Micron’s Rosy Memory Outlook Lifts AI Hopes

T-minus ten and holding on Blue Origin’s New Shepard, flight NS37, an eleven-minute, fully autonomous hop from West Texas with six on board and a milestone first: a spaceflight for a traveler who uses a wheelchair; Lauren, what’s behind the pause?

Holds like this are normal for New Shepard, and high winds pushed the schedule earlier; it looks like a weather-driven pause rather than a technical issue.

The FAA lists a backup window as soon as tomorrow; Chad, how important is it for Blue Origin to execute smoothly on this thirty-seventh mission?

They’ve built a strong safety record with New Shepard, which paved the way for New Glenn’s early successes this year; what really stands out is how space is opening to more people, including flyers with accessibility needs.

Lauren, zooming out, there’s a rush to avoid a single rocket monopoly and to capture new profit pools like space data centers—how are companies positioning?

Launch alone won’t sustain most business models, so players are stacking lines of work—satellites, space stations, crew flights—and now see orbital computing as a potential gold mine.

And for New Glenn’s next phase, what should we expect, Chad?

Blue Origin tests quietly and launches after heavy validation, which is why New Glenn’s first two outings and a landing went well; with Starship and New Glenn scaling, 2026 becomes about lifting far more mass to orbit and unlocking big infrastructure like stations and orbital data centers.

Markets check: cooler inflation has stocks in rally mode, and Micron’s blowout quarter eased AI jitters; what’s the read-through, Kamran?

Micron’s strength reignited confidence after Oracle and Broadcom spooked investors, lifting memory peers and bringing AI heavyweights back into the green, though this remains a stock-picker’s tape.

On the other side, higher memory prices pinch buyers like Dell and HPE.

Exactly—leaders delivering real growth get rewarded, while companies with margin pressure or missteps lag.

Janet, is the market now demanding hard AI earnings proof?

Yes—investors want visible commercial impact and are wary of leverage as we head into 2026.

Does that mean more pain for big spenders like Oracle?

Long-dated commitments and debt draw scrutiny, while cash-rich hyperscalers look sturdier; some worry is already priced in.

Is betting on OpenAI exposure out and backing Alphabet the safer trade, or can multiple AI winners coexist?

The market is broad enough for several winners, and scale plus strong foundations should prevail despite near-term shifts in sentiment.

Can breadth improve without the leaders derating?

Leaders can stay expensive if they keep delivering, and breadth can build on economic resilience, rate cuts, and clear AI productivity gains.

There’s also rising interest in private assets; Patrick, how strong is demand for late-stage private tech?

Companies are staying private longer and scaling bigger, and investor appetite is deep because innovation in the private markets is robust.

Does Databricks’ lofty valuation rest on fundamentals the public market can support?

It’s growing far faster than comparable public software names with exceptional customer expansion and savvy M&A, which underpins the valuation and sets it up as a top-tier eventual listing.

And OpenAI—are they raising now?

No comment on the round, but we’re proud backers—product velocity and partnerships keep them at the front of large language models.

How should investors stay grounded amid AI concentration risk?

Step back: adoption is unprecedented, private companies are hitting revenue milestones faster with leaner teams, and while any cycle has excess, it also mints enduring leaders.

Beyond gen AI, is 2026 about space and defense too?

Yes—defense tech and space are surging, while AI and cybersecurity remain resilient because the ROI from deployment exceeds the cost.

Back to West Texas—the countdown is back; Lauren, did winds clear?

Yes, a new T-minus is set, though another hold wouldn’t surprise me; for now, it’s green.

Chad, remind us how these suborbital rides fit Blue Origin’s roadmap.

New Shepard is a short, suborbital hop with an excellent safety record that seeds tech for bigger ambitions like New Glenn, stations, and a NASA-backed lunar lander.

Lauren, how tight is the government partnership?

Blue Origin is a key NASA partner on the Artemis lunar lander, and we’re watching a fresh space race with SpaceX to return people to the Moon.

Chad, does bringing in Dave Limp from Amazon matter?

He’s been pivotal in shifting Blue Origin from R&D to execution across launches, New Glenn, and lunar work.

We’re seconds from liftoff—then another pause; Lauren, what are the crew doing in these final minutes?

It’s an automated ride, so they’re strapped in and waiting as the system runs through its checks.

Today’s group includes a physicist-investor and the first wheelchair user headed to space; how are flyers chosen, Chad?

It’s a mix of paid seats and curated choices to tell a broader story, and seeing a SpaceX veteran like Hans take a Blue Origin seat is a notable crossover for the industry.

Back to Micron’s surge—Vasily, why the enthusiasm?

Results topped on every key metric, demand for high-bandwidth memory is intense, and they’re supply constrained for major customers.

But rising memory prices weigh on hardware makers.

Right—megacaps can absorb it, but others like PC and server vendors feel the squeeze.

Joanne, how central is HBM to Micron now?

It’s critical—AI data centers need it, capacity is tight, and that supports pricing power.

Does capex risk the classic memory bust?

Memory is cyclical, but AI buildouts could extend this upcycle even as capacity slowly catches up due to tooling bottlenecks.

Will competition intensify into 2026?

SK Hynix and Samsung are active, yet expansions take time because key equipment is scarce, giving some visibility to the cycle’s length.

And those financing worries after Oracle?

Anxiety will linger, but hyperscalers’ cash flows can fund the build; the real unknown is how quickly AI apps drive profits for users.

James, China’s racing to self-sufficiency in chips—how far along are they on advanced lithography?

The question has always been EUV access, and reports now point to a domestic prototype; the U.S. needs a leap-ahead plan, not just defensive moves.

And the talent battle?

China’s engineering talent is world-class and mobile, so assume they can catch up and plan accordingly.

What’s Substrate building, and how close are you with government?

We’re pursuing x-ray lithography via particle accelerators, have a working prototype in three years, and collaborate closely with U.S. agencies; progress on optics and tools is further along than many assume.

Are export controls too late?

Not too late, but the margin is thin; the U.S. should focus on speed and new approaches—that’s our play.

Trump Media is merging with TAE Technologies to create a public fusion company; Will, why this move now?

TAE has made steady progress but needs capital, and in fusion, money is becoming as critical as the physics.

What sets TAE’s approach apart?

They favor a different fuel mix that’s easier to source, though the real unlock is maintaining a superhot plasma with powerful magnets.

Realistically, when could we see power?

Fusion moves slowly, but timelines are tightening—TAE aims to start a first commercial plant build soon and talks about early power in the next decade, though definitions matter.

That’s it for Bloomberg Tech; the New Shepard attempt is scrubbed for today, and we’ll see if they try again tomorrow—find this show on the Terminal, online, or wherever you get your podcasts.