DARPA separately named IonQ among the selected entrants. Its description says the government team will test whether proposed systems can be built and perform as intended.
Entry into evaluation does not establish a completed utility-scale computer. The issuer's potential agreement value is conditional on funding actions and appropriations; this record does not establish recognized revenue or guaranteed proceeds. The original agreement and initial obligation were not inspected.
The original roadmap describes the proposed architecture and intermediate work needed to build the system. It establishes the issuer's target, not delivery of that target.
This entry checks the original announcement. It does not assess every subsequent IBM milestone or independently establish a delivered Starling system. Logical memory alone is insufficient for universal fault-tolerant computation.
Google Quantum AI and collaborators reported below-threshold surface-code memories: increasing the tested code distance suppressed logical memory errors under the experiment's conditions.
The paper reports an error-corrected memory experiment. The later correction changes Figure 3a's repetition-code axis and legend labels; it does not state that the main surface-code result is retracted.
Preserving encoded information is different from running a useful fault-tolerant application. This entry does not establish independent replication, commercial advantage or profitability.