SiMa.ai creates software and hardware for industrial robots, drones, security cameras, and eventually self-driving cars to use AI algorithms.
SiMa.ai, an AI chip startup based in Silicon Valley, announced on Tuesday that it had raised an additional $13 million from investors, including VentureTech Alliance, a significant Taiwanese fund that has a close strategic relationship with Taiwan Semiconductor Manufacturing Co. (2330.TW).
This is at least VentureTech Alliance’s third investment in American chip startups in the last month. When Ayar Labs raised an additional $25 million and Ethernovia an additional $64 million in May, VentureTech invested. Up to this point, SiMa.ai has raised a total of $200 million.
While ChatGPT’s viral success has sparked a new wave of investments in AI hardware, some startups that received significant funding in previous years are already experiencing difficulties.
According to Navin Chaddha, managing director of the investment firm Mayfield and the creator of several chip startups, “the over-funding is in the training market and in the data centre market.” Chaddha contributed personally to this most recent round.
SiMa.ai creates software and hardware for industrial robots, drones, security cameras, and eventually self-driving cars to use AI algorithms. According to Chaddha, there are few competitors in that market despite the sizeable market.
Although respectable, (SiMa.ai’s) valuations aren’t at the historic multi-billion dollar valuations that the datacenter players had at the time, according to Moshe Gavrielov, a member of the SiMa.ai board who also serves on the board of TSMC.
The struggles of British AI chip unicorn Graphcore have been widely reported.
AI chip Startup, SiMa.ai declined to say what the current valuation is but CEO and founder Krishna Rangasayee said the company, founded in 2018, was already generating revenue and over 50 customers were testing its chips.
Rangasayee also pointed to one recent benchmark testing result by SiMa.ai that beat AI chip giant Nvidia Corp (NVDA.O) in performance and power of chips used on devices like cameras, drones and robots.
The testing data is published by MLCommons, an engineering consortium that maintains testing benchmarks widely used in the AI chip industry. “To me, it’s a David versus Goliath. In reality, Nvidia is the primary competitor for us and…we are better than them in performance and in power,” said Rangasayee.