IBM Closed Its HRL Laboratories Acquisition Wednesday Without Disclosing a Price — or a Single Spin-Qubit Fidelity Figure

by Arash Shirazi · August 27, 2026

Quantum Computing Correspondent, Frontier Tech Wire

IBM said on Wednesday that it has completed its acquisition of HRL Laboratories, LLC, the Malibu, California research institution jointly owned until now by Boeing and General Motors. In the company's own words, financial terms of the transaction were not disclosed. That was also true of the July 23, 2026 announcement of intent, which said the deal was anticipated to close by the end of the third quarter of 2026, subject to customary closing conditions and regulatory approvals. The August 26 close falls inside that window.

IBM describes HRL as a leading research and development institution. The July release lists its capability areas as silicon-spin qubit engineering and quantum sensing, alongside advanced sensors, high-speed and high-power communications, electronics, advanced manufacturing, materials science and work on novel quantum materials. Boeing and General Motors, described in Wednesday's release as HRL's former owners, will continue to partner with IBM and HRL on quantum applications and advanced technology development.

What the strategic case rests on

The quantum rationale IBM offers is modality diversification. IBM's own quantum processors are superconducting; HRL's are silicon-spin qubits, a different physical platform built in structures closer to conventional semiconductor manufacturing. IBM also cites HRL work in cryogenics, control electronics, packaging and interconnect — the unglamorous layers that determine whether a qubit count on a die translates into a machine anyone can operate.

IBM situates the purchase against roadmap milestones it has previously published: a system it calls Starling, expected by 2029 and described in the July release as 20,000 times more powerful than today's quantum computers and capable of running 100 million quantum operations; a next-generation fault-tolerant machine called Blue Jay, targeted for the mid-2030s and described as capable of a billion quantum operations; and a facility IBM calls Anderson, described as the world's first pure-play quantum wafer foundry. Every one of those is a company target expressed in IBM's own forward-looking language — "expected," "targeted" — for machines that do not yet exist. None is an achieved result, none has been independently verified, and this article reports them only as statements of intent.

The numbers neither release contains

Here is what is not in either document: a purchase price, a headcount, a facilities list, a qubit count, a gate fidelity, an error rate, or a citation to any specific research paper. The only quantities either release attaches to quantum hardware are the forward-looking operation counts on IBM's own roadmap, and those describe machines IBM has not built. Nothing in either document quantifies what HRL has already demonstrated. IBM's language about HRL's silicon-spin work is qualitative — advanced expertise in silicon-spin qubit engineering, robust knowledge of silicon-based spin qubit platforms and surrounding infrastructure. For a transaction whose entire quantum thesis is a technology platform, that is a notable absence, and no responsible reader should infer a performance level from the strategic language.

For a peer-reviewed anchor, the best-known published HRL result in this area is the paper "Universal logic with encoded spin qubits in silicon," first-authored by Aaron J. Weinstein with Matthew G. Borselli as senior author, and published in Nature, volume 615, issue 7954, pages 817 to 822, online on February 6, 2023. Every listed author on that paper gives a single affiliation: HRL Laboratories, LLC, Malibu. The figures below are taken from the peer-reviewed version of record on the journal's article page, which carries no link to a preprint version of this work; that distinction is not pedantry, because preprint and published figures can differ in both magnitude and direction, and where they do, the published figures govern.

That paper reported universal quantum logic across two encoded qubits, each encoded in three physical solid-state spin qubits, using an array of six silicon-28/silicon-germanium quantum dots. Its abstract states fidelities, measured by interleaved randomized benchmarking, of 96.3 percent plus or minus 0.7 percent for encoded controlled-NOT operations and 99.3 percent plus or minus 0.5 percent for encoded SWAP. The frequently quoted average two-qubit Clifford gate fidelity of 97.1 percent plus or minus 0.2 percent is not an abstract figure; it appears in the paper's benchmarking figure, and is cited here as such. All of these are results from a single research demonstration published three and a half years ago, not a description of any product, and not a claim about what HRL can build today or about anything IBM acquired on Wednesday.

A comparison that is easy to make badly

The exchange-only spin qubit approach HRL is associated with has continued to advance elsewhere. A paper titled "Demonstration of an always-on exchange-only spin qubit," published open access in Nature Communications, volume 17, article 4794, on April 3, 2026, reported an average Clifford gate fidelity of 99.86 percent. That is again the peer-reviewed version of record, not a preprint. Two caveats attach to the figure. It is a single-qubit result: the authors use blind randomized benchmarking to characterize, in their words, "the full AEON single-qubit Clifford gate set." And it is the better of two operating modes the paper reports. The 99.86 percent, the leakage error of 0.015 percent per Clifford and the error per exchange pulse of roughly 0.076 percent are all the simultaneous-exchange configuration; the paper's single-exchange configuration shows a higher leakage error, 0.029 percent per Clifford, alongside a lower error per exchange pulse of about 0.060 percent. Quoting the headline number without the mode is how a mixed set of measurements becomes a false progression.

That work is peer-reviewed, but it is not HRL's, and this is the point most likely to be got wrong by anyone reading the two papers together. The article page gives the senior author, Jason R. Petta, at the Center for Quantum Science and Engineering, University of California, Los Angeles, and HRL Laboratories does not appear among the affiliations listed on that page. Nothing in the 2026 paper is a result IBM acquired on Wednesday. Nor are the two sets of figures directly comparable in any direction: the 2026 headline is a single-qubit randomized-benchmarking Clifford fidelity, while HRL's 2023 headline figures describe encoded two-qubit operations across two encoded qubits. A single-qubit gate fidelity and an encoded two-qubit gate fidelity are different quantities measured on different objects. Placing 99.86 percent next to 96.3 percent as though one superseded the other — as though spin qubits had improved from 96.3 to 99.86 percent between 2023 and 2026 — would be a category error, and this desk states plainly that no such progression is claimed here.

Read-through for listed quantum names

Because no price was disclosed, the transaction produces no valuation mark that anyone can apply to a publicly traded quantum company. It does establish that a buyer of IBM's size was willing to absorb an entire multi-disciplinary research institution rather than license technology or hire a team, and that two industrial owners were willing to sell the whole thing while retaining a partnership.

Investors in listed quantum hardware companies should keep the structural facts in view regardless of what any single transaction implies. A large portion of the publicly traded quantum sector is pre-revenue or generates revenue measured in single-digit millions of dollars per quarter against operating losses many times larger, funds itself through equity issuance that dilutes existing holders, and reports scientific progress through company-issued material that has not been peer-reviewed or independently replicated. None of that is changed by a private lab changing hands.

Nothing in this article is a recommendation to buy, sell or hold any security.