The lab where GM is cooking up new EV batteries to beat China

Inside General Motors’ fast-growing battery labs in suburban Detroit, scientists and engineers are analyzing stresses on lithium-ion cells: desert heat, arctic cold, jungle humidity, enough charging and discharging for a half-dozen Frankenstein reboots.

For The Verge’s exclusive tour of these secretive labs, I watch researchers peer at cell chemistries down to the atomistic level, using electron microscopes. Others work at a larger scale, all the way up to the Megashaker. Inside a cavernous hall, an enormous sliding test chamber envelops one of GM’s double-stacked, 205-kilowatt-hour battery packs — the type that powers hulking models like the Cadillac Escalade IQ^[1].

The 2,900-pound pack is suspended in midair from a ceiling-mounted crane, bolted to a surrounding structure to mimic its connection to a moving vehicle. Like a lithium-soaked smoothie, the battery is hydraulically shaken in the atmospherically controlled chamber, simulating anything from potholes to low-speed collisions. With dozens of test chambers for individual cells, modules, or entire packs, GM can simulate 10 years and 250,000 miles of real-world battery use and abuse in about six months.

That real world has become especially abusive to EVs. Some consumers have grown ambivalent, in part due to stubbornly high prices. Pollution and fuel-economy rules^[2] face an anachronistic rollback, encouraging automakers to lean into fossil-fueled cars. In the latest indignity, the $7,500 federal clean-car credits, beloved by EV buyers and a wellspring of sales for GM and other automakers, have been choked off by the Trump administration^[3]. All that took a $1.6 billion toll on GM’s bottom line this week, in the form of a writedown on third-quarter earnings.

GM directly attributed that $1.6 billion loss to the Trump administration eliminating credits and loosening emissions rules. The dirty ripples from those moves will surely be more polluting internal combustion engine (ICE) cars on the roads, and fewer EVs, as GM and other automakers scale back production to match lowered expectations for consumer demand. For all their sleek new models and multibillion-dollar investments, GM and other automakers continue to lose money on the electric side of the business. And this was before all these new headwinds began to blow.

GM must somehow withstand these industry shake-ups and stresses. But even as it plays defense by shoring up its ICE business, GM plans to keep on the EV offensive. Inside its sprawling Technical Center, the classic midcentury campus designed by Eero Saarinen, GM brings a 21st-century weapon to light: lithium manganese rich batteries, or LMR.

Is LMR the real deal?

Some analysts and investors keep saying Detroit has little chance of competing with China on the global EV stage. GM executives and engineers beg to differ, even as the Trump administration kneecaps the top-down energy and manufacturing policies and government supports that China used to go from automotive also-ran to a dominant force in EVs and batteries.

GM says its affordable, more sustainable LMR batteries will decisively outperform China’s best lithium iron phosphate (LFP) specimens — currently the world’s favored low-cost solution. The company’s new cells should deliver one-third more driving range than LFP, at virtually identical cost. All things equal, that’s the difference between an EV that can cover 300 miles and one that can keep going for 100 miles more. Indeed, GM promises better than 400 miles of EPA-rated range from their largest SUVs and pickups when these batteries arrive in 2028. They’ll also save GM at least $6,000 per battery pack, versus today’s pricey high-nickel cells. That’s nearly enough, by itself, to offset the loss of those $7,500 credits and bring EVs closer toward elusive price parity with gasoline models.

“This unlocks premium long-distance range at an affordable cost,” says Andy Oury, a lead GM battery engineer.

Kurt Kelty, who helped transform the global industry as Tesla’s battery cell development chief, says these cells — unlike solid-state chemistries — aren’t the usual battery clickbait, promising on the page yet perennially stuck in the lab. Talking with more than a dozen battery executives, engineers, researchers, and scientists, their confidence was palpable: These batteries, they say, are tested, proven, and ready for mass production. GM is spending $900 million to build a pair of sprawling new battery centers here, on top of more than $5 billion already invested in US battery operations.

