Simulations on ATERUI III grow Little Red Dots from radiation-forged black hole seeds
A team led by Sunmyon Chon ran nested cosmological simulations on Japan's ATERUI III and got black holes growing dozens of times faster than anything can today, with properties the release says resemble Webb's dots.
The Scientist · Science desk
What happened
- New simulations run on Japan's ATERUI III supercomputer suggest the Little Red Dots that JWST keeps finding are early-universe black holes caught in a phase of very fast growth.
- The team reports that the properties of the simulated black holes closely resemble the Little Red Dots seen by Webb, which would make the dots an early stage of that rapid growth.
- The puzzle behind the work is that black holes of millions to billions of solar masses already existed less than 600 million years after the Big Bang.
Why it matters
- What it limits constraint If ordinary radiation hydrodynamics and known accretion physics suffice, the early black hole problem stops being a search for exotic seed channels and becomes a question of whether the cloud-scale physics is modelled correctly.
- What's now possible capability A single run that resolves both young-galaxy scale and individual gas clouds lets seed formation and the first accretion phase be computed together, so neither has to be inserted as an assumption about the other.
- Who has to decide decision The abundance claim is the one observers can attack first: how often young galaxies sat next to bright enough neighbours sets a predicted sky density that Webb's growing catalogue of dots can be counted against.
Next story · Science · 5h ago
Baker lab's from-scratch zinc proteases cut their chosen target in 36% of 135 designsConfirmed2 publishers
Far-ultraviolet radiation is the part of this picture that has to be common. In the simulations, FUV light from neighbouring galaxies keeps a gas cloud from cooling into many ordinary stars, so the cloud collapses into a single supermassive star [5], which then collapses into a black hole seed [6]. Seeds made that way start heavy, and a heavy seed needs fewer doublings to reach a billion solar masses [13].
Speed is the second ingredient. Dense disks around the seeds trap radiation and let them take up material more efficiently than black holes can in the present-day universe [7], by dozens of times [8]. "Dozens" is the release's word and not a measured multiplier. Take it as thirty: a doubling that would need 100 million years at modern rates takes a little over 3 million [16]. Compression of roughly that order is what the observations require, because black holes of millions to billions of solar masses already existed less than 600 million years after the Big Bang [10].
The design is a zoom-in. The run begins at the scale of a young galaxy, then refines into progressively smaller regions until individual clouds of gas are resolved, which is what the high-resolution capacity of ATERUI III was for [4]. Seed formation and the first phase of accretion happen at scales far apart, and one calculation covers both.
The dots themselves are still unidentified objects: tiny, extremely red, and present in numbers nobody expected before Webb [11]. The National Institutes of Natural Sciences release, dated September 17, says the simulated black holes closely resemble them [3][12], and that no unusual physics or highly unlikely event was needed to produce them [9]. It does not say which observed properties were compared, or name a journal for the work [14]. Matching properties in a simulation makes this growth channel viable; identifying the dots would take an observation that separates it from other ways of making the same appearance.
What to watch
- Whether independent groups reproduce collapse into a single supermassive star under the same far-ultraviolet conditions.
- Whether JWST spectra of Little Red Dots show the dense, radiation-trapping gas the model puts around these seeds.
- Whether the simulated growth phase ends at the millions-to-billions of solar masses already seen before 600 million years.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence62
- Adoption
- Not scored
- Hype gap+35
- Incentives40
- Confidence58
Perspective Coverage
3 publishers- Builder
- Builder 69%
- Operator
- Operator 23%
- Investor
- Investor 8%
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
New supercomputer simulations performed with Japan's ATERUI III supercomputer suggest JWST's puzzling Little Red Dots are rapidly growing black holes from the early Universe.
- [2]
The research was led by Sunmyon Chon of the Max Planck Institute for Astrophysics, using the ATERUI III supercomputer at the National Astronomical Observatory of Japan.
- [3]
The properties of the simulated black holes closely resemble the Little Red Dots observed by the James Webb Space Telescope, suggesting the red objects could represent an early stage of extremely rapid black hole growth.
- [4]
The simulations began on the scale of a young galaxy and then progressively focused on smaller regions, eventually reaching individual clouds of gas; this demanding approach required the high-resolution computing power of ATERUI III.
- [5]
Intense far-ultraviolet radiation from nearby galaxies can prevent ordinary stars from forming inside some gas clouds; instead of breaking apart into many smaller stars, the gas can collapse into a single supermassive star.
- [6]
That enormous star can then collapse and create a black hole seed.
- [7]
Black hole seeds can become surrounded by thick disks of dense gas; these gas-rich environments trap radiation and allow the black holes to consume material much more efficiently than black holes can in the modern Universe.
- [8]
Under these conditions, the black holes can grow dozens of times faster than would be possible today.
- [9]
The findings do not require unusual physics or highly unlikely events; no exotic assumptions or rare coincidences were necessary.
- [10]
Some early supermassive black holes contain millions or even billions of times the mass of the Sun, yet they already existed less than 600 million years after the Big Bang.
- [11]
JWST revealed a large population of tiny, extremely red objects that researchers began calling Little Red Dots, and their true nature has remained unclear.
- [12]
The release is dated September 17, 2026 and is credited to the National Institutes of Natural Sciences.
- [13]
A more massive black hole seed requires fewer mass doublings to reach a given final mass such as a billion solar masses.
- [14]
The release does not name a journal for the study and does not report which observed Little Red Dot properties the simulated objects match or how many simulated objects were produced.
- [15]
According to the model, the conditions needed to create and rapidly grow these black holes developed naturally in the early Universe, which could also help explain why Little Red Dots appear to be so common in JWST observations.
- [16]
If 'dozens of times faster' is taken as thirty times, a mass doubling that would take 100 million years at modern accretion rates takes about 3.3 million years.
Sources
2 independent publishers whose own reporting we read for this story.
- nature.comOvermassive black holes and little red dots naturally form in simulations
1 article · September 15, 2026
- phys.orgJapanese supercomputer simulations may explain Webb's Little Red Dots
1 article · September 17, 2026
- sciencedaily.comJWST’s mysterious little red dots may be black holes growing at incredible speeds
1 article · September 17, 2026
Topics and entities
Topics
- Black Hole Formation and Growth
- Cosmological simulations
- High-redshift astronomy
Entities
- National Institutes of Natural Sciences
- National Astronomical Observatory of Japan
- Max Planck Institute for Astrophysics
- Nature
- James Webb Space Telescope