At Arbor, a company developing power systems that also capture carbon, an early engineering question sent the team to Home Depot. They needed data on an oxy-combustion approach. The model they built cost $2,000. For a business contemplating power plants, this is an amusingly small receipt. It also captures something about the investor behind it: Gigascale Capital wants founders to learn what matters before the expensive decisions become permanent.
- Backs climate-focused physical technology from pre-seed to Series A.
- Pairs investment with help on science, hiring, customers and financing.
- Raised a $250 million first institutional fund in June 2026.
Founded in 2023 by former Meta CTO Mike Schroepfer, climate investor Victoria Beasley and chemical engineer Evaline Tsai, Gigascale occupies a particular corner of venture capital. Its subjects include electricity, industrial materials and the equipment underneath modern life. These are businesses where success eventually requires something heavier than a download.
The expensive distance between works and sells
A laboratory can establish that an idea works. A factory must establish that it works repeatedly, within tolerances, at a price someone will pay. Between those achievements sit suppliers, production yields, permits and purchasing departments. The distance is easy to underestimate when the prototype is impressive.
That distance gives Gigascale its purpose. The firm invests in young companies and helps with the decisions that connect technical progress to commercial survival. Founders receive capital and access to people who can discuss engineering, recruiting and financing together. Investors in its funds supply money in pursuit of venture returns; the portfolio companies remain independent businesses.
Consider the financing mismatch. A prototype can be funded with equity. A commercial facility may require lenders, project finance or other capital with different expectations. The next technical milestone therefore needs to make sense to the next source of money. A clever machine can still be a poorly financed company.
Three kinds of useful experience
Schroepfer’s biography supplies the familiar Silicon Valley connection. At Meta, he helped scale products to billions of users and oversaw substantial data-center construction. That second detail matters here. Even a digital business eventually encounters land, buildings, cooling and electricity. Someone has to arrange the physical world underneath the screen.
Beasley brings experience closer to the factory floor. Before investing at Prelude Ventures, she led finance at solar manufacturer Suniva, working on manufacturing expansion and cost reduction. A solar cell’s silver content is a financial question as well as a materials question. Her background joins those conversations.
Tsai supplies another bridge. She earned a Cambridge doctorate in chemical engineering and biotechnology, co-founded Cambridge Cancer Genomics and worked on AI products at Samsara. Gigascale promoted her to partner in April 2026. Scientific judgment and company-building experience sit together in that résumé.
“The best hardware companies are the fastest iterators.”
Mike Schroepfer · South Park Commons conversation
Buy the answer first
Arbor’s inexpensive test illustrates the appeal of speed. Its proposed system converts organic waste into energy while separating carbon dioxide for storage. Before committing to larger equipment, the team found a modest way to investigate a specific combustion question. The useful lesson is about choosing the size of the experiment.
Another portfolio company, Xcimer Energy, faced the same issue at a different price. Its founders were planning Phoenix, a laser-development platform for their fusion approach. They initially intended to test one configuration. Then came the uncomfortable arithmetic: construction could take two years and tens of millions of dollars.
Technical advisers asked what would happen if the team’s understanding changed. Xcimer redesigned the platform to explore multiple parameters, including different energies and pressures. The original plan had become too narrow for the learning ahead. This was a design reconsideration before construction, rather than a story about a failed reactor.
What can this test teach us before we commit?
A customer has to want the thing
Gigascale’s portfolio spreads across technologies, but the commercial question stays fairly constant. Dioxycle, for example, is developing a process that uses electricity to turn carbon emissions into ethylene, a chemical building block. Its stated design constraint is price: industrial buyers need a product that competes economically with conventional supply.
The distinction is useful. Emissions savings explain why a technology matters. Delivered cost and performance help explain why a customer changes suppliers. Dioxycle’s founders made that purchasing decision part of the engineering problem from the beginning.
Other investments attack different bottlenecks. Form Energy develops multi-day energy storage. Radiant is developing nuclear microreactors. Mill makes food-recycling systems. Panthalassa is working on ocean-wave energy nodes for uses including offshore computing. These companies have different customers and development schedules; listing them together does not make their technical risks interchangeable.
The market also contains investors such as Breakthrough Energy Ventures and Lowercarbon Capital. Climate ambition alone offers little distinction. Gigascale’s pitch is the combination of early scientific conviction and practical assistance through commercialization. Founders can test that claim by asking for examples of hires, pilots and financing work.
That help need not look glamorous. A founder may need a manufacturing leader before another researcher, or a conversation with a utility before another prototype revision. Gigascale’s team includes dedicated talent, partnerships and finance roles alongside investors. Its public invitation to founders is unusually direct: bring the problem you are working through. The working relationship is part of the product.
Before the first factory
For a founder, the most portable idea is to investigate the buyer while the design can still change. Who owns the budget? What evidence gets a pilot approved? How long does procurement take? Those answers can alter which engineering problem deserves attention first.
The same approach offers something to people outside the fundraising process. An engineer evaluating a new system can ask which assumption has actually been tested. A prospective customer can distinguish an attractive laboratory result from evidence of dependable operation. A potential hire can ask what must happen before the company earns its next financing. The questions make progress more legible, even when the answers are inconvenient.
There are limits. A cheap experiment cannot prove a plant will run reliably for years. A customer’s enthusiasm cannot finance construction. Manufacturing economics, follow-on capital and deployment timelines still have to cooperate. Gigascale’s bet is that people who understand those constraints can help founders discover them earlier. The $2,000 receipt is a useful place to start.