Wave Power Twin Forecasts Output for Coastal AI
The model is at the research stage, and its next phases aim to link wave energy forecasts to data-center demand.
Updated on Nov. 3, 2026 in Data Centers
Live Poll
Do you support using clean energy to power AI data centers?
Eco Wave Power has completed the first phase of an AI-powered digital twin program developed with AI engineering, modeling how ocean waves interact with its proprietary floaters. The development is at the research and development stage; the companies have not reported a deployed commercial system.
Why it matters
Coastal AI data centers need steady power, and wave energy is variable, so a digital twin that forecasts wave energy output and structural loads could help match generation to demand. The effort reflects a broader push to pair computing infrastructure with alternative power sources near the coast.
The first phase covered simulations of wave loading, floater movement, structural forces, and potential electricity generation under varying sea conditions. No benchmark or accuracy figures were disclosed, so the model's predictive performance remains unquantified.
The players
Eco Wave Power
A wave energy company that operates Israel's first grid-connected wave energy power station and has launched an onshore wave energy pilot at the Port of Los Angeles.
NVIDIA
A chipmaker whose Omniverse and Warp software and accelerated computing platforms underpin the digital twin's simulation and machine-learning work.
The details
The initial phase built a physics-based digital model, meaning a software replica that solves the equations of fluid motion and structural mechanics for how ocean waves act on Eco Wave Power's floaters, the buoyant structures that capture wave movement. The team added early machine-learning models to forecast energy output and structural loads, and tested how the model could transfer to future sites with different wave conditions. The program runs on NVIDIA Omniverse for visualization and simulation, NVIDIA Warp for modeling, and NVIDIA-accelerated computing for the machine-learning work.
Timeline
The first phase of the digital twin program has been completed; the source does not give a completion date.
Proposed next stages, including integration of operational sensor data, follow the completed initial phase.
The Tech Race
The project applies NVIDIA's Omniverse digital twin platform, widely used for industrial simulation, to a marine energy problem. It positions wave power as a candidate supply for coastal AI data centers, a niche where developers are still testing whether variable renewables can be paired with round-the-clock compute loads.
The work is at the modeling stage, so no wave-powered data center service is available to users yet. Any practical effect would depend on whether the next stages of sensor integration and deployment are completed, a timeline the companies have not specified.
The takeaway
The digital twin is a step toward forecasting wave output well enough to schedule compute against it. Watch for the publication of accuracy results against operational floater data, which would show whether the model's forecasts hold up in the field.
What happens next
The companies propose expanding the model to the hydraulic and mechanical energy-conversion system, then integrating operational sensor and generation data with a broader machine-learning layer; no dates were given for these stages.
Further reading
For the infrastructure side of this power question, see our coverage of Data Centers.
Source note: This article includes information reported by Offshore Engineer.
Live Poll
Do you support using clean energy to power AI data centers?
