Microsoft boffins cook up archival storage using Pyrex glass they say can last over 10,000 years
- Reference: 1771460918
- News link: https://www.theregister.co.uk/2026/02/19/microsoft_glass_storage/
- Source link:
The boffins at Redmond have been [1]exploring this concept as part of Project Silica since 2019, using femtosecond lasers to encode data as voxels (3D pixels) inside glass.
Glass is resilient to water, heat, and dust. It also breaks down more slowly than the plastics or mechanical components found in hard disks or magnetic tape drives, both still widely deployed for archive storage media.
[2]
Up until recently, Microsoft's experiments into glass-based storage have required special fused silica glass, which is both challenging and expensive to manufacture. In a paper [3]published in the journal Nature this week, Microsoft researchers now say these long-term storage qualities can be achieved using the same kind of borosilicate glass found in oven doors and Pyrex glassware.
[4]
[5]
In their testing, they were able to etch 258 layers of data totaling roughly 2.02 TB onto a 2 mm thick borosilicate glass plate while achieving write speeds of between 18.4 and 65.9 Mbps depending on the number of laser beams used. That top speed is significantly faster than the 25.6 Mbps achieved using fused silica glass, but with less than half the density, at 2.02 TB vs 4.84 TB per platter. Additionally, they were able to reduce the amount of equipment required to read back data from the plates from three or four cameras to just one.
The way Microsoft is etching the voxel data into the glass hasn't changed — it's still using femtosecond lasers — but its method for doing so has. Early attempts at glass-based storage, including Microsoft's, used "birefringent" voxels, which means they refract light differently depending on their polarization. According to Microsoft, this required multiple laser pulses to encode the data, which they were eventually able to reduce to two. They've now developed a different phase-based voxel, which requires just one laser pulse.
[6]
In addition to reducing the number of laser pulses required to write data to the plates, Microsoft has also managed to increase the number of voxels written in parallel, significantly boosting transfer rates.
As we [7]noted last year, this is just what Microsoft has managed to achieve in the lab and isn't necessarily reflective of the technology's storage density or transfer rate if it's ever productized. For example, the researchers suggest that 16 or more beams writing in parallel could dramatically increase write speeds over the one to four beams used in its borosilicate glass trials.
Also, while more efficient, the approach isn't perfect. The phase-based voxels showed a greater propensity for interference, but the researchers note that this can be effectively mitigated by machine learning-based classification models.
[8]As memory shortage persists, vendor price quotes are not long remembered
[9]AI gets all the good stuff, including Micron's speedy 28 GB/s PCIe 6.0 SSD
[10]UNIX V4 tape successfully recovered: First ever version of UNIX written in C is running again
[11]HAMR time: Seagate unleashes 30 TB disks to feed the AI beast
Finally, Microsoft conducted an accelerated aging test to extrapolate the media's viable storage life. Despite our best efforts, existing storage media have a set life span before they begin to break down and are subject to bit rot. This can happen relatively quickly in the case of flash media, particularly when left unpowered in warm climates.
This doesn't appear to be an issue with glass, with Microsoft's trials supporting a viable storage life of more than 10,000 years.
[12]
Despite Microsoft's ongoing progress into glass-based storage, its future remains somewhat vague. In a [13]blog post published alongside the paper, the company notes that the "research phase is now complete, and we are continuing to consider learnings from Project Silica as we explore the ongoing need for long-term preservation of digital information."
That's not exactly a development road map. "Microsoft, like other companies, takes many considerations into account when determining productization and investment priorities. Microsoft continues to value the intellectual property in Project Silica and is exploring options for how to apply the research learnings," a Microsoft spokesperson told El Reg ®.
Get our [14]Tech Resources
[1] https://www.microsoft.com/en-us/research/blog/project-silicas-advances-in-glass-storage-technology/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/storage&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aZaY8iNsr7TxmJmbjnrfJAAAAYo&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://www.nature.com/articles/s41586-025-10042-w
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/storage&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aZaY8iNsr7TxmJmbjnrfJAAAAYo&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/storage&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aZaY8iNsr7TxmJmbjnrfJAAAAYo&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/storage&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aZaY8iNsr7TxmJmbjnrfJAAAAYo&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2025/06/12/archival_storage_feature/
[8] https://www.theregister.com/2026/02/18/memory_shortage_persists_vendor_change_terms/
[9] https://www.theregister.com/2026/02/17/micron_pcie_6/
[10] https://www.theregister.com/2025/12/23/unix_v4_tape_successfully_recovered/
[11] https://www.theregister.com/2025/07/15/seagate_hamr_drives/
[12] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/storage&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aZaY8iNsr7TxmJmbjnrfJAAAAYo&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[13] https://www.microsoft.com/en-us/research/blog/project-silicas-advances-in-glass-storage-technology/
[14] https://whitepapers.theregister.com/
Do they make their own Pyrex?
