Eclipse Ventures Closes $1.3B Across Two Funds to Double Down on Physical AI
TL;DR
Palo Alto-based Eclipse Ventures has closed $1.3 billion across two new funds to double down on what founder Lior Susan calls physical AI, the intersection of artificial intelligence with robotics, hardware, and the industrial systems that actually move, make, and power the real world. Eclipse Fund VI brings in $720 million for early-stage bets, and Early Growth Fund III adds $591 million for companies approaching commercial scale. The raise pushes Eclipse AUM to roughly $10 billion and makes it one of the largest dedicated hard-tech platforms in the world.
Key Takeaways
Physical AI is no longer a thesis, it is a category. When a firm can raise $1.3 billion for a strategy explicitly positioned against software-native AI, the market has accepted the framing. Eclipse has been beating this drum since 2015, when physical-world venture was deeply unfashionable. The Lior Susan patience trade is now paying a category-definition premium.
The $720M early-stage fund is the headline, not the $591M growth vehicle. Early-stage hard-tech is where alpha actually hides. Eclipse writing $720M into seed and Series A physical-AI companies is one of the biggest concentrated early-stage bets on industrial robotics, autonomous systems, and AI-plus-hardware ever assembled. That is where the 30x outcomes will come from if the thesis holds.
Incubation is the real moat. Eclipse has made clear that part of the $1.3 billion will go toward building companies in-house, not just backing them. For deep-hardware categories where founder sourcing is thin and time-to-first-prototype is long, incubation is a structurally better model than waiting for pitch-deck inbound. Expect more Eclipse-internal spinouts in the next 24 months.
Softbank, Tiger, and crossover investors cannot compete here. The hardware, manufacturing, and regulatory complexity of physical AI filters out tourist capital. Eclipse is one of maybe three firms globally with the operational muscle and LP patience to actually back these companies from seed to scale. That is a defensible position, not just a marketing line.
Fund Overview
Fund Names: Eclipse Fund VI and Early Growth Fund III
Fund Sizes: $720 million (Fund VI, early-stage) and $591 million (Early Growth III), totaling approximately $1.3 billion
Stage: Seed and Series A for Fund VI, Series B and later for Early Growth III
Check Size: Not publicly disclosed, but consistent with Eclipse's historical pattern of $5M to $30M initial checks with meaningful reserves
Geography: US-centric, with selective international exposure including UK (Wayve)
Focus: Physical AI, robotics, autonomous systems, manufacturing, battery and energy infrastructure, defense-adjacent hardware
Total AUM: Approximately $10 billion across the platform
Why This Fund Matters
Eclipse Ventures has spent a decade quietly building the most credible hard-tech investment platform in Silicon Valley while the rest of the market chased pure software. The core insight, which Susan has repeated for years, is that software ate the easy parts of the economy first and the hard parts are still up for grabs. Manufacturing, logistics, construction, energy grids, and defense all run on infrastructure that was designed in the 1970s and is now being rebuilt with AI and robotics at the center. The prize is enormous, and the number of firms actually equipped to invest in it is small.
This raise matters because it validates that thesis with real LP capital at scale. Eclipse is reportedly sitting on an LP base that includes top endowments, pension funds, and sovereign wealth allocators. Getting $1.3 billion committed in a market where LPs are cutting back on venture exposure is a direct signal that the physical AI narrative has crossed from contrarian to mainstream. For the rest of the ecosystem, that means more capital will follow, more generalist funds will pretend to have a physical AI thesis, and valuations in the category are about to decouple from the broader AI correction.
The portfolio tells the story. Wayve, a London-based autonomous driving platform, has become one of the highest-valued AI-for-mobility companies in Europe. Cerebras, the wafer-scale AI chip company, has filed for an IPO and is on the cusp of the kind of exit that returns a small fund by itself. Redwood Materials, Arc, Bedrock Robotics, and Mind Robotics fill out a portfolio that is both technically deep and commercially ambitious. These are not prototype-stage science projects, they are companies shipping product into real customers.
The risk for LPs is correlation. A fund that bets on physical AI is implicitly betting on industrial capex cycles, global supply chains, regulatory flexibility around autonomous systems, and the continued willingness of US customers to buy hardware at software-economics multiples. If any of those break, the whole portfolio rerates together. Software portfolios can hide behind long-tail winners, hardware portfolios often cannot.
