Optical data storage

Every byte you have ever saved is living on borrowed time. Hard drives wear out, flash cells leak their charge, and magnetic tape has to be copied to fresh cartridges every decade or so just to stay readable. For most data that is fine. For the records a society actually needs to keep — land registries, court rulings, medical history, scientific results, cultural heritage — it is a slow-motion problem that costs money, energy and attention for as long as the data exists.

That is why one of the oldest ideas in computing is getting a second look. Optical data storage, the technology behind the CD and the Blu-ray disc, is being reinvented in glass, ceramics and film as a way to store information for centuries rather than years. Here is what it is and how it works.

Reading data with light

Optical data storage means that information is read back using light. Each bit is stored as a tiny, local change in how a material interacts with light. A reader shines light on the medium and measures what comes back or passes through: more or less reflection, a shift in colour, a dark spot or a clear one.

That is the fundamental difference from the storage in most data centres today. A hard drive or a tape cartridge stores bits as magnetised regions, read by a magnetic head. A solid-state drive stores them as electrical charge trapped in silicon. In both cases the information is held in a physical state that slowly drifts, and the medium needs power, cooling or regular rewriting to stay healthy.

In optical storage, the change is usually structural or chemical: a pit, a burned mark, a modified patch of glass. Once it is made, it tends to stay made. Light does not wear the medium out when it reads it, and the data does not depend on a magnetic field or an electrical charge holding steady.

It is worth being precise about what the word covers. "Optical" describes how the data is read, not how long it lasts. A cheap recordable disc with an organic dye layer can degrade within a couple of decades, while a pattern etched into quartz glass may outlive the building it is stored in. The material, and the way the mark is made, decide the lifespan.

Plastic, glass, ceramics and filmCopy Anchor Link

Optical storage is not one technology but a family. What separates the members is the material that holds the data and the physical trick used to mark it.

The newer glass and ceramic approaches share one idea: write data into a material that is already chemically stable for geological timescales, rather than protecting a fragile material from its environment. Glass does not rust, does not care about magnetic fields and survives water, heat and dust. Some of the glass systems go further by encoding information in the orientation and strength of each structure as well as its position, packing several bits into a single point. That is where the popular term "5D storage" comes from.

Film sits slightly apart, and it is the oldest member of the family. Microfilm has been used by libraries and archives since the 1930s to preserve newspapers, land registers and manuscripts, and it is optical storage in the most literal sense: a photograph of information, read with light and a lens. Preservation microfilm uses a silver-gelatin image on a polyester base. When it is made and stored to international standards it carries a life-expectancy rating of 500 years.

The principle has since been extended from documents to digital data. Modern archival film, like piqlFilm, can carry any kind of file encoded as high-density 2D codes, alongside text and images written in plain view. It is the only optical medium where information can be stored in a form a person can look at directly.

The pros and consCopy Anchor Link

The strongest argument for optical storage is that it costs almost nothing to keep data once it is written. Magnetic archives have to be migrated to new media every few years, consuming staff time, hardware budgets and electricity. A written piece of glass or film can sit on a shelf, unpowered, for decades.

Advantages

• Longevity. The most stable media are rated for centuries or more, compared with roughly 10–30 years for tape and hard drives.
• No power at rest. Data does not need electricity, cooling or periodic refreshing to survive.
• Immutability. Most archival optical media are write-once. Data cannot be silently altered or encrypted by ransomware after the fact.
• Resilience. Optical media are unaffected by magnetic fields and electromagnetic pulses, and glass and ceramic media tolerate heat, water and radiation.
• Fewer migrations. Less copying means fewer chances to introduce errors, and a lower total cost over the life of the archive.

