SAIHEAT Ltd
SAIHEAT Ltd is a Singapore-incorporated, Nasdaq-listed company that operates distributed computing infrastructure through modular computing power systems designed to help energy owners utilize local power resources. Its business encompasses computing-power infrastructure and, subject to the proposed Canopy Wave combination, cloud-based AI inference services for open-weight generative AI models.
Fiscal year 2025 · change from previous year
Development 2021–2025
Business
Share of revenueModular computing-power systems help energy owners consume local energy and improve resource utilization.
Cloud infrastructure provides rented GPU capacity for AI and other compute-intensive workloads.
OpenAI-compatible interfaces support serving open-weight generative AI models in production.
Security, data isolation, zero-data-retention controls, and resource scheduling support enterprise deployments.
Customers
SAIHEAT's core customer base consists of energy owners seeking productive uses for local power and improved computing-resource efficiency. The proposed combination with Canopy Wave would add developers and enterprises that purchase GPU capacity and AI inference services for coding, AI-agent, and other production workloads. These customers value controllable costs, secure model serving, compatible APIs, and access to open-weight models.
Use modular computing systems to consume local energy and improve utilization of power resources.
Purchase GPU cloud capacity and API access to develop and operate AI applications.
Deploy open-weight models with security, data-isolation, and zero-retention requirements.
Use scalable inference capacity for coding assistants, agents, and production workloads.
Strategy
SAIHEAT's strategic direction is to combine distributed computing infrastructure with AI inference and GPU cloud capabilities. The proposed Canopy Wave transaction is intended to accelerate development of an inference-at-scale platform while providing infrastructure depth and a public-company financing platform. The strategy targets open-weight model adoption by emphasizing control, security, compatibility, and cost efficiency.
The transaction would add an established AI inference platform and engineering team.
Greater infrastructure depth is intended to support inference workloads at commercial scale.
Security and model control can differentiate open-weight inference from less flexible alternatives.
Geography
SAIHEAT is incorporated in Singapore and reports from a Singapore principal office, while its stated business model is global distributed computing infrastructure. Canopy Wave, the proposed combination partner, is based in Santa Clara, California and serves developers and enterprises through its AI inference and GPU cloud platform. No country-level or regional revenue percentages are disclosed in the available excerpts.
Risks
The business is exposed to execution risk because the Canopy Wave transaction, financing, and Nasdaq listing conditions must be completed before the planned combination takes effect. Its AI platform would depend on third-party open-weight models, computing infrastructure, and volatile GPU or per-token pricing, while limited operating history and customer concentration increase uncertainty. Capital needs, dilution, concentrated voting power, regulation, competition, litigation, and macroeconomic conditions could materially affect the combined company.
Closing depends on shareholder approval, Nasdaq approval, and completion of the private placement.
Exposure: Failure or delay could prevent the planned business combination.
The AI platform relies on infrastructure the combined company does not own and that may be terminated.
Exposure: Service interruptions or higher infrastructure costs.
Models developed outside the United States may be affected by export controls, trade restrictions, or procurement policies.
Exposure: Restricted model access or customer eligibility.
Declining per-token pricing or GPU rental rates could reduce revenue and platform economics.
Exposure: Inference and infrastructure pricing.
Dependence on a limited number of customers makes revenue vulnerable to contract loss or reduced usage.
Exposure: Revenue volatility and bargaining pressure.
Scaling infrastructure and completing the transaction may require additional financing.
Exposure: Shareholder dilution and funding dependency.
Substantially larger providers may have greater capital, infrastructure, and model-development resources.
Accounting
Judgemental itemsThe proposed transaction involves new Class A and Class B shares and negotiated pre-money valuations.
Impact: Could affect purchase accounting, ownership percentages, dilution, and transaction-cost recognition.
A planned approximately US$4.5 million private placement is part of the transaction structure.
Impact: Terms may affect equity classification, proceeds, dilution, and offering-cost presentation.
Distributed computing and GPU infrastructure require estimates of useful lives, residual values, and recoverability.
Impact: Changes in utilization, pricing, or technology could accelerate impairment charges.
The combination could create goodwill or identifiable technology and customer-related intangibles.
Impact: Future impairment testing could produce material non-cash charges if projected growth weakens.
SAIHEAT replaced Audit Alliance LLP with Assentsure PAC in March 2026 without reported disagreements.
Impact: Investors should monitor audit continuity, control observations, and future reporting judgments.
Similar listed companies
Publicly listed companies with a comparable business focus.
Bit Origin Limited
SGRX · Finance Services
CoreWeave, Inc.
CRWV · Services-Prepackaged Software
SuperX AI Technology Ltd
SUPX · Services-Management Consulting Services
SuperX AI Technology Ltd is an AI infrastructure and services company focused on hardware, software, and operational support for AI data centers.
Tuya Inc.
TUYA · Services-Prepackaged Software
Nexus Advanced Technologies Inc.
NXATW · Services-Allied To Motion Picture Production
Science Applications International Corp
SAIC · Services-Computer Integrated Systems Design
Last updated: 08/09/2026