Everything got faster.Except the battery.

A while back I read Empires of Light by Jill Jonnes. It's the story of Edison, Tesla and Westinghouse tearing each other apart in the 1880s over how to light the world — Edison's system against Westinghouse and Tesla's, fought out in New York, at the Chicago World's Fair, and finally at Niagara Falls. What the whole war was really about was delivery. How do you get electricity from the place it's made to the place someone needs it? Their answer was wire. Miles and miles of wire. Electricity would be everywhere, as long as everywhere had a cord attached.

But the book opens with the older names, and one of them stayed with me. Alessandro Volta, in 1800, stacked discs of two different metals with brine-soaked cloth between them and produced a steady current out of a thing sitting on a table. No generator, no waterfall, no wire to anywhere. More than eighty years before Edison lit lower Manhattan, someone had already worked out how to hold electricity still and carry it around. That is the part I still find remarkable. Energy that fits in your hand.

Today they're in everything, and we barely think about them. When one stops working we throw it out and buy another. But in the last two years we started asking far more of them than we used to. Phones now run models that think, and thinking costs power. Smart glasses hold a camera, speakers and an assistant that listens all day, inside a frame with room for a battery about the size of two sticks of gum.

Everything around the battery has improved by leaps. The chip in your phone is unrecognisable from ten years ago. The screen, the camera, the software. The battery has improved by a few percent a year — and it's now the thing holding the rest back.

I've been watching one company that thinks it has the answer since the summer of 2024. I found it at around $8. Within four months it ran to $18.68 and I didn't touch it, because a story is not a business and I had no proof they could actually build the thing. It had opened that year at $27.16, so plenty of people were already underwater by the time I showed up.

On Monday it traded at the lowest price in its life as a public company, and closed down 18% on the day. A price like that usually means something is broken. Quite often it is. So I went and looked properly.

This is worth understanding properly, so here it is in plain terms.

Inside almost every rechargeable battery is a piece called the anode, and it's made of graphite — the same soft grey stuff in a pencil. Think of it as a parking garage for lithium. When you charge your phone, lithium drives in and parks. When you use your phone, it drives back out. That's the whole trick.

Engineers have known for thirty years that there is a far better garage available. Silicon — the same element in sand and in computer chips — holds roughly ten times more lithium in the same weight. Ten times. In an industry that celebrates 5% gains, that number has sat there for three decades like money on a table.

Nobody could pick it up. When silicon takes on lithium, it swells to about three times its size. Then it shrinks again. Do that a few hundred times and the material cracks, breaks contact with the rest of the cell, and the battery dies. The industry's compromise was to sprinkle a little silicon into the graphite — enough to gain something, not enough to tear the cell apart.

So the fix has always been obvious and always been unusable. Not a chemistry problem. A structural one. Somebody had to find a way to hold silicon still while it tried to burst.

The company is Enovix.NASDAQ: ENVX It was founded in Fremont, California, and it has been trying to solve exactly that problem for nineteen years.

Their answer wasn't a new material. It was a box.

I'm not an electrochemist. For the science underneath this I lean on a couple of research services I pay for, and I'd rather say that than pretend.

Instead of rolling up flat sheets the way a normal battery is built, Enovix stacks the layers and clamps the whole assembly inside a stainless steel constraint — a rigid frame that physically refuses to let the silicon expand outward. The silicon still swells. It just has nowhere to go. That is the entire thesis in one sentence: because the structure holds the material still, Enovix can build an anode that is 100% silicon, while every competitor is stuck blending a little silicon into graphite and hoping.

Chinese phone makers ship the best blended cells on the market today and they cap out somewhere around 25% to 32% silicon before the swelling wins. Enovix is at 100%. In August of last year, Polaris Battery Labs — an outside testing house, not the company's own lab — confirmed Enovix's smartphone cell as the highest energy density cell ever reported for a phone. Energy density just means how much power you can pack into a fixed amount of space, which is the only measure that matters when the space is the inside of a phone.

But lab results don't bring in any money. A battery has to be manufactured, certified, and sold to customers who will sue if it catches fire. So I went looking to understand what this company actually is, not just what it has invented. It was not what I expected.

In late 2023, Enovix bought a Korean battery manufacturer called Routejade for roughly $98M — mostly in its own stock, which was expensive at the time. Routejade sits in Nonsan, South Korea. It had already been making lithium-ion batteries for more than twenty years. Its customers were medical device makers, industrial buyers, and the South Korean military.

Then in April 2025, Enovix bought the plant next door — a 330,000 square foot battery factory that SolarEdge was walking away from — for $10M in cash. The assets were worth more than the price, so the accountants recorded a $4.8M gain on the purchase. They bought a working factory for less than the sum of its parts.

That Korean operation is now the company. In the second quarter, 70% of all revenue was billed to South Korea. Defence and drone customers were 65% of the quarter.

Say that slowly, because it's the thing most write-ups skip: the batteries generating the revenue are not the 100% silicon batteries. Korea builds conventional cells with some silicon blended in — good ones, rugged ones, ones you can buy today. The moonshot is built in Malaysia and has barely sold anything. One ticker, two businesses.

