Key facts
- An international research team created a pulsar-detection method that searches publicly available NASA data roughly one million times faster than earlier techniques.
- The journal Physical Review Letters published the paper describing the results on Thursday.
- The previous best technique did not locate PSR J1622−7204 even though it required computing power equal to 600 years of work on a standard processor.
- Researchers must simultaneously estimate the star's spin rate, its spin-down rate, and its exact position in the sky to isolate the signal.
An international research team found pulsar PSR J1622−7204 in publicly available NASA Fermi data on Nov. 19, 2026, using a detection method roughly one million times faster than earlier techniques. <newline> <newline> A pulsar is a fast-spinning neutron star that emits radiation beams toward Earth like a lighthouse, according to NASA Science’s definition. Finding these stars in Fermi gamma-ray data is extremely difficult because the telescope captures only a few photons from each pulsar, mixed with background noise from other gamma-ray sources and the wider universe. The weak signal must be separated from gamma-ray emissions coming from many other sources across the sky. To isolate the signal, researchers must estimate the star’s spin rate, its spin-down rate, and its exact position in the sky at the same time.
Daniel Pook-Kolb, who led the work at the Max Planck Institute for Gravitational Physics in Germany, said researchers cannot count rotations to locate a photon because they do not observe every pulsar rotation, and he said the work amounts to searching a four-dimensional grid with millions of trials in each direction, beyond what Earth‘s computing power can handle. Previous searches split Fermi data into smaller pieces and depended on volunteers around the world to donate spare computing power through crowdsourcing, NASA Science said. Even with that volunteer network, finding a single pulsar generally took hundreds of computer-years, NASA Science said. Dotan Gazith and Barak Zackay reframed the search as a geometry problem: finding the shortest vector inside a high-dimensional lattice of regularly spaced points, NASA Science said. The approach uses algorithms built by cryptographers trying to defeat encryption designed to resist quantum-computer code-breaking, and the technique has also boosted satellite navigation efficiency, NASA Science said. Each test takes roughly five minutes on a single computer core, so a standard laptop with two to eight cores can handle the work, NASA Science said. Once the lattice method flagged a candidate, the researchers verified it by studying the full Fermi data set, where the signal proved too strong to be random.
The researchers found PSR J1622−7204, an isolated pulsar spinning 261 times per second, using about 10 hours of computing time on a standard processor. Older methods missed PSR J1622−7204 despite using computing power equal to 600 years of work on a standard processor. The comparison shows the new algorithm turned an impractical search into a calculation that ordinary hardware can handle, NASA Science said. NASA Science pointed to the contrast between the new method’s modest computing requirements and the failure of older approaches, even though those older approaches had centuries of processing power available. The paper describing the results appeared in Physical Review Letters on Thursday. NASA Science said the peer-reviewed paper documents the international team’s findings and announced the publication alongside the discovery. The team plans to search for pulsars in binary systems, where a pulsar orbits a companion star, because the star’s gravity distorts the signal and has made these pulsars nearly impossible to find with older methods.
Colin Clark of the Max Planck Institute said the data set probably holds tens or even hundreds of binary pulsars that earlier searches missed. The team included Daniel Pook-Kolb of the Max Planck Institute for Gravitational Physics in Germany and Colin Clark of the Max Planck Institute. Their analysis relied on publicly available NASA data and software tools. Colin Clark said the team will broaden its searches by combining the new algorithm with volunteer computing power from the Einstein@Home program to find many pulsars believed to exist in the data. The program draws on spare computing power contributed by volunteers worldwide. Under NASA’s longstanding Gold Standard Science policy, Fermi science data is freely available to anyone. The discovery would not have happened without that policy, NASA Science said. Fermi launched in 2008, scans the whole sky once every three hours, and observes gamma rays, the highest-energy form of light.
After a one-year calibration period, the complete Fermi data archive became public, and new data is still released as soon as processing finishes. Colin Clark said all 18 years of Fermi data are public and continue to grow, and he described NASA data as entirely open, which allowed outside researchers including the cryptography-inspired team to try a new approach on the same data. NASA’s open software and tools, including Fermitools, supported the work by letting the team prepare the data set before analysis began. The tools let outside researchers work with the Fermi archive without needing mission-internal resources. The catalog now lists thousands of gamma-ray sources, up from a few hundred at the start of the mission. NASA Science said new discoveries keep surfacing as different research groups reuse the public archive. Elizabeth Hays, Fermi project scientist at NASA’s Goddard Space Flight Center in Greenbelt, Maryland, said another team can check any search performed by one team, which improves all the science.
Open data lets independent groups verify each other’s results using the same public observations.





Join the discussion