a16z Machine Age Fund Bets $1.1B on AI Hardware

Andreessen Horowitz raised $1.1B for the Machine Age Fund to rebuild AI's physical stack as rack power heads toward 1MW.

7 min read
Data center racks with high-density AI servers and power infrastructure
a16z says rack power will hit 1MW within three years, driving its $1.1B hardware push.· a16z Blog
Visual TL;DR
Rack power surgeDriver
data center rack power approaching 1 megawatt strains current infrastructure
From the article 3 mentionsRack power moved from roughly 5-10 kW to 100-250 kW to support today's systems and will increase to 1MW over the next 3 years.
Machine Age FundCore
andreessen horowitz raised 1.1 billion to rebuild ai's physical stack
From the articleThe a16z Machine Age Fund is here with $1.1 billion to build the physical AI stack.
Physical stackContext
investment spans chips memory networking storage and full systems from data centers to robotics
From the article 3 mentionsThe a16z Machine Age Fund is here with $1.1 billion to build the physical AI stack.
AI buildoutEffect
fund aims to accelerate ai physical buildout as a national imperative
From the article 2 mentionsAccording to a16z Blog, Andreessen Horowitz has raised the new vehicle to accelerate the physical buildout of AI.
Problem solvingOutcome
ai positioned as strongest tool for solving problems and bestowing abundance across sectors
From the articleThe firm calls AI the strongest tool ever developed for solving problems and bestowing abundance.
Rack power surgeDriver
data center rack power approaching 1 megawatt strains current infrastructure
From the article 3 mentionsRack power moved from roughly 5-10 kW to 100-250 kW to support today's systems and will increase to 1MW over the next 3 years.
Machine Age FundCore
andreessen horowitz raised 1.1 billion to rebuild ai's physical stack
From the articleThe a16z Machine Age Fund is here with $1.1 billion to build the physical AI stack.
Physical stackContext
investment spans chips memory networking storage and full systems from data centers to robotics
From the article 3 mentionsThe a16z Machine Age Fund is here with $1.1 billion to build the physical AI stack.
Hardware betsCore
From the article 8 mentionsThe firm pointed to Unconventional AI, Nexthop, Volta, Atoms, Heron Power and Mind Robotics as recent hardware investments.
AI buildoutEffect
fund aims to accelerate ai physical buildout as a national imperative
From the article 2 mentionsAccording to a16z Blog, Andreessen Horowitz has raised the new vehicle to accelerate the physical buildout of AI.
Problem solvingOutcome
ai positioned as strongest tool for solving problems and bestowing abundance across sectors
From the articleThe firm calls AI the strongest tool ever developed for solving problems and bestowing abundance.
Contents(4)

The a16z Machine Age Fund is here with $1.1 billion to build the physical AI stack. According to a16z Blog, Andreessen Horowitz has raised the new vehicle to accelerate the physical buildout of AI.

It is pitched as a social and national imperative. The firm calls AI the strongest tool ever developed for solving problems and bestowing abundance.

What the Machine Age Fund will actually buy

The fund will invest across all computer infrastructure on which AI runs, including chips, memory, networking and storage.

It also includes full systems, from data centers to robotics to home AI appliances.

The common thread is constraint. Every layer is hitting the wall of today's supply chain capability and the limits of physics and computer science.

Recent bets show the direction. The firm pointed to Unconventional AI, Nexthop, Volta, Atoms, Heron Power and Mind Robotics as recent hardware investments.

It is not a first hardware experiment. A16z led the Series A in Skydio in 2016, invested in SpaceX, wrote its first check into Anduril in 2019, and was among the first venture investors in Waymo's 2020 raise.

Why rack power and density force a rebuild

A16z frames this as a rearchitecture, all the way down to electricity. The numbers explain why.

Compute density per rack increased by 28X from an H100 rack to a Rubin rack.

Networking within a rack has grown similarly, hitting the limits of copper cabling.

Rack power moved from roughly 5-10 kW to 100-250 kW to support today's systems and will increase to 1MW over the next 3 years.

That last figure is not hypothetical. Nvidia (NASDAQ:NVDA) has told partners that Rubin Ultra on the Kyber rack pushes to 600kW plus in 2027 and Feynman aims for 1MW. By 2027, the firm expects to reach one megawatt per rack.

Data center scale is moving from tens to hundreds of MW and, in some cases, GW-scale campuses. Power is increasingly supplied not only from grid-only to grid plus behind-the-meter or captive sources.

The hardware industry supply side is used to growing 20% to 30% per year at most, not the triple-digit growth that is needed to catch up with demand.

A16z says it needs faster systems, cheaper and higher-bandwidth memory across the hierarchy, faster interconnects, and power-efficient edge devices. Plus all the cooling, materials, electrical and real estate buildout to support them.

Who is running the fund

Hardware is in the team DNA, the firm says. Guido Appenzeller was the CTO for Intel (NASDAQ:INTC) Data Center Group.

Raghu Raghuram and Martin Casado spent multiple decades in the data center space with system software that required deep hardware partnership.

Shangda Xu and David George have led investments across the AI infrastructure stack, from silicon and networking to large-scale systems and compute platforms. David Ulevitch and Erin Price-Wright lead many hardware and US manufacturing investments through the American Dynamism practice.

That mix matters. The bottleneck has shifted from model weights to watts, copper and concrete.

Why this matters now

Founders understand the opportunity. Over the last couple of years hardware startups grew from a small amount of deal flow to now over 20%.

A16z is making hardware an official motion. It says its GTM, talent and marketing machine is now ready to serve hardware founders with access to customers and suppliers.

For context, the timing follows a year where AI moved from chat to reasoning to coding, and token intensity per task rose by orders of magnitude. When each task burns more tokens, you need more racks and more power.

Competitors are not idle. Founders Fund, Lux Capital and DCVC have all pushed larger hardware and defense tech vehicles, and hyperscalers are co-designing power with chip vendors. A16z is betting a generalist platform can win hardware deals that once went to specialist firms.

For founders, the signal is capital for long cycles. For incumbents, the pressure is supply chain velocity. If 28X density and 1MW racks arrive in 36 months, the winners will not be those with the best model alone.

They will be those who can ship the rack that powers it.

© 2026 StartupHub.ai. All rights reserved. Do not enter, scrape, copy, reproduce, or republish this article in whole or in part. Use as input to AI training, fine-tuning, retrieval-augmented generation, or any machine-learning system is prohibited without written license. Substantially-similar derivative works will be pursued to the fullest extent of applicable copyright, database, and computer-misuse laws. See our terms.