The Team
Lior Susan founded Eclipse in 2015 after a career that included building and selling hardware companies, which is exactly the operating background this thesis requires. He sits on the board of Cerebras and has been a consistent public voice for the physical-world investment thesis for a decade, long before it became fashionable. That narrative discipline is one of the reasons Eclipse has attracted the operator talent it has.
The broader team includes operating partners with backgrounds in manufacturing, supply chain, and hardware engineering. Unlike most venture firms that staff investment teams almost entirely with ex-bankers and consultants, Eclipse has built a deep bench of partners who have actually run factories and shipped physical product. That operational depth is the hidden asset of the platform and the thing that would be hardest for a new entrant to replicate.
Early Portfolio
Wayve, autonomous driving AI based in the UK. Cerebras, wafer-scale AI chips, currently on the IPO path. Redwood Materials, battery materials recycling. Arc, electric boats for commercial and defense customers. Bedrock Robotics, autonomous construction vehicles. Mind Robotics, industrial robotics lab focused on manufacturing automation.
What This Means for Founders
If you are building at the intersection of AI and physical systems, Eclipse is now the first call. Full stop. The fund can write a lead check at seed, follow you through Series A and B, and then come back for Series C out of Early Growth III. That single-platform capital continuity is something almost no other firm can offer in hard tech, because very few have both early and growth vehicles dedicated to the same thesis. Founders who want to avoid the painful investor rotation between stages should prioritize Eclipse for that reason alone.
The incubation angle is also worth taking seriously. If you have a strong technical idea but a weak team, or a strong team but no product direction, Eclipse has historically been willing to bring founders in-house and help them build. For scientists coming out of national labs or university spinouts, that structure is often a better fit than a traditional YC-style accelerator.
Fund Momentum Take
Eclipse is the single best-positioned firm in venture today for the physical AI category, and this raise cements that position for at least the next five years. The thesis is clear, the team has the scars to execute it, the portfolio is already validating the approach, and the capital base is now large enough to play the category from seed to pre-IPO without needing to bring in coinvestors who do not understand the space. That is an enviable position.
Our honest concern is valuation discipline. When a thesis moves from contrarian to consensus, the dollars chase in and prices get distorted. Eclipse will now have to resist the temptation to overpay for trophy deals in physical AI, because every other generalist fund in the Valley is about to claim the same thesis and drive term sheet auctions up. The firms that win in crowded markets are the ones that walk away, not the ones that lean in. We will be watching Fund VI deployment cadence carefully.
Our bigger speculation, and this is flagged as speculation, is that Eclipse is positioning for an eventual public-market vehicle or a permanent-capital structure. At $10 billion AUM with a portfolio that includes at least two potential mega-exits in the next 24 months, Eclipse is reaching the scale where the traditional 10-year fund structure starts to feel constraining. A listed vehicle or a BDC-style permanent capital arm would let the firm hold winners indefinitely and compound on its best deals. Nothing has been announced, but the math is starting to favor it. Watch that space.
Frequently Asked Questions
What is Eclipse Ventures?
A Palo Alto-based venture capital firm founded in 2015 by Lior Susan, focused on investing in companies that build hardware, robotics, and AI-powered industrial systems. Total AUM is approximately $10 billion after this raise.
How much did Eclipse raise in this announcement?
Approximately $1.3 billion across two funds: Eclipse Fund VI at $720 million for early-stage, and Early Growth Fund III at $591 million for growth-stage companies.
What does physical AI mean?
Eclipse uses the term to describe the convergence of artificial intelligence with robotics, autonomous systems, manufacturing automation, and hardware-intensive infrastructure. It is the opposite of pure software AI.
Who are the biggest portfolio companies?
Wayve, Cerebras, Redwood Materials, Arc, Bedrock Robotics, and Mind Robotics are among the best known. Cerebras is reportedly on an IPO track.
How does Eclipse compare to a16z American Dynamism or Founders Fund?
Eclipse is more narrowly focused on physical-world categories and has a longer track record in the space. a16z American Dynamism and Founders Fund both have overlapping theses but broader mandates. For hard-hardware founders, Eclipse typically has deeper operational support.
Have a fund closing to announce? Submit your fund here.
Need help raising capital? Check out our Fundraising Advisory services.