VC
Value Add VC
⚡HomePulse⚡Helpful Apps📝Blog
← Value Add PulseBIG TECH$725B combined

Big Tech's 2026 AI Capex Will Hit $725B, up 77% Year-Over-Year

Alphabet, Amazon, Meta and Microsoft collectively guided to roughly $725 billion in 2026 capital expenditure — up 77% from $410B last year — as the AI infrastructure buildout enters its third and most expensive year. Amazon leads at ~$200B, Microsoft ~$190B, Alphabet $175-185B and Meta $115-135B, according to freshly restated guidance from the four companies' most recent updates.

~$725B
Total 2026 Capex
+77% ($410B → $725B)
YoY Increase
~$200B
Amazon
~$190B
Microsoft
$175-185B
Alphabet
TC
Trace Cohen
Early-stage VC & angel · Founder, New York Venture Partners
June 30, 2026
1 min read
KEY TAKEAWAYS FOR VCs & FOUNDERS
1

$725B is more than the annual GDP of Poland, funneled almost entirely into GPUs, data centers, and grid infrastructure

2

Meta and Microsoft trading like 'bear market names' shows public markets are finally questioning the payback math

3

AI stock 'air pocket' at June end suggests the sell-side is testing whether spending has front-run realized revenue

4

Every dollar of this capex is a downstream tailwind for Nvidia, TSMC, and hyperscaler-adjacent private startups

TC
The VC Read · Trace's TakeTrace Cohen

The $725B number is either the biggest infrastructure moat any four companies have ever built or the biggest coordinated capital misallocation since the fiber overbuild of 1999 — and honestly nobody knows yet because monetization is running 12-18 months behind the spend. What I'm watching is the shift in how founders are pitching: 'we save hyperscalers 30% on training' now beats 'we build a new frontier model,' because every startup near the buildout has priced in that the Big Four will keep spending regardless. For LPs, the second-order private opportunity is grid infra, cooling, and power PPA aggregators — those are the arms dealers to the AI arms race, and they're still under-covered. Watch which hyperscaler blinks first on 2027 guidance; that's the top-tick signal.

💸 AI Spending →🏗️ AI Buildout Tracker →

The Big Four hyperscalers just told markets they will spend roughly $725 billion on capex in 2026 — a 77% jump from 2025's $410B and the single largest one-year corporate infrastructure buildout in modern history. Amazon is leading at approximately $200B (a $60B jump from 2025), Microsoft ~$190B, Alphabet $175-185B, and Meta $115-135B based on the latest restated guidance.

The context: as of June 30, 2026, AI stocks entered what analyst Dan Ives called an 'air pocket,' with Meta and Microsoft trading like 'bear market names.' Public markets are finally asking whether the spending will produce enough monetizable AI revenue in 2026 and 2027 to justify the depreciation drag. Microsoft's most recent quarter posted 34% cloud growth — strong, but not the acceleration bulls needed to underwrite $190B in spend.

“Microsoft's most recent quarter posted 34% cloud growth — strong, but not the acceleration bulls needed to underwrite $190B in spend.”

The money flow is almost entirely into three buckets: Nvidia GPUs (Blackwell and Rubin generations), custom silicon (Meta's MTIA, Google TPU v7, Microsoft's Maia, Amazon's Trainium 3), and physical data-center buildout — power, cooling, land. Grid interconnect delays are pushing hyperscalers into nuclear-power PPAs (Amazon-Talen, Microsoft-Constellation, Meta-Entergy) and even geothermal (Google-Fervo).

Second-order winners are massive: TSMC's 2026 revenue is expected to top $150B on this spend, ASML's EUV order book is booked out through 2027, and private AI-infra startups like Baseten, Groq, Modal and Together AI just posted their strongest bookings quarters ever. Contrarian read: the capex intensity looks a lot like the 1999 telecom buildout — durable long-term infrastructure but potentially followed by a brutal digestion period as depreciation catches up with monetization.

What to watch: Q3 earnings will be the first test of whether AI revenue growth is accelerating in line with capex growth. If not, expect one of the four to blink and cut 2027 guidance — the moment that happens, the entire AI-infra multiple compresses.

ShareXLinkedInEmail
More onMeta →Microsoft →

Originally reported by Fortune. Analysis and editorial commentary by Value Add Pulse.

← Back to Pulse

Read Next

BIG TECH$3.92B

Qualcomm Agrees to Buy Modular for ~$4B in All-Stock Deal, Taking Aim at Nvidia's CUDA Moat

Qualcomm agreed on June 24 to acquire Modular, the AI-infrastructure company founded by Chris Lattner (LLVM, Swift), in an all-stock deal valued near $4 billion — its largest ever. The play: use Modular's MAX platform to run AI models across Qualcomm silicon without hand-tuning code for each chip, directly attacking the software stack that has cemented Nvidia's data-center dominance.

BIG TECHDiscs end 2028

Sony to End Physical PlayStation Game Disc Production Starting 2028

Sony announced on July 1 that it will stop producing physical discs for new PlayStation games starting January 2028, citing that digital purchases already account for 85% of PlayStation game sales. The move doesn't affect games already released or shipping before 2028, but it marks the beginning of the end for physical media on the platform that helped define disc-based console gaming.

BIG TECH$5B valuation

Nvidia Challenger Etched Hits $5B Valuation on $1B in AI Chip Orders

Etched, the Harvard-dropout-founded AI chip startup building the transformer-specific Sohu chip, disclosed on June 30 that it has booked more than $1 billion in customer contracts for its 'frontier inference clusters' and sits at a $5 billion valuation following a $500 million round in December. The company says an 8-chip Sohu server can outperform 160 Nvidia H100 GPUs on Llama 70B inference while using less power and space -- a direct challenge to Nvidia's inference dominance from a company that nearly ran out of cash three years ago.

@Trace_Cohen·t@nyvp.com