Robinhood Ventures Fund II spread its second fund across roughly 80 startups funded through $250K SAFE notes, and beyond the software names sits a hardware cohort: energy, robotics, and defense companies tackling problems that software alone cannot solve. The fund’s shares begin trading on the NYSE on August 13 at $25 apiece, and its prospectus details the physical bets alongside the digital ones. Where the software slice chases the agent economy, this cohort chases the constraints around it: power, movement, and production capacity. It is a different risk profile from the software names, with longer development cycles and supply chains that cannot be fixed with a software update.
The hardware slice is smaller than the software side of the portfolio, but it is deliberate. The $250K SAFE structure the fund favors lets it seed companies across very different markets with the same instrument, and the physical bets run from grid-scale power to defense production. The companies in this cohort share a trait that software bets do not: they sell something that has to be manufactured, shipped, and installed, which makes their timelines longer and their moats harder to copy.
On the power side, Apollo Atomics is developing ultra-compact nuclear reactors designed to fit the footprint of AI data center campuses, where electricity demand is outpacing grid capacity. The company’s bet is that smaller, factory-built reactors can be sited and installed faster than conventional plants, which take years of construction and permitting. Voxel Energy takes a complementary path, pairing solar generation with repurposed battery storage to build data centers that run on their own power. Battery repurposing also extends the life of EV packs that would otherwise be recycled, a cost advantage the company leans on, and the pairing eases the strain on grids that cannot deliver new capacity fast enough. The two companies approach the same problem from opposite ends: Apollo Atomics from the supply side, Voxel Energy from the storage side, and both are chasing the same demand curve. For the fund, that means the energy bets reinforce each other even though they compete for the same customer.
The robotics cluster is equally direct. Eden Robotics sells autonomous robots as a service, letting warehouses and light industrial facilities rent capable machines instead of buying them, trading a big capital outlay for a predictable monthly cost. Shotwell builds observability for robotics, the logging and monitoring layer that fleets need once they leave the lab, covering uptime, debugging, and fleet-wide health. Ornadyne takes an unusual approach with robotic birds for surveillance, mimicking avian flight for low-profile monitoring where traditional drones stand out. Together the three cover the robot lifecycle, from deployment to monitoring to specialized form factors, and robots as a service has become one of the faster-moving corners of the market as operators look for flexibility over ownership.
Defense is a growing theme in the portfolio. Tenet Industries works on low-cost, mass-producible defense systems, betting that volume manufacturing matters more than exotic one-off hardware. The company’s approach echoes the broader shift in defense spending toward systems that can be produced at scale, a theme that has drawn a wave of venture money in recent years. The General Aviation Company is building internet-connected air traffic control, modernizing infrastructure that still runs on legacy radios and paper strips, a slower sell but one regulators are already moving toward.
Anoria rounds out the hardware bets with an emotion-reading wearable, a consumer device that reads physiological signals and turns them into feedback the wearer can act on, a reminder that the fund’s physical bets are not all industrial. It is early-stage consumer hardware, the kind of bet the fund can afford to make small, and it gives the cohort a presence in personal devices rather than just infrastructure. Wearables are a crowded market, so the company will need signal quality to stand out.
The through-line is capacity. Compute needs power, robots need observability, and defense needs production lines, and each of these companies sells into one of those bottlenecks. Each bet is small by design, a $250K check, but together they map the physical layer of the AI buildout, and the fund’s August 13 NYSE listing at $25 a share puts a public price on the whole basket. Between the reactors, the storage, the robots, and the defense lines, the cohort covers the physical systems the agent economy has to plug into.
