On June 9, 2026, in San Francisco, GM announced three energy-storage deals at once: a sodium-ion battery co-development program with startup Peak Energy, LFP cells supplied to LG Energy Solution for integration into storage systems, and a 7.2 MWh second-life battery system purchased from Redwood Materials. As TechCrunch noted, it is GM’s first formal step from vehicles into the grid and data center market.
The timing matters. GM is the third automaker to make this move, after Redwood Materials entered storage in 2025 and Ford followed in December of that year. The driver is the same across all of them: AI data center power demand is growing faster than utilities can build transmission and generation, which turns storage into must-have infrastructure.
Three Deals, One Direction
The first deal is the sodium-ion partnership. GM will co-develop sodium-ion cells with Bay Area startup Peak Energy at its Battery Cell Development Center in Warren, Michigan, targeting trial production in 2028. TNW reports Peak Energy has $100 million in backing, a pilot line in Escondido, California, and plans for a factory with 10 GWh of annual capacity.
The second is a bridge: GM will supply its own LFP (lithium iron phosphate) cells to LG Energy Solution, which integrates them into finished systems sold to data centers and utilities — a way to occupy the market before sodium-ion is commercially ready.
The third expands the Redwood Materials relationship. GM is buying a 7.2 MWh storage system for a Michigan facility, which TechCrunch estimates will save roughly $3 million over its life, and will route about 10,000 used EV battery packs into Redwood’s recycling pipeline. Redwood already operates a 12 MW / 63 MWh microgrid at a Crusoe data center in Sparks, Nevada — the largest second-life battery deployment in North America.
Why Sodium-Ion
Sodium-ion chemistry swaps sodium, iron, and manganese for lithium, cobalt, and nickel. The raw materials are cheap, the supply chain is not controlled by China, and the fire risk is far lower — low enough that Peak’s system design omits cooling and fire suppression entirely. The tradeoff is energy density: TNW cites roughly 120–160 Wh/kg, about half of automotive lithium-ion cells at 250–300 Wh/kg. That rules out vehicles, but for stationary cabinets it barely matters.
GM battery VP Kurt Kelty’s framing is that sodium-ion “will be a defining chemistry for grid-scale energy storage systems.” GM has committed $900 million to new battery chemistries since 2022, and the Warren development center opened in 2024; the company expects it to pull commercialization forward by about a year. Kelty was also candid about the entry strategy: “The way we’re getting into the market is the easy way, through ESS.”
Data Centers Are the Real Buyer
All three deals point at the same customer profile: AI data centers whose power bills and interconnection queues are out of control. The Crusoe microgrid in Sparks already proves the commercial model, and the LG supply line names its target customers directly — data centers and utilities.
When compute contracts are priced at $920 million per month, as in the Google–SpaceX deal, power is the obvious next bottleneck. GM’s bet is that data centers will pay a premium for firm, dispatchable power, and that an automaker’s cell capacity and chemistry expertise are monetizable assets in that market.
Competition and Risks
GM is not entering empty territory. Tesla Energy, Fluence, and BYD are already established in storage. CATL and BYD have shipped sodium-ion cells in China, but TNW notes the technology has yet to prove it can survive the 15–20 year cycle life utilities require. The domestic precedent is shaky too: startup Natron Energy shut down when its funding dried up.
Peak Energy’s own track record is a 3.1 MWh sodium-ion system completed near Denver in 2025. Benchmark Mineral Intelligence estimates sodium-ion’s North American market share is essentially zero today and will stay under 1% through 2030. GM, for its part, declined to disclose how much it is investing in the storage push.
What It Means for Developers and Infrastructure Teams
Three practical signals. First, automakers are becoming energy suppliers: teams planning new AI facilities gain another class of storage vendor to evaluate, backed by automotive-scale manufacturing. Second, mind the timeline — sodium-ion trial production starts in 2028 and commercial volumes come later, so it does nothing for the 2026–2028 power crunch; near-term relief still comes from LFP, gas, and demand response. Third, supply-chain de-risking pressure will keep lifting the priority of China-light chemistries; storage procurements built around China-dominated chemistries may face growing scrutiny in future tenders.
Sources
- GM joins race to build batteries for AI data centers and the grid — TechCrunch
- From cars to data centres, GM pushes into energy storage with three new battery deals — TNW
- Why GM Is Betting on a Future With Sodium-Ion Battery Storage — Mother Jones
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.