoman said Nvidia’s market dominance has not hindered fundraising efforts and that its hardware and software “continue to excite our investors.”

Rivos is embroiled in a lawsuit with Apple ( AAPL.O ), which accuses Rivos of stealing intellectual property, adding to fundraising challenges.

Investor demand

Chip startups looking to raise cash are facing stiff demand from investors. For them, companies need to have a product that is within months of launch or already generating sales, sources said.

About two years ago, new investments in chip startups were often $200 million or $300 million. That has fallen to about $100 million, according to Pitchbook analyst Brendan Burke.

At least two AI chip startups have overcome investor reluctance by touting potential customers or their relationships with well-known executives.

In raising $100 million in August, Tenstorant boasted CEO Jim Keller, a near-legendary chip architect who has designed chips for Apple, Advanced Micro Devices ( AMD.O ) and Tesla ( TSLA.O ).

D-Matrix, which had forecast revenue of less than $10 million this year, raised $110 million last week, bolstered by financial backing from Microsoft and the Windows maker’s commitment to test D-Matrix’s new AI chip after it launches next year.

While these chipmakers struggle in Nvidia’s shadow, startups in AI software and related technologies don’t face the same difficulties. They brought in about $24 billion in funding as of August, according to PitchBook data.

Despite Nvidia’s dominance in AI computing, the company doesn’t have an impenetrable lock on the field. AMD plans to launch a chip this year that will compete with Nvidia’s, and Intel ( INTC.O ) has stepped up development by acquiring a rival product in acquisitions. Sources see it as having long-term potential as an alternative to Nvidia’s chip.

There are also affiliate applications that can provide opportunities for competitors. For example, data-intensive computing chips for prediction algorithms are an emerging niche. Nvidia doesn’t dominate this area and it’s worth the investment.

Our Standards: The Thomson Reuters Trust Principles.