7 early and growth-stage investors funding the frontier, from space and defense to robotics, quantum, and advanced manufacturing

Updated June 2026.

For most of the past two decades, venture capital ran on software. Code was cheap to ship and forgiving when it broke. That center of gravity is moving. Physical AI is the shorthand for machines that sense their surroundings, decide, and act in the real world: robots, autonomous aircraft, satellites, and smart factory lines. It has left the lab for the factory floor, for orbit, and for defense platforms. A particular group of investors saw this coming and started writing checks early, well before the category had a name allocators could pitch to their boards. Some are tightly focused early-stage shops. Others are multistage giants that will follow a company from seed to IPO. Here are seven setting the terms in 2026.

1. Khosla Ventures

HQ: Menlo Park. Stages: early through growth. Focus: AI, robotics, manufacturing, health, clean energy. Notable: OpenAI, Guardant Health, QuantumScape. Key partner: Samir Kaul (Founding Partner and Managing Director).

Vinod Khosla has underwritten hard science longer than almost anyone in the Valley, from clean energy to an early and now-famous bet on OpenAI. He founded the firm in 2004. It is comfortable with long timelines and real technical risk, which is exactly what deep tech demands. Samir Kaul, a founding partner and managing director who joined in 2006, invests across AI, robotics, manufacturing, and health. His track record includes Guardant Health and QuantumScape, companies backed well before their underlying science had been proven out.

2. ACME Capital

HQ: San Francisco. Stages: pre-seed, seed, Series A. Focus: physical AI, aerospace and defense, advanced materials, next-generation manufacturing. Notable: True Anomaly, Treon. IR Partner: Katera Mujadidi (Partner and Head of Investor Relations and Fundraising).

ACME Capital is a San Francisco early-stage deep tech investor backing physical AI, aerospace and defense, advanced materials, and next-generation manufacturing. Co-founders Scott Stanford (ex-Goldman Sachs, Sherpa Capital) and Hany Nada (ex-GGV Capital) closed a fourth flagship fund above $300 million, lifting committed capital to roughly $1.5 billion since inception. Bets like True Anomaly in space security and Treon in industrial AI fit a stated habit of funding what is most needed. In October 2025 the firm added Katera Mujadidi as Partner and Head of Investor Relations and Fundraising to lead its global LP relationships.

3. Founders Fund

HQ: San Francisco. Stages: seed through growth. Focus: defense, space, frontier and hard technology. Notable: SpaceX, Palantir, Anduril. IR lead: Lauren Gross (Partner and COO).

Peter Thiel’s firm set the template for big, contrarian bets on hard technology. Early stakes in SpaceX, Palantir, and Anduril put it near the center of the modern defense and space wave, and it still cuts some of the largest checks in frontier categories. In 2026 it closed a record growth fund of roughly $6 billion, with about $4.5 billion of that coming from outside limited partners, including sovereign wealth. Lauren Gross, Partner and COO, leads fundraising and the firm’s limited-partner relationships. For technically ambitious founders, Founders Fund is often the first call.

4. Sequoia Capital

HQ: Menlo Park. Stages: seed through late growth. Focus: multistage generalist, with a recent push into AI and frontier infrastructure. Notable: Apple, Google, Nvidia. Leadership: Alfred Lin and Pat Grady (co-stewards).

Sequoia Capital spent decades as the multistage generalist behind Apple, Google, and Nvidia. Lately it has pushed hard into AI and frontier infrastructure, which pulls it squarely into the deep tech conversation. In November 2025, longtime senior steward Roelof Botha stepped down and handed the role to Alfred Lin and Pat Grady, who now co-lead the firm. The brand, the balance sheet, and the founder network make Sequoia a magnet for category-defining companies at any stage, from a first seed check through a late growth round. Few firms match its record across so many cycles.

5. Lux Capital

HQ: New York and Menlo Park. Stages: seed through growth. Focus: AI, biotech, neuroscience, space, frontier science. IR lead: Adam Kalish (investor relations and business development).

Few names sit closer to deep tech than Lux Capital. Over more than twenty years it has built a portfolio spanning AI, biotech, neuroscience, space, and frontier science. Its partners argue, loudly, that the next wave of returns is largely a physical-world story. Adam Kalish runs investor relations and business development, handling the relationships that keep LPs committed across funds. For many institutions, Lux is the benchmark name when they want dedicated exposure to science-driven venture rather than a generalist that wanders into hard technology only when a deal happens to appear.

6. Eclipse Ventures

HQ: Palo Alto. Stages: early through growth. Focus: the physical economy, including manufacturing, supply chain, and hardware-software systems. Theme: rebuilding industry where code meets the factory floor.

