Who Sells the Machines That Drill Glass? The Picks-and-Shovels Play
Who Sells the Machines That Drill Glass?
The Picks-and-Shovels Play
Everyone's betting on who makes the best glass substrate. But during a gold rush, the surest money is in selling shovels. Part 1: the bottleneck, and the machine makers behind every pilot line — LPKF, SCHMID, TRUMPF, and Philoptics.
In 1849, the people who got reliably rich off the California gold rush weren't the prospectors knee-deep in icy rivers. They were the ones selling them picks, shovels, and denim jeans. The glass substrate boom is shaping up exactly the same way — and the "shovels" here are some of the most precise machines ever pointed at a sheet of glass.
Here's the thing about the glass substrate story everyone keeps telling. It's always about the substrate makers: Intel, Samsung Electro-Mechanics, SKC's Absolics, TSMC. Will glass replace ABF? Whose pilot line ramps first? Those are great questions. But they're prospector questions. The quieter, arguably better, question is this: no matter which of those giants wins, who do they all have to buy their equipment from?
Because glass, it turns out, is a nightmare to manufacture. And the companies that solved the hardest steps — especially the one step that breaks most often — get paid by everyone, regardless of who ends up on top. That's the picks-and-shovels play. This first part introduces the bottleneck and the cast of machine makers. Part 2 turns it into an actual investment thesis — including one small German company doing something its income statement says it absolutely should not be able to do.
The Picks-and-Shovels Logic
Let me lay out why equipment is such an interesting place to look. When a brand-new manufacturing technology is in its pilot-to-mass-production transition — exactly where glass substrates sit in 2026 — nobody knows yet which substrate maker will win the volume game. Yields are still wobbly. Customers are still qualifying suppliers. It's a knife fight.
But every single one of those competitors needs the same physical tools to drill, etch, image, and cut the glass. The substrate makers are placing risky bets on demand. The equipment makers are selling them the means to place those bets. One sells lottery tickets; the other sells the pens.
A typical greenfield glass production line costs a fortune to tool up — market researcher Mordor Intelligence pegs new fusion-draw or float lines at USD 200–500 million each, with the furnace alone eating up to 40% of that. Multiply that across Intel, Samsung, TSMC, Rapidus, DNP, and a dozen others, and you start to see why the toolmakers are smiling.
What's striking is that the glass-core substrate market itself is still tiny today. According to Yole Group and QY Research estimates (cited by industry analyst PhotonCap), the glass-core packaging substrate market was only around USD 244 million in 2024, projected to reach roughly USD 460–586 million by 2030. That sounds small — until you remember it's being pulled along by Mordor Intelligence's estimated USD 150 billion wave of new fabs from Intel, Samsung, and TSMC through 2030. The substrate is the appetizer. The capex is the feast. (One caution: you'll see much bigger "glass substrate market" numbers in the $7–17B range floating around — those lump in display and solar glass, which is a different animal entirely. The narrow packaging number is the one that matters here.)
The One Process That Decides Everything
To understand why a handful of equipment firms hold so much leverage, you need to understand the single step that keeps factory managers up at night: drilling the holes.
A glass substrate is useless unless you can pass electrical signals from the top of it to the bottom. That means punching tens of thousands of microscopic vertical channels — Through-Glass Vias (TGVs) — clean through the glass, each one maybe a tenth the width of a human hair. Then you line them with copper. Sounds straightforward. It is not.
Glass is brittle and chemically standoffish — wonderful for clean signals, terrible for being drilled. Traditional mechanical drilling or crude laser ablation leaves microscopic cracks around each hole. Then you fill it with copper, the two materials expand at different rates when heated, and the glass shatters. Industry reporting from Korea's THE ELEC has described early yields where roughly nine out of ten boards cracked and had to be scrapped. Nine out of ten. That's not a production line; that's a very expensive shredder.
So the company that figures out how to drill clean, crack-free vias at scale doesn't just have a nice product. It holds the chokepoint for the entire industry. Honestly, this is the whole ballgame.
The Arms Dealers — Meet the Machine Makers
Now for the fun part: the actual companies. The glass substrate process line is a relay race — glass core, TGV drilling, metallization, build-up layers, imaging, singulation, inspection — and different firms own different legs. But the leg everyone watches is TGV. Here are the players holding the picks.
And the supporting cast keeps growing. Taiwan's E&R Engineering is another TGV contender. Japan's DISCO brings decades of expertise in precision dicing and grinding — critical for cutting and finishing brittle glass without chipping it. Applied Materials has been exploring hybrid laser-mechanical drilling. And Korea's Hanwha Precision Machinery has adapted its packaging equipment for the delicate handling of glass, drawing on its display-industry roots. Notice the pattern: a lot of these firms come from displays. Glass substrate manufacturing is, in many ways, the semiconductor industry quietly poaching the flat-panel industry's hardest-won skills.
The Process Line — Who Owns Which Leg
Before we move on, a picture helps. Here's the relay, simplified, with the equipment vendors mapped to the steps they're known for.
One of These Names Doesn't Add Up
So that's the cast. A clear bottleneck, a chokepoint process, and a handful of firms — mostly German and Korean — selling the only tools that solve it. If the picks-and-shovels logic holds, these are the names that get paid no matter how the substrate war ends.
But here's where it gets strange, and where Part 2 picks up. Among these toolmakers, one small company has become the stock market's favorite proxy for the entire glass substrate thesis. Its revenue is actually shrinking. Its share price is doing the opposite — emphatically. Part 2 digs into that anomaly, names the gold miners whose orders are driving it, and lays out the risks nobody prints on the brochure.
Comments
Post a Comment