NewMoneyMoves
PerspectivesResearchTerminalsWeekend RWABoard
VIEW
InsightEditorialAdd a viewContextThesisReadingOutlook
››Sector Deep Dive
EDITORIAL VIEW

AI Infrastructure 4 Bottlenecks — Big Tech Q1 2026 Earnings Summary

Memory · Power · Custom Chips · Optics — Four Bottlenecks Converging Simultaneously

HHaelangdal·Founder AnalystApril 30, 202627 min readSector Deep Dive
NewMoneyMoves

Global Investment through Themes

Content

ThemesIdeasReportsEarnings

Explore

GuideETFsPerspectivesTelegramNVIDIA PortfolioWeekly ReviewCredit & Leverage

Community

Daily News

Legal

Privacy PolicyTerms of ServiceDisclaimer

© 2026 NewMoneyMoves. All rights reserved.

Bottom Line

Four bottlenecks converging simultaneously — prioritize true global #1 by sector; recommend Korean names only where Korea genuinely leads.

Reader's Brief — 30-second TL;DR

Advanced
Why Now

Big Tech 4 2026 CapEx raised to $700B+; Microsoft CFO stated '83% of increase is component costs,' confirming memory price pass-through structure.

Winners ?? Losers

Winners — SK hynix, Doosan Enerbility, TSMC, Marvell, Lumentum. Caution — Hanmi Semiconductor (share loss), OE Solutions (awaiting profitability).

Watch For

SK hynix quarterly OPM trend, Doosan new order disclosures, TSMC CoWoS expansion schedule, Lumentum InP capacity expansion, Apple quarterly memory-cost guide commentary.

Reading depth
  1. 01What Big Tech Said in Q1All four Big Techs said they would spend more, for different reasons. MSFT pinned $25B of the rise on component inflation, and all four cited compute constraints.Jump to section
  2. 02Bottleneck 1: Memory — Why Prices Are SpikingMemory prices spike as the big three shift general-memory capacity to HBM. SK Hynix's 71.8% operating margin is the first quarter HBM hikes fully hit the P&L.Jump to section
  3. 03Bottleneck 2: Power — How to Feed DatacentersOn the power bottleneck, Big Tech builds its own plants (BYOP). The three turbine makers fill slots to 2030, and Doosan fills the trio's 5-7 year delivery gap.Jump to section
  4. 04Bottleneck 3: Custom Chips — Why Big Tech Builds In-HouseBig Tech builds its own chips because NVIDIA GPUs are expensive, not optimal, and hard to get. The ASIC has moved from a cost tool to a revenue source.Jump to section
  5. 05Bottleneck 4: Optics — How Chips CommunicateThe 1.6T transceiver market was revised 3.5x in 12 months, and Lumentum effectively monopolizes the key 200G EML laser. It stays short no matter how much is made.Jump to section
  6. 06Conclusion — Top Picks by BottleneckLaid side by side, the four bottlenecks sort the true No. 1 by sector. Korea is valid in memory and power; global names come first in custom chips and optics.Jump to section

What Big Tech Said in Q1

More Spending, Different Reasons

Combined 2026 investment plans increased from $645B to ~$700B, up approximately $60B. The headline looks like "AI investment grew again," but the actual meaning is different.

CompanyCapEx ChangeKey Comment
Google+$50BSundar Pichai: "Cloud revenue would have been higher if we could meet all demand"
Microsoft+$300BAmy Hood: "$250B of the $300B increase is component price inflation"
Meta+$100-150BZuckerberg: "Most of the increase is component costs, especially memory"
AmazonNo changeAndy Jassy: "We have $225B in Trainium revenue commitments"
Company
Google
CapEx Change
+$50B
Key Comment
Sundar Pichai: "Cloud revenue would have been higher if we could meet all demand"
Company
Microsoft
CapEx Change
+$300B
Key Comment
Amy Hood: "$250B of the $300B increase is component price inflation"
Company
Meta
CapEx Change
+$100-150B
Key Comment
Zuckerberg: "Most of the increase is component costs, especially memory"
Company
Amazon
CapEx Change
No change
Key Comment
Andy Jassy: "We have $225B in Trainium revenue commitments"

The most important statement came from Microsoft CFO: "$250B of the $300B increase is component price inflation." They're spending more for the same quantity of servers because memory prices have risen.

> Amy Hood (Microsoft CFO): "Our 2026 capital expenditure of $190B includes approximately $25B of component price increases. Despite this additional investment, we will remain capacity constrained throughout at least 2026."

All four companies used the same phrase in their earnings calls: "compute constrained" — not enough computing capacity.

Custom Chips Become Revenue Streams

The most significant change this quarter is that custom chips (ASICs) are no longer just "internal cost-saving tools." They're now generating external revenue.

Amazon is the clearest example. CEO Andy Jassy disclosed "$225B in Trainium revenue commitments" — Anthropic at 5GW, OpenAI at 2GW, plus Uber and Meta.

> Andy Jassy (Amazon CEO): "Our custom AI chip business has entered the top 3 datacenter chip businesses globally. We now have over $225B in Trainium revenue commitments."

Google also announced that they will "deliver TPU hardware directly to certain customers' datacenters" — effectively turning TPU into an external revenue stream.

Takeaway

All four Big Techs said they would spend more, for different reasons. MSFT pinned $25B of the rise on component inflation, and all four cited compute constraints.

Full access requires 🥉 Bronze tier

Sign in with Google — your tier will be checked automatically and access granted if eligible.

Sign in with Google

This site runs on ads — the tier system rewards community contributions.

Comments

This report is provided for informational purposes only and does not constitute a recommendation to buy or sell any financial instrument. Investment decisions should be made based on your own judgment and responsibility. The analysis and opinions contained herein are based on information available at the time of writing and are subject to change.

All Reports
Home
Perspectives
Company
Macro
Theme