Build2 distinct publishers3 min readPublished
The Machine Age Fund treats power and heat as the binding constraint on AI scaling, which hands hardware founders a dedicated buyer and hands the rest of us a rack-power forecast that a16z produced itself.
The Engineer · Build desk
Compiled by The EngineerSomething wrong?How this is made
Take the rack numbers as a spec sheet rather than a slide. Moving from 5 to 10 kW per rack to 100 to 250 kW is a 10x to 50x increase in power delivered into the same floor position [7][1]. The 1 MW rack a16z expects within three years is another 4x on top of that range [7][2]. These are a16z's own projections, and the announcement carries no independent forecast [7], so the useful question is which of the two numbers you are buying against. At 250 kW you are placing purchase orders. At 1 MW you are filing for a utility interconnect. Those are different calendars, and only one of them speeds up when you add capital.
That distinction is why a fund like this behaves unlike a software fund. a16z's own framing is that computing infrastructure is being redesigned down to the electricity, with economics closer to industrial development than to the low-capital software businesses venture returns were recently built on [13]. Industrial development means money in front of revenue for years, on schedules set by permitting and construction. The fund says it wants to "open the throttle and accelerate the physical buildout of AI" [4]. Throttles assume the fuel line is already connected.
Now the number I would push on at review. a16z says hardware and infrastructure already accounts for over 20% of its recent startup deal flow, and discloses neither a time period nor a deal count [3]. Twenty percent of forty deals in one quarter and 20% of four hundred over three years describe different firms with different conviction. Without a denominator it is positioning.
The roster is better evidence than the percentage. Horowitz co-founded Loudcloud in 1999, turned it into Opsware, software for automating data-center operations, and sold it to Hewlett-Packard in 2007 [9]. The lesson a16z draws from that period is what happens when a technically sound product arrives before the hardware and connectivity around it are ready [c9b]. Casado's Nicira made network infrastructure programmable and went to VMware for $1.26 billion in 2012 [10]. Ulevitch's OpenDNS went to Cisco for $635 million in 2015 [12]. All three were software sold into infrastructure, not manufacturing. a16z seems to know it, which is why the announcement names Guido Appenzeller, Shangda Xu and Erin Price-Wright as hardware and infrastructure specialists alongside the five partners who presented the fund [6][5].
Fund arithmetic sharpens the problem. The $1.1 billion reserved here is 87% of what Nicira sold for [3]. For one Nicira-scale exit to return this fund, a16z would have to own nearly the entire company at exit, which venture ownership does not deliver. So the thesis requires outcomes larger than the ones on its partners' resumes, in a category where every company carries manufacturing and supply-chain risk stacked on top of engineering risk.
The 2009 franchise was organized around the argument that software would eat the world [8]. This one is capital aimed at the layer where the answer is amperage and heat rejection, and where a rewrite buys you nothing.
Ranked by verification strength, evidence, and original report placement.
a16z launched the Machine Age Fund on August 28, saying it raised $1.1 billion to back the chips, memory, networking, power systems, data centers and robots required to keep AI scaling.
a16z's fund announcement describes AI development as a "social and national imperative" and says it wants to "open the throttle and accelerate the physical buildout of AI."
a16z's announcement argues computing infrastructure is being redesigned "all the way down to the electricity," and the source notes the economics look closer to industrial development than the low-capital software businesses on which venture funds built much of their recent performance.
a16z says its hardware practice already accounts for over 20% of its recent startup deal flow; that percentage lacks a disclosed time period or deal count.
Ben Horowitz presented the fund with Martin Casado, Raghu Raghuram, David Ulevitch and David George.
The announcement separately identifies Guido Appenzeller, Shangda Xu and Erin Price-Wright among a16z's hardware and infrastructure specialists.
Distinct publishers with included, body-backed reporting in this cluster.
mezha.net
1 article · August 28, 2026
runtimewire.com
1 article · August 28, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
product
a16z's $1.1bn hardware fund prices AI roadmaps in kilowatts per rack1 distinct publisher
product
A DOJ probe of a16z board seats asks whether venture portfolios are interlocking directorates1 distinct publisher
invest
DOJ's a16z probe puts a venture standard on trial: partners on boards of rival portfolio companies2 distinct publishers
product
a16z puts $1.1bn into the hardware its software portfolio is waiting on2 distinct publishers
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One announcement, two retellings
Strip out the history and nothing here has a second witness. RuntimeWire says TechCrunch had it first, mezha.net says the same, and both are working from a16z's launch text. The old prices check out as public record — HP buying Opsware, VMware's $1.26 billion for Nicira, Cisco's $635 million for OpenDNS — but those are resumes, not evidence about this fund. Everything a reader would actually act on is unverified: no limited partners, no terms, a deal-flow share with no denominator, and a power curve the firm drew for itself.
A pool announced, not yet a pipeline
Reserved capital is a commitment, not a deployment. The companies a16z points to — Unconventional AI, Nexthop, Volta, Atoms, Mind Robotics, and further back Skydio, SpaceX, Anduril, Waymo — were already in the portfolio, and the firm conspicuously declines to say whether any were funded from this vehicle. The only demand signal is 'over 20% of recent deal flow.' What is observable is a large number, an operator bench and a hiring pattern; what is not observable is a single check written under the new mandate.
Throttle language, unpriced physics
'Open the throttle,' a 'social and national imperative' and a megawatt rack three years out carry more weight in this story than any verified figure. The overreach sits in the forecast rather than the fundraise: $1.1 billion is concrete, and RuntimeWire is careful to mark the power curve as a16z's own and the 20% share as undated. Mezha.net passes the imperative language through with no such qualification. Overstated, then — but by the firm, and partly flagged by one of the two publishers reporting it.
The fundraiser drew the curve
The only party supplying the power projections is the party raising money against them, and it is simultaneously advertising to the founders it wants to buy. A steeper rack-power line makes both the fund and the firm's hardware bench look more necessary, and the 'national imperative' register does work with limited partners and policymakers alike. Mezha.net's own sidebar mentions a nearly year-long Justice Department look at a16z over board conflicts at AI companies — unrelated to this fund, but a reminder that the firm's interests are already being examined elsewhere.
Confident on shape, cautious on substance
We can see clearly who is claiming what and who gains if it holds. What we cannot see is anything past the announcement: two publishers, one origin, and only RuntimeWire adding reporting of its own — the operator histories, the rival funds, the missing denominator. That is enough to characterise the bet and its incentives, not enough to judge whether the physics arrives on schedule or where the money actually lands.