Kelty, GM’s vice president of battery, propulsion, and sustainability, has spent a lifetime in the field. He was at Sony in 1991 when the company debuted its revolutionary Camcorder, the first video recorder with a lithium-ion battery. He spent 11 years guiding Tesla’s battery tech as roughly Employee No. 50 at Elon’s fledgling outfit, then worked on the anode side for Sila Nanotechnologies before coming to GM.

“I’ve kind of seen it all at this point,” Kelty tells me. “So when I say this a game-changer, well, I’ve seen a lot of technologies come and go. Or usually not come. They have an announcement, and then they disappear. That’s kind of the standard for the battery industry.”

Where batteries tend to be about incremental improvement — a superior coating here, a better copper foil there — Kelty says these prismatic LMR cells represent a serious step change. One without the usual tradeoffs.

“That’s why we’re excited about it,” he says. “You can typically get high energy but kill your life cycle, kill your cost, or whatever. But in this case, it’s a really balanced chemistry.”

These batteries could stage an end-run around China in another way, relying more on a US-based supply chain that’s just getting off the ground. Asked about China’s current dominance, Kelty says GM knows what it’s up against. In a bitter irony for American innovation and manufacturing, that Nobel-winning LFP tech was developed right here at the University of Texas. But as American companies dithered, China helped itself to the patented technology and built itself an LFP empire.

“They violated the IP for years and years, but they only sold within China, so it was never an issue,” Kelty says. “Now those patents have expired, so they don’t have any problems exporting [LFP].”

“So you’ve got to keep in mind who your competitors are. And by using LMR, we can now actually exceed their LFP performance, but give it a similar cost.”

A hot ticket in batteries

Oury says LMR has been around for a decade, but researchers couldn’t overcome issues with voltage that petered out over time. Accelerating its development five years ago, starting with coin-size cells, GM solved the durability issues and steadily worked its way up to a larger prismatic form factor — stacks of thin cells, roughly the size of a laptop case — that will go into vehicles.

“We got the confidence to scale up to larger designs,” Oury says.

Among North American automakers, GM placed early bets on building its own “high nickel” or “nickel rich” batteries. That NCM (for nickel-cobalt-manganese) chemistry is the most energy-dense and currently powers nearly every EV sold in America. The automaker initially struggled to build those pouch batteries at scale, bringing production of some EVs to a crawl. But now those bets are paying off, giving GM a potential edge against EV latecomers, including tariff-tossed import brands with no domestic EV or battery factories. GM is now producing more lithium-ion cells than any automaker in North America, in partnership with LG Energy Solution at plants in Ohio and Tennessee.

Kelty says GM can produce those cells at lower cost than any domestic rival. Based in part on teardowns of roughly two dozen competitors’ packs here, Kelty says GM isn’t yet the lowest-cost producer at the pack level — but will reach that milestone with its next-generation EVs. For one, prismatic LMR packs will require 50 percent fewer parts.

High-nickel may be the long-distance champ, but it’s expensive, heavy on pricey nickel and cobalt. Cobalt is also ethically fraught, including sourcing from African mines where children perform hazardous work for meager wages. Those battery costs, roughly 30 percent of a typical vehicle’s cost, remain the number one factor in making EVs unaffordable for many Americans. It’s the reason Cadillac’s impressive Lyriq-V, the 615-horsepower SUV, starts at over $81,000. It’s why a Cadillac Escalade IQ costs $130,000, or a Chevy Silverado EV pickup starts at $75,000.

China’s answer was those LFP batteries, which sparked that nation’s EV revolution, especially for small budget models that dominate that market. Safe and durable, those batteries replaced cobalt and nickel with cheap, plentiful iron and phosphate. GM, Ford^[4], and Stellantis^[5] are among Western automakers now looking to build their own LFPs. The tradeoff is mediocre driving range and performance. The newest “Blade” LFP battery from China’s BYD delivers roughly 350 watt-hours per liter of volume. That compares with more than 600 watt-hours for GM’s high-nickel batteries, or Tesla’s 4680 cylindrical cells.

