AZIO AI, a next-generation artificial intelligence and high-performance computing infrastructure platform, and Envirotech Vehicles, Inc. (NASDAQ: EVTV) ("EVTV"), announced continued execution of a scalable AI infrastructure program designed to support sustained, real-world compute workloads using purpose-built cooling systems, self-owned behind-the-meter power, and modular deployment architecture.

The infrastructure is engineered to operate under continuous, production-level demand, generating live operational data that directly informs system design, efficiency, and economics as AZIO AI advances toward larger-scale multi-megawatt deployments.

Purpose-Built Infrastructure Designed for Scale
AZIO AI is deploying proprietary cooling and power systems engineered specifically for high-density AI computing. Unlike legacy data-center designs adapted for AI workloads, AZIO AI's architecture is optimized from inception for sustained thermal performance, power efficiency, and operational reliability as compute density increases.

By integrating cooling and power at the infrastructure level, AZIO AI aims to improve performance predictability while reducing dependency on third-party utility constraints that increasingly limit large-scale AI deployments.

Self-Owned, Behind-the-Meter Power Strategy
The infrastructure utilizes on-site, behind-the-meter power generation, providing AZIO AI with increased control over cost structure, uptime reliability, and expansion timelines. This approach is designed to mitigate grid congestion and long interconnection lead times that have become a growing challenge for AI-driven compute facilities.

The power strategy is intended to be repeatable across future AZIO AI sites, supporting faster deployment of additional capacity as demand scales.

Texas-Based Reference Deployment Supporting Global Infrastructure Development
AZIO AI's initial deployment is located in Texas, at operating oil field sites that produce on-site natural gas from underground reserves, which is used to generate power on a continuous, 24/7 basis through dedicated generation equipment.

The Texas deployment is structured as a reference operating environment, enabling AZIO AI to evaluate power generation, cooling performance, system reliability, and operational economics under sustained, real-world conditions. Rather than being site-specific, the infrastructure is designed to generate data and operational insights that can be applied across AZIO AI's broader platform.

Learnings derived from this deployment—including power management, thermal efficiency, uptime characteristics, and modular deployment practices—are intended to inform the design, deployment, and operation of future AZIO AI infrastructure across international markets, including locations where grid capacity, energy reliability, and deployment timelines present similar constraints.

By utilizing behind-the-meter natural gas generation and modular infrastructure in a controlled operating environment, AZIO AI is developing repeatable deployment frameworks intended to support future domestic and overseas data center and AI infrastructure projects.

Built for Continuous, Real-World AI Workloads
The system is designed to operate under sustained, full-time computing demand, enabling AZIO AI to measure real-time performance across power utilization, cooling efficiency, uptime, and system economics. Data gathered from live operations is used to refine infrastructure design, improve cost efficiency, and accelerate deployment of larger-scale facilities.

This operational data forms a core component of AZIO AI's broader infrastructure roadmap, supporting expansion from initial deployments to multi-megawatt and, over time, larger-scale AI compute campuses.

Modular Architecture Enables Incremental Expansion
The modular design allows compute and power capacity to be added incrementally while maintaining consistent performance standards. Initial configurations are capable of supporting approximately 500 kilowatts (kW) of compute load, with expansion pathways designed to scale to multiple megawatts as deployment milestones are achieved.

Based on typical high-density AI configurations:

• Approximately 500 kW can support 1,000–1,250 compute units
• Approximately 5 megawatts (MW) can support 10,000+ compute units
• Larger deployments in the tens of megawatts can support 100,000+ compute units, subject to final configuration and commercial arrangements

Commercial Framework with EVTV
Under the current structure:

• AZIO AI supplies and sells compute hardware
• EVTV owns the deployed compute assets
• Compute units are hosted within a managed hosting environment
• AZIO AI participates through hardware sales and platform-level infrastructure economics
• EVTV participates through infrastructure ownership and hosting economics

This framework is designed to align long-term incentives while supporting scalable deployment across future facilities.

Market Opportunity and Industry Context
Global demand for artificial intelligence compute and supporting infrastructure continues to accelerate, driven by the rapid adoption of large language models, enterprise AI deployment, sovereign AI initiatives, and increasing power constraints across traditional data center markets. According to publicly available research and commentary from leading global institutions including McKinsey & Company, Bloomberg, Goldman Sachs, NVIDIA (NASDAQ: NVDA), and the International Energy Agency (IEA), AI-driven data center expansion, power infrastructure, and advanced cooling requirements represent a rapidly expanding global market measured in the tens of billions of dollars annually, with continued growth expected as AI workloads scale and energy-efficient, behind-the-meter architectures become increasingly critical.

AZIO AI's vertically integrated approach—combining purpose-built cooling, self-owned power, modular infrastructure, and high-density compute—positions the Company to participate in this expanding market as deployments scale across future facilities.