Chad Rigetti just raised $139 million for a company that does not yet have a product. Four years ago, his last company went public at a $1.5 billion valuation and crashed to $0.38 per share. That is the actual news.
Sygaldry Technologies, founded in 2024, announced the funding round on Tuesday: a $34 million seed led by Initialized Capital closed in August, and a $105 million Series A led by Breakthrough Energy Ventures closed in March, according to a press release on GlobeNewswire. The pitch, in the company's own words, is to build quantum-accelerated AI servers for data centers and have them in commercial production by the end of the decade. Rigetti, who founded the original Rigetti Computing in 2013 and ran it until investors pushed him out in November 2022, is the CEO and co-founder. He is joined by Idalia Friedson, who helped take Rigetti Computing public in 2022 as its chief strategy officer, and Michael Keiser, an AI scientist.
The money is real. The timeline is not established.
Here is what the press release does not say: no quantum computer has demonstrated an unambiguous speedup on a real AI workload. Prediction markets tracking quantum progress show overwhelming skepticism that any quantum system will deliver an unambiguous, classically impossible computation by 2026. The category that Sygaldry is selling, quantum-accelerated AI, requires integrating quantum processors alongside classical chips in a way that actually accelerates the parts of AI inference that matter. Nobody has published a peer-reviewed result showing that a hybrid quantum-classical system beats a GPU cluster on a task anyone cares about.
This is not a secret. It is the entire problem the field has been trying to solve for a decade.
The energy framing is where the pitch gets more interesting. Sygaldry's announcement cites an estimate that the AI industry will need $5.2 trillion in capital expenditure and 125 gigawatts of new power generation by 2030 to meet global demand. If a quantum-classical hybrid could match GPU performance at meaningfully lower energy consumption per inference, that would matter regardless of whether the speedup story holds. This is the angle that makes the pitch testable and interesting, because energy efficiency is something you can measure in a data center in a way that quantum advantage is not.
Rigetti has made this argument before. In 2022, he told investors that quantum computing would transform drug discovery and materials science. The stock peaked at $8.81 in May 2022 and fell 96% over the following year. Two delistings followed. He was removed as CEO in November 2022 and replaced by Subodh Kulkarni. Rigetti Computing, which had been one of the most prominent US-based quantum hardware companies, eventually ended up on the Nasdaq delisting watchlist twice in eighteen months.
The co-founder and investor community around Sygaldry appears to believe the second act will be different. Breakthrough Energy Ventures, which led the Series A, does not fund science experiments. It funds companies that have a credible path to a product in a large market. That is meaningful signal. Initialized Capital, which led the seed, has a portfolio that skews toward infrastructure. The fact that both funds participated suggests the pitch passes a commercial bar that the first Rigetti pitch apparently did not.
But the technical question remains open. Sygaldry is building superconducting qubit systems, the same physical modality that Rigetti Computing used. Superconducting computers are relatively cost-intensive to manufacture compared to other qubit architectures, which has been a persistent structural challenge for the sector. Whether that cost profile is compatible with the energy-efficiency argument depends entirely on whether the quantum acceleration claim holds at scale, and on workloads that data centers actually run.
The most honest version of this story is not quantum computing startup raises money or even founder redemption arc. It is a specific technical bet: that hybrid quantum-classical chips will hit energy efficiency thresholds for AI inference that GPUs cannot match, and that this will happen before the end of the decade. That is a bet worth tracking. It is also a bet that the entire field has been making, with varying degrees of honesty, since 2013.
What Sygaldry has that the original Rigetti did not is a co-founder who was not there for the first version. Michael Keiser's AI science background is the piece that did not exist in 2013. Whether that changes the calculus is the question this company will answer in the lab, not in the press release.
