Investors claim that Nvidia’s dominance in the development of computer chips for artificial intelligence has curbed venture funding for potential competitors, as evidenced by the 80% decrease in U.S. deals this quarter compared to the same period last year.
The Santa Clara, California-based business controls the market for processors that handle enormous volumes of linguistic data. Training is the process by which generative AI models become gradually smarter by being exposed to additional data.
Companies trying to develop rival chips have a harder time doing so since Nvidia has gotten better in this area. Venture capitalists are now reluctant to offer significant cash infusions because they view these firms as a risky gamble. The withdrawal has seriously damaged the prospects of the businesses because it can cost more than $500 million to advance a chip concept to a functioning prototype.
Greg Reichow, a partner at Eclipse Ventures, stated “Nvidia’s continued dominance has put a really fine point on how hard it is to break into this market.” “As a result, investment in these companies—or at least into many of them—has decreased.”
According to data from PitchBook, American chip startups had raised $881.4 million as of the end of August. In comparison, the first three quarters of 2022 saw revenue of $1.79 billion. Through the end of August, there were only four deals, down from 23.
Nvidia opted not to respond.
Technology website The Register stated that AI chip startup Mythic, which has received over $160 million in total, ran out of money last year and almost had to stop operations. But it was able to attract a very small $13 million investment in March, many months later.
Because investors only demand “home run only type investments with a huge investment, huge return,” according to Mythic CEO Dave Rick, Nvidia has “indirectly” contributed to the overall fundraising difficulties for AI chips.
According to Rick, the slowdown in the cyclical semiconductor business has been exacerbated by challenging economic conditions.
According to two insiders acquainted with the company’s financial position, a covert business called Rivos that is developing semiconductor designs for data servers has recently having problems acquiring money.
Nvidia’s market dominance, according to a spokesperson for Rivos, hasn’t interfered with the company’s efforts to raise money, and its hardware and software “continues to excite our investors.”
The funding difficulty has been made worse by Rivos’ legal spat with Apple, which has accused Rivos of stealing intellectual property.
Obstinate investors
Investors’ requirements are becoming more stringent for chip firms wanting to raise capital. They demand that businesses have a product that has either just launched or is making sales, according to sources.
New investments in chip firms were frequently $200 million or $300 million about two years ago. According to Brendan Burke, a PitchBook analyst, that amount has decreased to roughly $100 million.
At least two AI chip startups have managed to win over investors by highlighting their connections to high-profile executives or potential clients.
Tenstorrent boasted about its CEO Jim Keller, a nearly legendary chip architect who has created chips for Apple, Advanced Micro Devices, and Tesla, to raise $100 million in August.
D-Matrix, which expects revenue of less than $10 million this year, raised $110 million last week thanks to support from Microsoft and a promise from the maker of Windows to test the company’s new AI chip when it becomes available in 2019.
While these Nvidia-shadowed chip makers struggle, startups in AI software and related fields do not experience the same limitations. According to data from PitchBook, they received about $24 billion in funding this year up until August.
Despite Nvidia’s dominance in AI computing, the market is not entirely under its control. AMD intends to release a chip this year that will compete with Nvidia’s, but Intel jumped ahead of AMD in development by acquiring a competitor’s product. According to sources, they could eventually replace the chip used by Nvidia.
There are also nearby applications that could offer doors for rivals. For instance, a growing market is data-intensive computing chips for prediction algorithms. This market is open to investment and is not dominated by Nvidia.
