By Sophie Lindqvist

TU Wien and data storage firm Cerabyte have produced a QR code measuring 1.98 square micrometers, the smallest on record.

The code, engraved into a thin ceramic film, is smaller than most bacteria. It can only be read with an electron microscope. The Guinness Book of Records confirmed the result on March 29, 2026. The new code is 37% the size of the previous record holder, according to TU Wien.

How small is 1.98 square micrometers?

A single square micrometer fits inside the cross-section of a human hair. The new QR code occupies just under two of those squares. Each pixel measures 49 nanometers across. That is about ten times smaller than the wavelength of visible light. The pattern remains invisible under any optical microscope.

Researchers used focused ion beams to carve the code. Focused ion beam milling removes material atom by atom at nanometer precision, a serial process that limits throughput but enables exact placement of every pixel in the grid.

What makes ceramic different?

The QR code sits in a ceramic layer used to coat high-performance cutting tools. Those coatings survive extreme heat and mechanical stress. Erwin Peck and Balint Hajas of TU Wien argued that durability transfers to data: "We conduct research on thin ceramic films, such as those used for coating high-performance cutting tools. For high-performance tools, it is essential that materials remain stable and durable even under extreme conditions."

The same property, the team says, suits archival storage. The medium needs no power. It needs no cooling. Magnetic drives lose data within years when unpowered, Mayrhofer's group noted. Ceramic carriers could hold data for centuries, the researchers claim.

The team projects more than 2 terabytes per A4 sheet at the demonstrated pixel density. That figure is a theoretical extrapolation. Independent benchmarking has not been published.

How durable is "centuries"?

Alexander Kirnbauer of TU Wien framed the work in civilizational terms. "We live in the information age, yet we store our knowledge in media that are astonishingly short-lived," he said. He compared the approach to stone inscriptions readable after thousands of years.

The "centuries" claim rests on material stability at room temperature, not on field testing. No sample has been aged long enough to validate multi-decade retention in independent work.

What still needs proving?

Throughput is the most pressing gap. Focused ion beam writing is slow. Scaling to industrial volumes requires new writing tools. The team has not published a cost model. Competitive benchmarks against LTO tape, fused silica, and synthetic DNA storage have not appeared.

The University of Vienna served as independent verifier for the record attempt. TU Wien's USTEM center provided the high-resolution electron microscopes. Verification of the record itself is documented. Independent verification of the durability claim is not.

What comes next?

Kirnbauer said the team will pursue new ceramic materials, faster writing methods, and scalable manufacturing. The stated goal is to write complex data structures "robustly, quickly, and energy-efficiently" into ceramic films.

TU Wien and Cerabyte have set a record, not shipped a product. The path from laboratory demonstration to commercial archive media will require independent endurance testing, throughput engineering, and unit-economics data before ceramic storage competes with established archival formats.