Somewhere in the foreground of the Large Magellanic Cloud, two dead stars are locked in an orbit so tight that they complete in just over six minutes what takes Earth a full year. They swing around their common center of mass once every 374 seconds — and according to a paper posted to arXiv on August 10, 2026, that orbit is shrinking faster than any comparable system astronomers have clocked before.
The object is eRASSU J060839.5−704014, a mouthful of a name that encodes its discovery: the eROSITA all-sky survey (eRASSU) picked it up as an X-ray source at that patch of sky. It first showed up in the scientific literature in 2024, when a team led by Chandreyee Maitra and Frank Haberl identified it as a rare double-degenerate ultracompact binary — two white dwarfs orbiting each other so closely that one is actively stripping material from the other. That earlier work, published in Astronomy & Astrophysics, flagged the system as a supersoft X-ray emitter whose 374-second X-ray pulsations were modulated at nearly 100 percent, a signature of a compact, well-behaved accretion geometry rather than a messy or variable one.
Two years on, a new study led by Rahul Sharma of India's Inter-University Centre for Astronomy and Astrophysics (IUCAA), with collaborators from the Max Planck Institute for Extraterrestrial Physics, MIT, and institutions in Germany, the United States, China and Greece, has done something the 2024 discovery paper couldn't: nail down exactly how fast the orbit is decaying.
How do you time a binary that orbits in six minutes?
You watch its pulsations — precisely, and for a long time. The team stitched together "phase-connected" timing data, meaning every single X-ray pulse cycle from the object could be counted continuously without ambiguity, across observations from NASA's NICER instrument on the International Space Station, China's newly operational Einstein Probe, and archival data going back through XMM-Newton. Phase connection is the same technique pulsar astronomers use to track the rotation of neutron stars with extreme precision — it turns a string of individual timing snapshots into one continuous clock.
That clock revealed an orbital period of 374.15013 seconds — accurate to within about 20 microseconds — and, more importantly, a decay rate of −4.7×10⁻¹¹ seconds per second. In other words, the orbital period itself is shrinking by roughly 4.7 hundred-billionths of a second for every second that passes. It sounds infinitesimal, but stacked up over the years spanned by the archival data, it's a clean, statistically significant signal. And it's steeper than the decay rates measured in the two previous benchmark systems for this type of binary, HM Cancri and V407 Vulpeculae — the objects that, until now, defined how fast an ultracompact binary could plausibly shrink.
The Einstein Probe data added a second layer of information: blackbody temperatures of about 126 eV from NICER and about 144 eV from Einstein Probe's Followup X-ray Telescope — soft thermal components that confirm the system's supersoft nature. Phase-resolved spectroscopy also showed the temperature dropping across the bright phase in both instruments, evidence of a structured emission region with real temperature gradients rather than a uniform hot spot.
What's actually driving the shrinkage
Binaries this close don't lose orbital energy the way, say, a satellite in low Earth orbit does, via atmospheric drag. Instead, the dominant loss channel is gravitational radiation: the orbiting masses warp spacetime enough, and fast enough, that the system continuously radiates away orbital angular momentum as gravitational waves, exactly as general relativity predicts for any two massive bodies whipping around each other at close range. Sharma's team's analysis assumes the observed decay is driven almost entirely by this gravitational-wave angular momentum loss, rather than by mass transfer between the stars pushing the orbit around (mass transfer can either widen or tighten an orbit depending on which star is losing mass and how) — an assumption that lets the decay rate be translated directly into a mass estimate.
From the decay rate and the physics of gravitational-wave emission, the team derived a "chirp mass" — a combined-mass parameter that gravitational-wave signals are especially sensitive to — of about 0.43 solar masses, placing the system among the most massive known binaries of its class. The system is also classified as a rare "direct-impact" ultracompact binary, meaning the stream of material torn off the donor white dwarf slams directly onto the surface of its companion rather than first forming a disk, a geometry only a handful of these ultra-short-period systems are known to have.
Why It Matters
The Laser Interferometer Space Antenna (LISA), the space-based gravitational-wave observatory now in development, is expected to be exquisitely sensitive to exactly this kind of source: compact binaries emitting continuous, predictable gravitational waves in the millihertz band. Before LISA flies, though, mission teams need "verification sources" — binaries whose orbital parameters are already known precisely from electromagnetic observations, so that once LISA is collecting data, scientists can check that its gravitational-wave signal for that object matches what's already been measured from the ground and from X-ray telescopes.
A system with a firmly measured period, a firmly measured decay rate and a derived chirp mass, like eRASSU J060839.5−704014, is close to an ideal calibration target: it's now on record as decaying faster than the previous benchmark systems, giving LISA a well-characterized signal to test against. Beyond instrument calibration, tight double-degenerate binaries like this one are also laboratories for a longstanding astrophysical question — how systems like this evolve, whether they merge, and whether some mergers of this kind can detonate as Type Ia supernovae, the "standard candles" cosmologists use to measure the expansion of the universe.
The Einstein Probe's contribution is also notable in its own right. Launched from Xichang on January 9, 2024, aboard a Long March 2C rocket, the Chinese Academy of Sciences-led mission (with the Max Planck Institute for Extraterrestrial Physics, France's CNES, and ESA as partners) carries a wide-field Lobster Eye X-ray telescope alongside two narrower Followup X-ray Telescopes. With operations now extended through 2027–2029, it's one of the newest tools available for exactly this kind of high-cadence timing work — and this result is an early demonstration of what it can add to targets first flagged by eROSITA.
The paper has been accepted to The Astrophysical Journal Letters, and with the underlying timing baseline still growing, further observations should tighten the decay-rate measurement even more — useful, since the tighter that number gets, the more valuable the system becomes as a fixed point against which to check LISA once it launches.
