a16z Raises $1.1B Machine Age Fund for AI Hardware, Data Centers, and Power
TL;DR
Andreessen Horowitz has raised the $1.1B Machine Age Fund for chips, memory, networking, storage, data centers, robotics, and home AI devices, while its deployment pace and portfolio remain undisclosed.
The hardware thesis behind this Andreessen Horowitz fund has a measurable test over the next three to six months. If the first disclosed Machine Age Fund investments remain concentrated in software services, or if they do not include capital-intensive businesses in chips, power equipment, and robotics, the $1.1B fundraising figure will not demonstrate that a16z has built a hardware investing capability. The firm announced the fund on August 28, 2026, but did not disclose a deployment period, typical check size, or any already signed transactions.
The mandate reaches across the physical supply chain for AI computing. At the component layer, a16z names chips, memory, networking, and storage. Its systems category includes data centers, robotics, and home AI appliances. The firm also extends the rebuilding argument down to electricity, saying that current supply-chain capacity and limits in physics and computer science constrain expansion. That argument translates model demand into specific equipment that companies must deliver. The announcement, however, does not allocate the fund among those categories.
The Machine Age Fund also makes hardware an official investment motion at a16z for the first time. Ben Horowitz, Martin Casado, Raghu Raghuram, David Ulevitch, and David George signed the announcement. The firm points to their experience in data-center software, chips, networking, US manufacturing, and growth investing. It says its go-to-market, recruiting, and marketing teams will now support hardware founders. Those services may help startups reach customers and suppliers, but they do not remove the capital risks created by wafer capacity, grid interconnections, factory schedules, and inventory cycles.
SiliconANGLE independently confirmed the amount and the stated investment scope. It also reported that Andreessen Horowitz has backed several AI infrastructure companies during the past two years. The examples include solid-state data-center transformer developer Heron Power, data-center operator Volta Infrastructure, chip company Unconventional, and several robotics startups. These positions show that the strategy is not starting from zero. Neither source states how much of the new $1.1B will support existing portfolio companies, how much will go to new deals, or how the expected returns compare with software funds.
The independent report adds a useful competitive reference. Kleiner Perkins raised $3.5 billion across two new funds in March 2026, with one vehicle focused on early-stage AI companies. The amounts are not directly comparable because investment stages and sector mandates differ. The a16z strategy can instead be tracked through physical outcomes: whether portfolio companies win production orders, whether data-center power equipment shortens delivery times, and whether robotics companies move from pilots to repeat purchases.
Over the next three to six months, the useful disclosures will be the number of investments, their hardware categories, and the stage of each company. Investors should not treat the entire $1.1B as money already deployed into the AI supply chain. When a16z reveals its first transactions, the use of proceeds will matter as well: research, factory construction, and capacity expansion carry different schedules and technical risks. For now, the August 28, 2026 announcement proves that Andreessen Horowitz raised a dedicated fund. It does not yet prove that the capital has produced additional chip, power, or robotics supply.
Sources:
Related Articles
Huawei Unveils Ascend 960 SuperPoD: Optical Efficiency Still Needs Chips and Supply
Huawei details optical interconnects and an accelerated 2027 chip schedule. Claimed power savings exceed 550 kW, but testing and supply constraints still separate the architecture from delivered capacity.
NVIDIA Posts $96.2 Billion Quarter and Still Guides Higher With Zero China Data Center Compute Revenue
NVIDIA's fiscal 2027 second-quarter revenue rose 106% to $96.2 billion, including $89.0 billion from Data Center; its $108.0 billion outlook assumes no China Data Center compute revenue.