Analysis
Ambiq Micro filed a Form S-1 on June 22, 2026 to list on the New York Stock Exchange. The Austin, Texas company designs ultra-low-power AI silicon — its Apollo processor family runs machine learning inference at power budgets under 1 milliwatt, targeting always-on wearables, hearing aids, industrial IoT sensors and battery-powered edge devices.
The filing lists ~200 million cumulative shipped units, making Ambiq the largest deployed base of AI-capable ultra-low-power silicon in the industry. Customers include a mix of consumer electronics OEMs (smartwatches, fitness bands) and industrial IoT designers.
“The filing lists ~200 million cumulative shipped units, making Ambiq the largest deployed base of AI-capable ultra-low-power silicon in the industry.”
Market context: on-device AI is now a real design category, spurred by frontier model providers who explicitly encourage local inference (Apple Intelligence, Google Gemini Nano, Anthropic's soon-to-launch on-device Claude). Every one of those tech platforms needs sub-milliwatt silicon partners, and Ambiq's Apollo generation is one of two or three credible options.
Comparable deals: SiFive raised at $2.5B in 2022; Groq trades in secondary around $6B; Cerebras is marketing its own S-1. Ambiq's likely valuation range at IPO is $1.5-2.5B based on comparable trading multiples, though final pricing depends on 2027 growth guidance.
What to watch: the S-1's revenue disclosures (Ambiq has been private about specific numbers), any strategic anchor investor commitments, and the read-across to other ULP-AI competitors like Syntiant (private) and Kneron (private, planning its own IPO).