Corning, the inventor of Pyrex (and patented although I would be surprised if the patent hadn't expired by now) sold all their commercial cooking glassware to another company and went all in on glass for fiber optics several years ago. The company they sold it to immediately stopped making borosilicate glass (AKA Pyrex) and went to regular glass without changing the name. The fake "Pyrex" is thicker than the original, I still have one bowl of the real thing (with a patent notice on the bottom) and noticed the change immediately. And it has a warning that it can no longer go directly from refrigerator (freezer?) to oven. So buyer beware.
Re: Do they make their own Pyrex?
There's a complicated relationship[ between Pyrex, pyrex and borosilicate: https://www.youtube.com/watch?v=2DKasz4xFC0&t=173
Re: Do they make their own Pyrex?
I came here to say the same.
You will know which one you have in a couple of years time. If its the old one that just keeps going, then its the original, good recipie. If however you have forgotten about it and replaced it several times, when it randomly went ping and wrecked your meal idea for the night, then you had the new cheaper one. Just remember to say thanks to the accountant that changed the recipie as it saved them a few pennies and wrecked the reputation in the process.
I believe you can tell the difference by the hue across the glass when you look at it a bit percularly. There are many articles on line about how to do this.
Re: Do they make their own Pyrex?
Not wrecked enough to bring down prices apparently. Of course I stopped buying new Pyrex approximately when this happened after being sorted to find the new bowls had completely different physical properties.
Re: Do they make their own Pyrex?
There are plenty of specialty glass providers, Corning to make a massive one. I believe they even delivered into this topic a decade or so ago.
There is absolutely no reason for Microsoft to re-invent this wheel, when they need only come up with a performance expectation and ask for a quote.
Re: Do they make their own Pyrex?
I imagine laboratory glassware is still borosilcate glass unless there is something better. I think polycarbonate testtubes were used for introductory chem lab courses at one point.
You would not think kitchen glassware would be in line for enshittification but clearly nothing is safe. Ironically with traditional stoneware you get what you see even if it has a shit brown glaze.
Re: Do they make their own Pyrex?
Most laboratory glassware that needs to handle heat or cold swings is still borosilicate glass, although there's also quartz glass (can handle very high temperatures), actinic glass to block UV and IR radiation and siliconized glassware to prevent organic samples from sticking to the glassware.
There are plenty of non-laboratory grade test tubes available made out of the cheapest glass available at the time (soda lime glass?), but plastic is cheaper and works fine if not producing something particularly corrosive or very exothermic or endothermic (most introductory chem lab courses).
It is rumored that some cases of new large-scale meth production had to use stolen laboratory glassware instead of normal Pyrex, after Pyrex was changed to soda-lime glass which could well shatter with a large batch, but there are other sources of borosilicate glass
Re: Do they make their own Pyrex?
One easy to tell difference between real Pyrex and ersatz "pyrex" is the color. If the glass has a greenish tinge, that's a dead giveaway the cookware is made from soda-lime glass and not borosilicate glass.
This is incredibly old as far as digital storage technology goes, it looks like the actual development is entirely in write speed, which would obviously get a boost from using modern pulsed laser designs for both bandwidth and density. I would say this is within reach as a DIY project in quite a few cases already, so Microsoft's showing is just on the level of a tech demo by an intern, which it likely was.
I had been thinking about dusting off the same thing to work on a very old project combined with something new, is it something in the air?
Eh, it's Microslop
The big question here is how, if they decide to move forward, they'll ruin it with AI. Slopya wouldn't let anything like this move forward without somehow ramming Copilot sideways up its arse. Perhaps every cartridge will ship with an LLM that will be your only way to actually read the data? Yes, there is a slight bit of CNN involved in reading the data, but that won't be annoying enough for Slopya.
Oh, and then you'll need a monthly subscription so MS can pretend it's making any actual AI revenue. So 10,000 years in the future, the aliens will find these in the smoking ruins of Earth, attempt to access the information and, 'Glork, it says we need an Enterprise Subscription to Office, do we have one?'
Heechee fans?
In 10,000 years who will be able to read the data when knowledge and tools to do so will itself be lost? Will what's left of humanity (never mind some alien) decorate their walls with "this pretty engraved glassy stuff"? Or will there be bootstrapping instructions engraved on granite slabs: "first, create a civilisation able to read these tablets".
With every mistake, we must surely be learning.
Which company's knowledge would we most likely want to protect the future from?