analysis Some AI chip startups are managing to raise capital from investors despite operating in a crowded market of competitors, where venture capital funding has plummeted over the past year.
Venture-backed AI chip designers who have raised funding rounds in recent months include Israel-based NeuReality, Netherlands-based Axelera, and SiMa.ai, Quadric and EnCharge AI, three US firms headquartered in Silicon Valley.
These startups convinced investors to part with tens of millions of dollars in an investment environment that is much more conservative due to this year’s shaky economy, which has been plagued by inflation and rising interest rates.
As The registry As recently reported, global VC funding for semiconductor startups fell 46 percent in 2022 to $7.8 billion as of December 5, reflecting increased scrutiny of these capital-intensive companies.
While funding for semiconductor startups took a deep tumble this year, the decline in the total number of known funding rounds was not as pronounced, falling 20 percent to 618 deals in 2022.
Of the five startups raised in recent funding rounds, four – Axelera, SiMa.ai, Quadric, and EnCharge AI – are focused on AI chips for running inference on edge devices, where data is processed at high speeds at the lowest Power consumption to be processed possible in constrained form factors is of paramount importance.
This can’t be a complete coincidence. As of October 31, VC funding for inference-based AI chip designers had exceeded funds raised by startups focused on processors for training, the first step in developing AI applications before committing to inference for to use in the real world. That’s according to a research note from PitchBook senior analyst Brendan Burke in November.
Burke said this is a reversal of a trend seen over the past four years where funding for chip startups has been focused on training, or both training and inference exceeded capital for those working only on inference.
“There is a cyclical component to this pattern given the limited need for education firms to raise funds each year, but we see inference-focused firms forming significant commercial partnerships during economic recessions,” he said.
This increased need for inference chips coincides with a projected increase in spending on edge-of-the-road AI chips, a fairly nebulous notion that, in the Pitchbook report’s case, encompasses the PC, automotive, and industrial markets. According to Burke, the PC and automotive markets for AI chips will have grown more than 22 percent in 2022, faster than spending on such silicon for data centers.
“Demand for automotive and edge computing is driving more commercial deals for inference-focused chips than cloud training chips,” the analyst added.
Not all startups working on AI chips were lucky this year.
Mythic, a Texas-based startup developing analog chips for edge AI use cases, ran out of investor funding before it could generate any revenue, a top executive said in November.
Then there’s Graphcore, a British startup that has commercialized data center chips for training and inference. The company’s private valuation was reportedly reduced by $1 billion after it lost a key deal with Microsoft, among other financial woes.
Ruta Belwalkar, a private investor and chip designer, previously told us she wouldn’t be surprised if more chip design startups were acquired or closed over the next year because they didn’t move quickly enough from R&D to commercialization.
It’s not just investors who still find some AI chip startups palatable in a weakening economy. Some industry veterans are also taking the plunge.
For example, Lightelligence, a Boston-based optical AI chip startup, recently hired Weifeng Zhang, Alibaba’s former chief scientist for heterogeneous computing, along with Wayne Wu, the former head of AMD’s PCIe design team, and Hal Conklin, who was most recently Arm’s , hired vice president of global sales channel.
We wish them good luck. ®
https://www.theregister.com/2022/12/16/ai_chip_vc_funding/ Venture capital slowdown hasn’t hit AI chip developers yet • The Register