Samsung SDI Buys Out GM’s Stake in the $3.5 Billion Indiana Cell Plant and Turns It Toward Grid Storage
- GM stake acquired: 49.99%
- Joint venture investment: $3.5 billion
- First-phase capacity: 27 GWh per year
- Mass production start: 2027
Samsung SDI has bought General Motors out of SynergyCells, the battery joint venture the two companies put up in New Carlisle, Indiana. The Korean cell maker disclosed the purchase in a regulatory filing on August 11, 2026, and named no price. GM held 49.99 percent. The site carries roughly $3.5 billion of committed investment and about 1,600 jobs. Its first phase is rated at 27 GWh a year. Samsung SDI now plans to open it on stationary storage cells rather than the electric-vehicle cells the venture originally targeted.
The deal dates to 2023 and closed in 2024. It centered on nickel-rich NCA prismatic cells under the PRiMX brand, with mass production targeting 2027 and a stated path to 36 GWh. Samsung SDI tied the ownership change directly to demand, saying the shift reflected “the slower-than-expected growth of EV demand.”
The plant does not stop, it changes product
Samsung SDI called New Carlisle its first independently operated battery production base in North America. An energy storage line comes first. That choice has a signed contract behind it. In December 2025 the company signed a three-year ESS supply deal worth more than 2 trillion won. That is about $1.35 billion, with North American deliveries starting in 2027.
GM did not walk away from the chemistry. The two signed a fresh joint development agreement covering prismatic cells for possible future EV use. GM keeps a claim on the technology and a supply relationship. It no longer carries half of a factory it does not need yet.
Why It Matters
A 27 GWh line pointed at storage is a large addition to US grid-battery supply, and it arrives from a plant underwritten for cars. Developers of utility-scale and commercial storage have spent two years short of domestic cells. Tariff and domestic-content rules sharpened that squeeze. A dedicated North American ESS line eases both problems. It eases them in Indiana rather than overseas.
Read it the other way and the signal is sharper. One manufacturer weighed both markets against 27 GWh of nameplate capacity and judged storage the safer bet. Anyone sizing charging load or fleet infrastructure on forecast EV volume should note who just moved, and how far. The 2027 deliveries also carry a prior claim, so buyers shopping for 2027 cells are shopping behind an existing contract.
Critical Perspective
Samsung SDI has not said which chemistry the storage cells will use, and that gap carries the whole 27 GWh. The venture was tooled for nickel-rich NCA prismatic cells aimed at cars, while most US grid-storage procurement now runs on lithium iron phosphate, the chemistry Tesla and CATL ship into the same market. Retooling a line is not a product decision announced in a filing. The $1.35 billion ESS supply deal also predates this purchase by eight months, so it commits volume without proving the Indiana plant is the source. GM walked away from 49.99 percent of a factory it helped specify and kept only a development agreement. How much of the 27 GWh is genuinely ESS-capable in 2027, and on which chemistry?
Sources
- Electrek (2026-08-11)
- WardsAuto (2026-08-12)
- Charged EVs (2026-08-14)