Powering the next century of growth
Amazon is funding a 960 MW nuclear plant in Indiana that nobody has decided to build. The joint development agreement locks in a bankable structure, but not a budget, a schedule or a final investment decision.
A verified 1 keV electron temperature is a real first for magnetized target fusion. It is also a tenth of the way to the next milestone, with liner cost, tritium breeding and pulse economics still unproven.
Texas turned grid-forming inverters from a vendor feature into an interconnection condition. New storage must pass plant-level model and unit tests, and a $1,500/MW retrofit pool reaches only about a third of the 50 GW now exempt.
Voltus raised $225M to scale its virtual power plant to 20 GW by 2030. One day later, FERC rejected PJM's fast lane for data-center load. The two events point at the same thing: grid capacity is now the binding constraint, and whoever aggregates flexibility owns the short-term answer.
Linkener closed a €7.5 million round led by Klima to consolidate Spain's fragmented business energy management market, pairing proprietary metering and billing data with AI and acquisitions.
RX raised 10 billion won in a pre-Series A bridge, with Korea Development Bank, Smilegate and Kiwoom joining. The RX-50 aims at industrial heat rather than electricity, but five numbers stay unpublished.
Epyr turns cheap midday electricity into stored industrial heat. With €15M from SlateVC and a first reference at WEPA's Troyes paper mill, the bet is that a firebrick can undercut a gas boiler.
Northvolt's collapse stranded a recycling process that already ran at industrial scale. Its former engineering team bought back the 134 patents, relaunched as Refinyx, and is betting circular battery materials can compete on cost alone.