Every week I want to pull together a few stories from the markets and tech world that actually matter — not everything that happened, just the stuff worth knowing. This is a new format for me, so bear with me while I find my footing. This week: AI chip stocks had a rough ride, America’s biggest power grid is straining under AI’s appetite for electricity, and a quantum computing project just got a very real industrial test case.
1. AI Chip Demand Is Real — But So Is the Volatility
On Wednesday, Dutch chipmaking-equipment giant ASML posted second-quarter net income of €2.92 billion, beating analyst expectations of €2.62 billion. More striking: it raised its full-year 2026 revenue guidance for the second time this year, from a range of €36–40 billion to €43–45 billion — a 16% increase at the midpoint — citing surging AI-driven demand for advanced logic and memory chips. Shares on the Amsterdam exchange closed up 3.7% at €1,613.
Then came Thursday. The Nasdaq fell 0.82% and the S&P 500 slipped 0.16% as chip stocks pulled back for a second straight session. TSMC dropped nearly 2% even after reporting record quarterly revenue and raising its own capital spending outlook — investors got nervous about high valuations and warnings of further price increases. Micron fell more than 5%.
The takeaway: demand for AI chips looks genuinely strong across the industry, but after a long rally, the stocks themselves are getting jumpy about how much of that good news is already priced in.
2. The Grid Is Starting to Show the Strain of AI
PJM Interconnection, the largest power grid in the United States (it covers 13 states plus Washington, D.C.), announced the results of its latest capacity auction on July 14 — for electricity to be delivered starting mid-2028 — and by PJM’s own account, it was a first: this is the first time in PJM’s history that the entire grid fell short of its own reliability requirement, missing the target by 6,831 megawatts, or roughly the output of seven nuclear reactors. The auction procured 138,318 MW, plus another 10,864 MW through PJM’s Fixed Resource Requirement areas.
Prices cleared at PJM’s regulatory price cap of $325 per megawatt-day. Without that cap, prices would have landed around $554.72/MW-day in most of the footprint — and $776.69/MW-day in the Chicago-area (ComEd) zone, according to PJM’s own auction data. FERC Chair Laura Swett pointed to these results as new evidence that PJM’s governance needs an overhaul, and FERC has scheduled a technical conference for July 23 focused specifically on reforming PJM’s stakeholder process. Separately, PJM already updated its emergency playbook back in June, adding steps like advance “capacity advisories” and, as a last resort, calling on large users such as data centers to run their own backup generators.
This is the part of the “AI boom” story that doesn’t get as much attention as chip earnings, but it’s arguably just as important — the physical grid has to keep up with the digital demand, and right now it’s falling behind.
3. Quantum Computing Gets a Real-World Industrial Partner
On a lighter note: quantum computing company Quantinuum announced a multi-year partnership with Rolls-Royce, quantum error-correction firm Riverlane, and the University of Edinburgh’s supercomputing centre (EPCC) to explore how quantum computers could help with industrial design and simulation — starting with fluid dynamics for gas turbines.
The team plans to test core computational building blocks on Quantinuum’s Helios quantum computer, with an eye toward scaling up to its future Sol and Apollo systems. It’s part of a broader UK push toward what researchers call “TeraQuOp” computing — quantum machines capable of a trillion error-free operations, a benchmark many in the field see as the point where quantum computing starts solving problems classical computers genuinely can’t.
It’s early-stage research, not a product launch, but pairing a quantum computing company with an actual jet-engine manufacturer is a good sign that the field is moving from theory toward real engineering problems.
Why These Three Together
Different headlines, one common thread: AI is reshaping more than just software. It’s moving chip company earnings, straining the physical power grid, and now pulling next-generation computing research into old-economy industries like turbine design. I’ll be back next week with three more.
This post is for informational purposes only and isn’t investment advice. I’m not a financial advisor — please do your own research (or talk to a licensed professional) before making any investment decisions.