Canopy Wave is being acquired, but its founders will run the combined company. SAIHEAT, a Nasdaq-listed Singapore-based distributed computing operator, has signed a definitive merger agreement to buy the California AI inference and GPU cloud platform at a pre-money equity valuation of $60 million, against SAIHEAT's own negotiated value of $40 million. Under the terms, former Canopy Wave shareholders will hold 54.19% of the combined entity's economic interests and 78.44% of its voting power.
Who ends up in charge
The merged business will operate under the Canopy Wave name, trade on Nasdaq under the ticker "CWAV," and be headquartered in Santa Clara. Canopy Wave CEO Tao Zhang and CTO James Liao will lead the combined entity. Zhang and Liao are expected to collectively hold a majority of both the economic interests and voting power in the enlarged company. SAIHEAT CEO Jianwei Li said the combination positions the business "where the AI market is going: inference at scale."
Zhang framed the deal around enterprise demand: companies are evaluating open-weight large language models for performance, control, and cost efficiency. The combined company's stated mission is to make serving those models "simple, secure, and economical."
What each side brings
Canopy Wave contributes GPU cloud infrastructure, orchestration software, and API access. It holds SOC 2 Type II certification and enforces a zero-data-retention policy, two items that matter to enterprise buyers weighing compliance risk. Its current limitation: it relies on leased third-party infrastructure. SAIHEAT adds modular data centre expertise and energy-efficient computing, which the companies say will address that gap.
The combined entity will target enterprise and developer clients globally, offering inference services for open-weight models. SAIHEAT expects to transition from foreign private issuer to domestic issuer reporting in the next fiscal year.
The counterargument
Both boards unanimously approved the transaction, and SAIHEAT management's central argument has real grounding: enterprise AI investment is shifting from model training to inference, and the combined company's compliance stack targets regulated buyers who cannot use hyperscaler products off-the-shelf.
The risk is that the combined entity enters a market crowded with hyperscaler inference offerings and well-funded AI infrastructure players, at valuations that leave little room for a slow ramp. Canopy Wave still depends on leased capacity. SAIHEAT's data centre expertise narrows that gap on paper; the line to watch is whether it does so at competitive cost and speed. Completion is expected by the end of 2026, subject to SAIHEAT shareholder approval and Nasdaq listing approval. The merged entity will trade under "CWAV."