Astronomers just released a 5.6-trillion-pixel map of the universe — with 4 billion objects

The biggest 2D map of the universe is now public, and it's almost incomprehensibly large. Built by the Dark Energy Spectroscopic Instrument (DESI) team, the map contains 5.6 trillion pixels and captures nearly 4 billion stars, galaxies, and other celestial objects across 75% of the visible sky. For anyone who's ever wanted to zoom through the cosmos on a screen, the Legacy Survey Sky Viewer lets you do exactly that — for free.

13 years in the making

The map is the result of 263,407 individual exposures taken over 13 years using telescopes at Kitt Peak National Observatory and Cerro Tololo Inter-American Observatory, among other NSF facilities. Images span visible light and near-infrared wavelengths, giving scientists a detailed look at objects far too faint for any amateur setup to detect. According to the NOIRLab press release, the final dataset was processed over eight weeks on the Perlmutter supercomputer — one of the most powerful machines in the world — using code that took a year to develop.

The imaging data was originally designed as a targeting tool for the DESI spectrograph, which measures the redshift of galaxies and quasars to calculate how far away they are. That redshift data is what turns a flat photograph of the sky into a full 3D model of the universe.

Dark energy under question

In April 2026, DESI completed its five-year survey ahead of schedule, logging 47 million galaxy and quasar observations — well above the original target of 34 million. As the Berkeley Lab news reports, early results hint that dark energy — the mysterious force driving the universe's expansion — may be weakening over cosmic time. That would challenge a cornerstone assumption of modern cosmology that has stood for 25 years. Full five-year dark energy findings are expected in 2027.

For context: the European Space Agency's Gaia telescope has catalogued around 1.8 billion objects, but its focus is primarily our own Milky Way. DESI's survey looks far deeper into the universe, reaching structures billions of light-years away.

What it's used for

Previous versions of the Legacy Surveys catalog have already been cited in more than 1,800 published scientific papers. The new release will serve as a calibration reference for the upcoming Vera C. Rubin Observatory, and the dataset has been designated for AI classifier training under the US Department of Energy's Genesis Mission. In practical terms, that means machine-learning models will be trained on this data to automatically identify and categorize rare cosmic objects at a scale no human team could manage manually.

The map is publicly accessible now — no login, no paywall.