General Motors says silicon is the anode material it's betting on next, and the timing tells you why. While solid-state cells stay stuck in lab prototypes, batteries that swap some of their graphite for silicon are already shipping in phones, hypercars and a growing list of luxury EVs. That gap between promise and product is what's pushing carmakers and cell startups to pour money into a chemistry that, until recently, most engineers treated as too fragile to trust.
Kurt Kelty, GM's vice president of battery and sustainability, put it plainly at the company's Empower conference in San Francisco last month. "We believe silicon is the next anode technology," he told InsideEVs. "We're definitely deep on silicon." He added that buyers should expect "silicon anodes being deployed in greater percentages" over the short to mid term, well before solid-state cells reach dealers.
Why the anode is the bottleneck
The anode stores lithium ions while a battery charges and releases them as it discharges. For decades that job has fallen almost entirely to graphite. It's cheap, stable and dense enough to have carried lithium-ion batteries from camcorders to freeway-capable EVs. It also carries baggage: graphite mining is dirty and costly, and more than 90% of the world's processing sits in China. That concentration has become a supply-chain headache for U.S. and European manufacturers trying to build batteries at home.
Silicon fixes the performance side of the equation. A silicon atom can hold far more lithium than the same mass of graphite, so a silicon-heavy anode packs more energy into the same space. The catch has always been swelling. Silicon expands and contracts as it takes on and sheds lithium, cracking the anode and shortening cell life. That's why the practical path isn't pure silicon but a blend that trims the graphite and adds enough silicon to boost capacity without tearing the cell apart.
The numbers cell makers are quoting explain the enthusiasm. Amprius Technologies, based in Fremont, California, claims an EV rated for 310 miles on a conventional pack could reach 574 miles using its silicon anode design. Sila Nanotechnologies says its high-silicon anodes can add 20% to range with no increase in pack size. Neither figure requires a new form factor or an exotic electrolyte, which is the whole point.
Already on the road
The clearest sign that silicon anodes have left the lab is that you can buy cars using them. Mercedes-Benz installed silicon-containing anodes in the new AMG GT, and the company says the pack can charge from 10% to 80% in about 11 minutes at a peak rate of 600 kilowatts. Owners on enthusiast forums note the same chemistry is spreading across Mercedes' newer electric models, not just its halo performance cars.
Motorsport got there first. The McMurtry Spéirling, the fan-car that set a record hill climb at the Goodwood Festival of Speed, runs on cells from Taiwan's Molicel built with Group14's silicon anode material. The result is enough discharge power to hit 60 miles per hour in 1.55 seconds. That's a demonstration of raw power rather than range, but it shows silicon cells can dump energy fast without failing.
High-end smartphones have quietly normalized the technology too. Several flagship handsets now ship with silicon-carbon batteries that squeeze more capacity into thin bodies, giving cell makers real-world cycling data across millions of devices before the chemistry scales up to vehicle-sized packs.
The manufacturing race
Performance claims mean little without factories, and that's where the recent activity has concentrated. Sila's plant in Moses Lake, Washington, is already running, with initial capacity to supply anode material for up to 50,000 EVs a year. The company holds supply agreements with Mercedes-Benz and with Panasonic, Tesla's main cell partner, and says it could expand the site to serve 2.5 million vehicles if demand holds.
Group14 has started producing silicon anode material at a plant in South Korea after taking full ownership of a joint venture with SK Inc. That facility is slated to turn out up to 10 gigawatt-hours of material, enough for more than 100,000 EVs. The company frames domestic and allied production as a hedge against the graphite chokepoint in China.
Amprius has taken a lighter-weight route. In February the company signed its first U.S. manufacturing partner, Nanotech Energy, to build its silicon anode cells domestically. The deal targets defense and aerospace buyers, including L3Harris Technologies, who need supply chains that satisfy federal sourcing rules under the updated National Defense Authorization Act. The two firms refined Amprius' SA128 cell, a 21700 cylindrical format rated at 6.8 amp-hours and 320 watt-hours per kilogram. That energy density sits well above typical graphite cells and pushes Amprius' total capacity beyond 2 gigawatt-hours across its Asian and U.S. sites.
GM hedges across chemistries
GM isn't going all-in on any single approach, and Kelty was upfront about that. The company still leans on high-nickel cells across most of its lineup, uses lithium-iron-phosphate in the Chevy Bolt, and is developing a lower-cost lithium-manganese-rich cell aimed at large SUVs and pickups arriving in 2028. It recently added sodium-ion batteries for grid-scale storage to the mix, a reminder that different jobs call for different chemistries. Readers tracking that side of the market can follow our ongoing Battery Tech coverage for the grid storage angle.
Solid-state stays on the watchlist rather than the roadmap. "We've got a bunch of solid-state prototypes in our labs," Kelty said, describing active testing but no production timeline. Most automakers don't expect solid-state cells ready for sale before the end of the decade, which is exactly the window silicon anodes aim to fill.
What to watch next
The open question isn't whether silicon anodes work. Cars on the road settle that. The question is cost. Today's silicon cells show up first in performance models and premium trims where buyers absorb the premium. Getting the chemistry into mainstream EVs means driving the price down and proving the cells survive years of daily charging without the swelling problem eating away at capacity.
The scale-up underway suggests the industry believes it can clear that bar. Between Sila's Washington plant, Group14's Korean output and Amprius' U.S. contract manufacturing, the supply base for silicon anode material is filling out faster than solid-state's ever has. For an industry that has spent years waiting on a battery breakthrough that keeps slipping, a chemistry that's already in showrooms looks like the safer bet.
Detailed specifications and partner information are available through Amprius' investor materials at ir.amprius.com.