Disadvantages

• Speed. Writing is slow. Several emerging systems write at single-digit megabytes per second today, far below tape.
• Write-once. What makes the media secure also makes them unsuitable for data that changes.
• Expensive readers and writers. New formats need new hardware, and early equipment costs thousands to tens of thousands of dollars.
• A shrinking disc market. Several large manufacturers have stopped producing recordable discs and disc-based archive systems in recent years, leaving fewer suppliers for the established formats.
• Immature market. Lots of the new emerging media being used is still just in testing phase, and have not been proven at scale, both in terms of writing and reading. Some of the research projects will never be commercially available.
• Dependence on a reader. A medium that lasts 10,000 years is of little use if the only machine that can read it disappears in 20.

The last point is the one the industry talks about least. Durability of the medium and longevity of the information are not the same thing.

Where things stand in 2026Copy Anchor Link

The optical storage market is in the middle of a changing of the guard. The disc-based archive systems that were sold to broadcasters and data centres in the 2010s are being wound down, and recordable Blu-ray production has shrunk to a handful of suppliers. At the same time, a new generation of start-ups and research labs is moving glass, ceramic and holographic storage out of the lab.

Most of that new generation is at the pilot stage. None of these is yet a product you can order off the shelf, and the timelines most of them publish point to 2028–2030 for full-scale systems. Still, there is a lot of uncertainty around these initiatives as several of them already have been postponed multiple times in the past. What is commercially available today is narrower: archival-grade discs, and photosensitive film, which is already in operational use in national archives and long-term vaults.

Self-contained: storage that carries its own keyCopy Anchor Link

Most of the race in optical storage is about squeezing more data into less space: hundreds of terabytes on a glass disc, petabytes in a rack. It is an impressive engineering contest, but it skips an awkward question. What is the point of storing a nation's archive on a coaster-sized piece of glass if, in 200 years, nobody can read it back?

The denser the data, the more sophisticated the machine needed to recover it. Reading nanostructures inside glass takes precision optics, polarised-light microscopy and a machine-learning model trained on that exact format. All of that has to survive, be understood and still work for as long as the medium does. History is not encouraging here. Clay tablets in Linear A have survived for over 3,000 years and are still unreadable, because the knowledge needed to decode them did not survive with them.

piqlFilm, a photosensitive film developed in Norway, starts from the other end. Instead of asking how much data can fit, it asks what the simplest possible reader is. Digital data is written as 2D codes at a resolution an ordinary camera or scanner can capture. At the start of each reel, in plain human-readable text, sits everything a future user needs: a description of how the data is laid out, the source code for the decoding software, and specifications for the file formats stored on the film. The decoder is also stored in digital form, so it can be rebuilt from the film itself.

In other words, the reel is self-contained. It does not depend on a particular drive, a software version or a supplier still being in business. Recovering the data takes light, magnification and something that can take a picture — technology that is far more likely to exist in 500 years than any specialised reader built today. Documents and images can also be written to the film in visual form, readable with nothing more than a light source and a magnifying glass.

That is why the "low-tech" label is the point, not the weakness. In long-term preservation, every piece of advanced technology between the medium and the reader is a dependency that can be lost. piqlFilm removes as many of those dependencies as possible and writes the rest down on the film. The medium does not just store the information; it explains how to get it back.

ReferencesCopy Anchor Link

• Canadian Conservation Institute: Longevity of Recordable CDs, DVDs and Blu-rays (CCI Notes 19/1)
• Business Wire: New M-DISC to provide up to 1000 years of permanent data storage
• Phys.org: A 3D nanoscale optical disk memory with petabit capacity
• TechRepublic: Microsoft unveils glass storage that could preserve data for 10,000 years
• JetStor: Cerabyte targets 100 petabytes per rack with 5,000-year ceramic storage
• ITdaily: HoloMem aims to replace magnetic tape with holographic tape
• National Archives of Australia: Preserving microforms
• ISO 18901:2010: Processed silver-gelatin-type black-and-white films — Specifications for stability
• Norner Research (2021): Statement of lifetime expectancy (LE) of the piqlFilm
• The Broadcast Bridge: Film preservation for 500 years on film