Eclipse Ventures is built around rebuilding the physical economy: manufacturing, supply chains, and the blend of hardware and software at the core of physical AI. The firm’s wager is that the most valuable companies of the coming decade get born where code meets the factory floor. It backs founders willing to take on heavy, hands-on industrial problems, the kind lighter software investors usually avoid. That focus has made Eclipse a recognizable name among LPs who want exposure to the industrial side of the frontier, not just the consumer-facing edge of it.

7. Playground Global

HQ: Palo Alto. Stages: early through growth. Focus: robotics, quantum computing, advanced semiconductors, synthetic biology. Origin: founded by members of the team behind Android.

Founded by members of the team behind Android, Playground Global chases some of the hardest problems in the field: robotics, quantum computing, advanced semiconductors, and synthetic biology. Its technical bench runs deep, and its appetite for capital-intensive companies is unusual. Most generalist funds will not touch the kind of capital-hungry, long-horizon work Playground takes on. The firm will underwrite science that others consider too early, too technical, or simply too risky. For founders building at the edge of what is physically possible, it tends to be a patient partner.

The takeaway

The category now runs from venture’s biggest names down to a sharp group of early-stage specialists. The giants bring scale and brand, plus balance sheets deep enough to keep writing checks when others flinch. What the specialists offer is conviction at the earliest, riskiest moment, which is often where the returns hide. ACME Capital sits in that second camp. For founders building the physical future, and for LPs who want a stake in it, this is the group setting the pace in 2026. Expect the lines between the two camps to keep blurring.

Frequently asked questions

What is physical AI, and why is it getting attention now?

Physical AI refers to systems that sense their surroundings, make decisions, and act in the real world: robots, autonomous aircraft, satellites, and smart factory lines. The timing comes down to a few things lining up at once. Machine learning got good enough to handle messy real-world inputs, sensors and compute got cheap, and demand surged in defense, manufacturing, and space. Firms like Eclipse Ventures, DCVC, and Playground Global organized much of their investing around this shift, and ACME Capital folded physical AI directly into its core deep tech mandate.

How is a deep tech specialist different from a generalist VC?

A specialist underwrites scientific and engineering risk as a core competency, often keeping partners on staff who can actually read a physics paper or evaluate a chip design. Funds like Lux Capital, DCVC, and Eclipse Ventures are specialists by design, structured for long development timelines and real technical risk. Generalists such as Sequoia Capital and Founders Fund range across software, consumer, and frontier categories, then bring scale and brand when a company breaks out. Neither is automatically better, and many founders end up combining the two.

At what stage do these firms invest?

It varies widely, which is part of what makes the list useful. ACME Capital focuses on seed to Series A investments, operating at the earliest stage of company formation where both the potential upside and the risk are highest. Sequoia Capital and Founders Fund operate across the full arc, from a first seed check to late growth; Founders Fund closed a roughly $6 billion growth fund in 2026 aimed at very large later-stage checks. Lux Capital, DCVC, and Khosla Ventures tend to enter early and keep investing as companies mature. Founders should match a firm’s focus to where their company actually sits.

How can an LP get access to top deep tech funds?

Access is the hard part. The strongest funds are often oversubscribed and lean on long-standing relationships. Many firms now have dedicated people for exactly this: Adam Kalish handles investor relations at Lux Capital, ACME Capital brought on Katera Mujadidi to lead the function, and Founders Fund relies on Lauren Gross, Partner and COO. Beyond the listicle names, large franchises such as Bessemer Venture Partners (Anita Umarji, Partner and Head of Investor Relations) and NEA (Anna Golynskaya, Partner, Head of Capital Formation and LP Relations) run formal capital-formation teams. New LPs usually build in through smaller commitments and warm introductions, though you would be surprised how often LinkedIn catches a partner’s attention.

Which firms invest in aerospace, defense, and advanced manufacturing?

This is a crowded, competitive corner of the market. Founders Fund helped define it with early stakes in SpaceX and Anduril. ACME Capital invests across aerospace and defense and next-generation manufacturing, with portfolio companies like True Anomaly in space security. Eclipse Ventures concentrates on manufacturing, supply chains, and the hardware-software systems that run them. Khosla Ventures, well known for OpenAI, and DCVC back industrial and frontier technologies too. For founders in these sectors, the practical question is which firm has done the most relevant work before, and which partner will actually pick up the phone.

Is deep tech riskier than software venture?

In some ways, yes. Deep tech companies tend to need more capital and take longer to reach a product, and they carry a risk a software startup does not: the underlying science has to work at all. The flip side is that success can be very hard to copy. That moat is part of why investors like Lux Capital, DCVC, and Khosla Ventures accept the longer odds and timelines. The risk profile differs from software venture rather than strictly exceeding it, and matching investor expectations to that reality matters for everyone at the table.