LMR’s innovation finds a middle ground through a healthy dose of manganese, the abundant silver-gray transition metal. The cells sacrifice only modest range versus the most powerful high-nickel designs. But they could be vastly cheaper and whip LFP’s driving range or performance. Ford has announced its own breakthrough in LMR.

In this chemistry competition, the action here is mainly in the cathode. That’s the positive battery electrode that generates electricity for a car or smartphone, and accepts electrons in a chemical reaction as the battery discharges.

Today’s NCM batteries contain up to 85 percent nickel in those cathodes, 10 percent manganese, and 5 percent cobalt. LMR batteries flip that. Their cathodes contain up to 70 percent manganese, about 30 percent nickel, and 2 percent or less of cobalt.

“Manganese is dirt cheap, so at a raw materials level, it gives you that benefit to start with,” says Kushal Narayanaswamy, GM’s director of advanced battery cell engineering.

GM in the house

Cheaper materials are only a start. Analysts say some Chinese companies are bringing new cars or batteries to market in as little as two years. That unrivaled pace may be China’s biggest advantage, more than the technology onboard their EVs. GM says its in-house operations could cut the time it takes to bring new batteries to showrooms by a full year, at lower costs and with higher quality.

I meet Narayanaswamy at the Wallace Battery Cell Innovation Center, a linchpin of GM’s DIY approach to batteries. It’s taking a page from the vertical integration that characterizes China’s BYD, the world’s largest maker of EVs and batteries. Here, GM is developing and testing its own chemistries — mixing black slurries, coating cathodes — without constant back and forth with outside suppliers that can slow projects to a crawl. The Wallace lab can now produce up to 100 cells a day, and has created 26 variants of LMR cells in three form factors. Next door, I see the Ancker-Johnson Battery Cell Development Center under construction. The 500,000-square-foot facility will house a small pilot assembly line for batteries, bridging the often daunting gap between labs and factory production.

The automaker has cut testing time for new batteries by 60 percent. Digital twins and virtual modeling simulate hundreds of real-world scenarios — from condensation in battery packs to catastrophic collisions — before GM builds a single physical cell or pack. The company has simulated more than 1.4 million miles of driving. Radu Theyyunni, global director of virtual electrification and powertrain, says the company’s physics-based models have logged 150 million hours of compute time, to optimize LMR from the atomistic level to full packs. That’s the equivalent of a single, powerful CPU running nonstop for 171 years.

An obvious question: If LMR is all that, why is GM messing around with LFP batteries at all, with plans for production in Tennessee? GM executives argue that LFP can still be the right fit for lower-cost, lower-range models. And LFP shows great potential for second-life uses in stationary energy storage.

Forget the cavalry — it’s up to automakers and consumers

With government credits gone kaput, GM has chosen to pick up those $7,500 tabs itself for now through subsidized lease deals, to keep EV sales from flatlining. That’s generous of GM, but another hit to its bottom line, on top of a tariff bill that could reach $5 billion by year end. Dealers are surely being asked to share the pain.

I ask Kelty if the end of credits is like tearing off the Band-Aid: painful, but ultimately necessary. The credits were never meant to last forever. Now automakers have nowhere to hide, no subsidies, and no excuses. Analysts say the move may separate EV contenders from pretenders. Only automakers who can deliver affordable, compelling EVs, at a profit, will compete or perhaps survive, including against China on the global stage.

“Would I prefer that had come later?” Kelty says of the precipitous end of credits. “Yes, I think we all would. But it forces us to go in the right direction. We’ve got to get our costs down.

“We did this at Tesla as well, to ignore all the incentives out there. You’ve got to be able to stand on your own. And that’s the way we’re looking at it here at GM.”

Kelty says EVs can already match ICE cars in total cost of ownership. But that hasn’t won over enough Americans who look at an EV and see only a higher monthly payment.

“So once you get the sticker price to be similar, then it’s game over for ICE in most applications. Because it’s a better experience when you’re driving an EV.”