The Deep Synoptic Array, a network of small ground dishes working in concert across 12 by 10 miles of high desert, would be the first instrument of its size to release its data openly as it collects it.
Caltech has flipped the model on who gets time on a world-class radio telescope. A new array of 1,650 ground dishes, taking shape in the Nevada desert, will release its observations to the public in near real time, a first for an instrument this large. The change in access is the news, more than the dish count, and it could open a generation of fast-radio-burst, pulsar, and black-hole follow-up work that has had no realistic path to telescope time.
The project is the Deep Synoptic Array, or DSA, a network of small radio dishes working in concert across roughly 12 by 10 miles of high-desert site. Unlike the iconic single-dish radio telescopes most readers have seen in photos, modern radio astronomy leans on arrays: many smaller antennas pointed at the same patch of sky, their signals combined to mimic a much larger instrument. The DSA, according to Caltech's announcement reported by Gizmodo, reached final design the week of June 8 to 11, 2026, and is positioned to map the radio sky faster than any previous array of its kind.
Caltech says the DSA will survey the cosmos "100 times faster than any other known telescope in the world." That figure is the institution's own benchmark, not an independent measurement, and it is worth pausing on. "100 times faster" can mean different things: how quickly the array sweeps a given patch of sky, how soon it can issue a first useful map, or how fast it can return to a transient source for follow-up. The number, sourced to Caltech and reported by Gizmodo's Matthew Phelan, should be read as a project goal under defined survey conditions rather than a settled physics result. The point that does hold up is the operational one: the DSA is designed to cover more sky, more often, than any instrument that came before it.
That matters because the radio sky is not static. Some of the most interesting signals astronomers chase are short-lived: fast radio bursts, millisecond flashes from distant galaxies whose origin is still debated; pulsars, the rapidly spinning remnants of dead stars that act as cosmic clocks; and the low-frequency "burps" from supermassive black holes at the centers of galaxies. Capturing these events is a time-allocation problem as much as a hardware problem. Historically, access to a top-tier radio telescope has been rationed by competitive peer review, and the largest instruments, the now-collapsed Arecibo dish in Puerto Rico and the Green Bank Observatory in West Virginia, have been oversubscribed for decades. Smaller institutions and independent researchers often cannot get a look.
The DSA's explicit answer to that bottleneck is its data policy. The plan, as described in the Caltech announcement reported by Gizmodo, is to publish calibrated data products in near real time, making the petabyte-scale firehose available to any scientist, student, or citizen researcher with the bandwidth to download it. If the pipeline holds up, the bottleneck shifts from competing for time on a giant dish to running the analysis. That is a structural change, not an incremental one. Fast-radio-burst alerts could be issued to a global community in minutes. Pulsar timing campaigns, which currently require years of negotiated telescope time, could be assembled from a single open archive.
There are real caveats. The radio sky is increasingly crowded, and siting a new array in the desert is not the same as solving the radio-quiet problem. Green Bank, in a West Virginia valley, was protected for decades by its remoteness; the Nevada site sits closer to population and to the satellite constellations now blanketing low Earth orbit. NASA's 2023 proposal to place a radio telescope on the lunar farside was an attempt to step around that interference entirely, at much higher cost. Whether the DSA's site can deliver the quiet Caltech's sensitivity claims require is an open question, and the source material flags it as such. Arecibo's 2020 collapse is also a standing reminder that "biggest yet" is not the same as "robust yet." And an open-data promise is only as good as the pipeline behind it: release latency, calibration quality, and documentation are all things that have to be audited by users, not by the institution that built the telescope.
The prototype dishes already exist at Caltech's Owens Valley Radio Observatory, in the Mojave between the Sierra Nevada and the White Mountains. Scaling that hardware up to 1,650 antennas is the engineering task the next year or two of construction will be measured against. If the build goes to plan and the data policy holds, the result is a different kind of radio observatory: one that does not wait for a guest observer to fly to a control room, but simply publishes the sky and lets anyone with a question try to answer it. The science it will most directly enable is the kind that needs a lot of looks, at a lot of objects, on a tight clock. Fast radio bursts, the long pulsar-timing campaigns that test general relativity, and the slow variability of black-hole accretion are exactly that shape of problem, and they have been waiting for an instrument that treats them as a firehose rather than a hand-poured trickle.