EdgeStart is an integrated 8-GPU NVIDIA cabinet designed to enable building owners to generate income by utilizing unused solar energy and batteries.
New York, October 27th / PRNewswire / -- Distributed energy and artificial intelligence company Lektra, Inc. Starting today, they are accepting EdgeStart orders. This is an integrated server cabinet designed for commercial and industrial building owners, photovoltaic and energy storage sites, as well as energy developers. It can be installed in underutilized spaces within facilities and is powered by existing solar arrays and batteries. Lektra will install the system, operate it remotely, and sell its computing power through Lektra Cloud to AI customers for tasks such as inference and other GPU workloads, while paying the owners 80% of the revenue generated.
Many buildings equipped with solar panels generate more electricity than they consume themselves. In some markets, the prices of utility electricity and grid connection rules mean that selling excess solar power can only earn a few cents per kilowatt-hour. Lektra states that, with a patented process, EdgeStart provides owners with a more efficient way to generate additional income. Each cabinet contains an 8-GPU NVIDIA server (the same type of hardware as that rented by AI from large cloud service providers), an online UPS, a distribution unit with metering capabilities, firewalls, and switches. Real-time dashboards display usage, energy consumption, and earnings, while Lektra is responsible for customer service, billing, monitoring, and support. Owners can start with just one cabinet and expand as needed.
Lektra Founder and CEO Karl Andersen stated: “EdgeStart is the first building block in our vision, aimed at creating a more resilient and influential computing network where electricity, whether from solar power, batteries, or microgrids, is readily available. AI Infrastructure does not have to be confined within the walls of large data centers. Enterprises across the country have energy, available space, and the capability to support computing. EdgeStart connects them into a coordinated network, allowing them to be compensated for supplying power to AI computations.”
The development of large-scale centralized data centers can take several years, as they require extensive construction, licensing, land acquisition, coordination with utility companies, and access to a reliable supply of electricity. In addition, political and community consultations surrounding such projects often pose obstacles, as they raise concerns regarding electricity prices, water usage, land use, and the capacity of power grids. EdgeStart offers a flexible and opportunistic approach to deploying computing in buildings that can easily support it or that are already equipped to do so.
EdgeStart is the entry-level product of the Lektra EdgeScale data center product series, which also includes Edge Pro and Edge Max. Edge Pro is designed for larger-scale commercial deployments, while Edge Max supports installations of over megawatts. The EdgeScale data centers powered by Lektra have currently been deployed in 8 locations, with each site not exceeding a capacity of 20 MW.
Lektra launches two models. The EdgeStart RTX is equipped with 8 NVIDIA RTX PRO; it has a price of $310,197 and a peak power consumption of about 7 kilowatts. It is estimated that the owner's monthly income in the first year will be around $9,700. The EdgeStart B300 is also equipped with 8 NVIDIA B300 GPU; it is priced at $840,951 and has a peak power consumption of about 14.5 kilowatts, with an expected owner's monthly income of around $36,950. The price includes the cost of the installed hardware and transportation fees.
Homeowners can choose to pay in cash or finance through Blue Street Capital with a term of 48 months, a down payment of 20%, and an annual interest rate of 9%. Credit approval is required. The estimated monthly payment is approximately $6,200 for the RTX model and about $16,700 for the B300 model.
Andersen continues to state: "Uber has created a sharing economy for transportation, and Airbnb has done the same for the real estate industry. However, energy—one of the most important resources in the world—has largely not yet undergone such a transformation. EdgeStart is the first step towards a sharing economy, allowing owners of energy infrastructure to become active participants in the AI economy in a reasonable and prudent manner, rather than merely being customers of the power grid."
Lektra refers to the building owner of EdgeStart as energy hosts. host first conducts an on-site assessment at lektra.com, followed by project review and revenue model analysis. Subsequently, host can purchase the system with cash or financing, after which Lektra is responsible for installation, testing, and go-live.
EdgeStart is now available for purchase at lektra.com. Buyers can also apply for an on-site assessment or contact [ email protected ].
The launch of EdgeStart is part of Lektra's continuously expanding portfolio of distributed AI infrastructure solutions. The initial configuration will support some NVIDIA GPU architectures, and in the future, more hardware options and deployment configurations are also planned to be provided.
Lektra, Inc is a distributed energy and artificial intelligence infrastructure company that is developing a network of computing resources powered by distributed energy. By integrating GPU computing, energy management technologies, and cloud orchestration, Lektra enables enterprises and energy owners to participate in the AI infrastructure economy. Its EdgeStart, Edge Pro, and Edge Max solutions are designed to support scalable deployment in commercial, industrial, and distributed energy environments.
Forward-looking Statements
This press release contains forward-looking statements regarding product development, commercial availability, market opportunities, expected performance, and potential revenue generation. Actual results may vary significantly from expectations due to market demand, GPU utilization rates, power costs, equipment supply, operating expenses, and other factors. The examples of computing power revenue mentioned in this text do not constitute guaranteed earnings or investment returns.
