DOE's $215M quantum competition sets its pass mark at 100 logical qubits

A researcher who wants to run a real calculation on a quantum computer today is booking time on an experiment. The Energy Department's own description of the field is that these devices have so far been more the subject of scientific experimentation than useful tools in their own right. They are hard to build, and sensitive enough to outside interference that components are often held near absolute zero ^[15].

The $100 million pool and the two $50 million pools add up to $200 million. That leaves $15 million of the announced $215 million for phase one, where awards are fixed at up to $1.5 million for meeting early milestones ^[7]. At full value, that funds ten of them ^[8]. The department did not state that phase-one figure; it is what is left after the published pools are subtracted. Half the prize money sits on clearing the entry bar, and each 50-qubit step above it carries the same $50 million ^[5][6].

The definition is the part a planning team can use. DOE is counting logical qubits, meaning the qubits effectively available for computation after error-correction machinery has been paid for ^[4]. A physical qubit count, which is what most launch slides lead with, does not convert into that number without knowing the error-correction overhead.

"I see it as no different than getting achievements where we could build the first particle accelerators, or incredibly new, capable telescopes," said Darío Gil, the DOE's undersecretary for science, describing the target as a scientific instrument ^[10]. Tanner Crowder, quantum information science lead in the Office of Science, said the milestone requirements are meant to produce prototypes that scale: "Our goal is not to have a bunch of qubits that can't do much" ^[11].

Officials are hopeful early machines could be operational as soon as 2028. The department's stated goal is technology useful to scientists by then, in areas including drug discovery, astrophysics, and secure communications ^[9]. The prize pays on demonstration. The announcement does not commit a laboratory to buying a machine, or a company to delivering one on a fixed date, so the 2028 figure is a DOE hope with money attached to the specification behind it.

For a lab computing group writing a 2028 budget, the usable artifact is the acceptance test, and it converts into three questions for any vendor roadmap: how many logical qubits after error correction, how many fault-tolerant operations in a single run, and which pool the entry would qualify for, the $100 million one or the $50 million tiers. Crowder said awards could be announced as soon as November or December: "We hope to start work as soon as possible. There's an obvious urgency to us being able to deliver this capability for the nation" ^[12].