And the Korean one has a tailwind with a date printed on it. Under the 2026 defence bill, the Pentagon is barred from buying batteries made with materials from Chinese or Russian entities of concern starting January 1, 2028. China currently makes something like three quarters of the world's lithium-ion cells. Enovix's expanded Korean capacity is scheduled to come online in mid-2027 — about six months before that door closes.

The company says its pipeline for Korean-made products has grown to $183M from $130M in a single quarter. I want to be careful here, because that number gets quoted as if it were an order book. It isn't. It's the company's own estimate of what the opportunities it has identified would be worth annually if they all landed. Treat it as a direction, not a dollar.

That's the floor under the stock. The reason I finally bought sits on the other side of the company.

A pair of glasses is the cruellest place in consumer electronics to put a battery. No room, no weight allowance, and the whole thing sits on someone's face, which means it cannot get hot and it cannot swell. If a battery survives there, it will survive anywhere.

The market is arriving fast. The research firm IDC counts about 13.6M smart glasses shipping this year, up from 2.7M in 2024, heading toward 27.3M by 2030 — and it identifies battery life, not the quality of the AI, as the thing holding the category back.

Enovix has an order for 50,000 battery packs this year from a large customer it won't name. In the second quarter they shipped about 2,100 of them, cleared international safety certifications, and passed the customer's full punishment routine — drops, tumbles, thermal cycling, extended heat. The customer has since ordered about 19,000 packs for the third quarter, with the rest due in the fourth.

Small numbers. Deliberately small numbers. The revenue from it was, in the company's own word, modest, and ramping it up is currently making the profit margin look worse rather than better. That's not the point. The point is that this is the first time a 100% silicon anode battery has gone through commercial production, safety certification, and a demanding customer's testing lab and come out the other side as a product. It can be built. That was genuinely in doubt.

Which brings us to phones, where the real money is.

There is now one accelerated test left, expected to finish this quarter, then a small build of real phones for field testing, then volume production targeted for 2027. A second phone maker gets its first samples this quarter.

Now the part I'd want to know if I were reading this. That launch date has moved twice. In early 2025 the company said mass production late that year. By autumn it was 2026. It is now 2027. And the final test itself was rewritten mid-process, replacing the standard industry test with a new one designed jointly with the customer. Enovix says the old test was built for graphite and doesn't reflect how silicon behaves. A sceptic says they couldn't pass the old test so the test changed. Both readings are fair. What tips it for me is that the 1,000-cycle result came out of the customer's mouth, not the company's.

Second quarter revenue was $9.0M, up 21% from a year ago and the fifth straight quarter of growth. First half revenue was $16.6M, up 32%. It was also the seventh quarter in a row where they sold each battery for more than it cost to build. That sounds obvious, but it isn't — plenty of young manufacturers sell things at a loss for years to win customers. Building a battery and selling it for more than the materials and factory time cost is the first hurdle, and Enovix has been clearing it since 2024.

Now the honest half.

That gap between what a battery costs to build and what it sells for has been getting thinner — 14.4 cents on every dollar of sales, down from 26.0 cents a year ago, because more of what they sold was the cheaper Korean product. And that gap has to cover everything else: the engineers, the factories, the research. It doesn't come close. The company lost $43.1M in the quarter and $81.3M in the first half. Since it began operations, it has lost $1.06B.

The number I see repeated everywhere is $552.1M of cash. It's true and it's misleading. Against it sits $532.8M of debt — $172.5M due in May 2028 and $360M due in 2030. These are convertible notes, meaning the lenders can swap them for shares instead of cash if the stock is high enough. It isn't close. So once the debt is set against the cash, what's actually left over is about $19M, and that 2028 bill is a real bill. In the first six months of this year they burned through $67.7M.

At today's price the whole company costs about $836M. It sells roughly $33M of batteries a year. So you're paying about 25 times current sales for it — and that's sales, not profit, of which there is none. This is not a cheap stock by any normal measure.

One more limit is worth understanding, because it's the one I see people miss.

The 100% silicon batteries are made in a single factory in Malaysia, on one production line. That line is built to produce about 1,350 batteries an hour, and it isn't even fully staffed yet. Run it non-stop — day and night, every day of the year — and you get somewhere around 11M batteries. One popular phone model needs tens of millions.

There is also a shareholder lawsuit still working through the California courts, filed back in 2023 over what the company said about its manufacturing progress. That fits an uncomfortable pattern: this is a company with a long history of describing things as closer than they turned out to be.

Rodgers joined the Enovix board in 2012, back when these cells managed about 10 charge cycles before dying. Last week they passed 1,000, verified by a customer with every reason to be strict. He called it the event of the decade for the company, and given that he has watched all fourteen years of it, I'm inclined to let him have the phrase.

Here's what I keep coming back to. In the space of five days, this company answered the question it had spent nineteen years failing to answer — and then lost its chief executive. The market priced the second event and ignored the first. That happens. It's most of what I look for.

What I'm actually buying at $3.81 is three things. A twenty-year-old Korean factory with real customers and an American law working in its favour from January 2028. A 100% silicon battery that has now been built, certified, and shipped to a demanding customer, in small numbers, but out the door. And one remaining test.

I didn't catch the low. It went to $3.52 after I bought and closed at $3.60, and it may well go lower — a company without a permanent chief executive tends to drift until it has one. I've been watching this thing for two years waiting for a day bad enough to make the price sensible. Monday was bad enough.