SPhotonix – 360TB into 5-inch glass disc with femtosecond laser
27 comments
·December 15, 2025biglyburrito
toomuchtodo
Lots of archival applications that can use this today at these speeds, assuming you’re staging the data like you would for tape. It’s slow, but permanent from a longevity perspective. You could fit 200PB of the Internet Archive in ~600 of these 5 inch glass discs. Hopefully speeds improve, along with infra to treat the media similar to an automated tape library.
rbanffy
> You could fit 200PB of the Internet Archive in ~600 of these 5 inch glass discs
It'll only take 900 years to write using a single drive. ;-)
I'll need to be at least as fast (a 100x speedup) and at least as compact as LTO.
toomuchtodo
https://en.wikipedia.org/wiki/A_Canticle_for_Leibowitz
Hope to see rapid improvements in this tech!
rbanffy
With these numbers the use cases seem limited to small batches of data that need the extreme durability, but to fully use this durability, you'd need to launch your archives to highly eliptical orbits that would take them to the further reaches of the solar system, because the Sun will be a white dwarf long before the warranty expires.
fasteo
It might have a place in its current state,as far as write speed goes
war and peace[1] is 3.2 MB in the plain text version, so it will take less than a second to store it.
ycui1986
they claim to increase the speed to 500Mbps in 3-5 years. but femtosecond laser is not semiconductor. there is no exponential scaling law.
femtosecond laser has been running at 80MHz for decades, they cannot just talk to the laser manufacture and ask them casually to increase that to 500MHz or above. so, it is better to take a grain of salt of the claim speed can be increase to 500Mbps in 5 years.
The cost of the femtosecond laser also won't come down easily either. At $75k a pop for the laser alone, who can afford such technology?
summa_tech
We have GHz femtosecond lasers. It's commercialized! Amplitude Systemes has them for sale! We build time references out of them!
You can also make them out of semiconductors. VCSELs and other semiconductor gain media naturally have short upper state lifetimes. It makes for a happy stable femtosecond laser with a semiconductor saturable absorber.
ranger_danger
According to Amplitude's website, they are "GHz Burst" lasers that have a max repetition rate of 2 MHz, so I don't think that would be very practical for storage media.
0cf8612b2e1e
Surely the only intended customers are giant organizations with petabytes of data. Millions of dollars for their archival system does not seem unreasonable. I recall reading a story that movie studios refresh their physical media for their archives every N years to prevent data loss. If these slabs of glass are stable that removes an ongoing operational burden. Just make a single chip with all of the assets per movie.
aydyn
My guess is their plan is to multiplex the laser. How many lasers can be reading/writing at the same time?
reactordev
So cool yet so slow…
Maybe someday we’ll get Star Trek style clear “chips” that store our information and can be read at Gbps speeds.
Soon. Someday soon.
enzosaba
This would be good to dump the internet onto a disk and preserve it for the eternity. Something that Elon Musk would like; he wants to send its grokipedia to space.
rbanffy
If any alien race gets a copy of Grokipedia, we are doomed.
mrbluecoat
Sigh, where's my hologram hard drive?
typpilol
Sounds cool for long term data storage, but they need to get the read write speeds up.
4mbps write and 30mbps read is extremely slow. Even if they achieve their roadmap 500mbps is still slow compared to modern drives.
Better not keep any data you need access to within like 90 days on it or you're toast.
What market is this even aiming for? Bitcoiners?
dummydummy1234
Long term storage, along the lines of tape replacement/supplement?
There are plenty of things that need to be archived in a basement and never read unless the more readily available forms get corrupted.
Having the ability to say as long as the item exists the data exists is valuable, especially with not having to worry about degradation (which happens with tapes/flash/hard drives)
The ability to say that the data is good.
abound
> 4mbps write and 30mbps read is extremely slow.
It's even slower when you consider the 360 TB capacity -- it'd take nearly three years to write to the whole thing.
qdotme
The better question is the seek latency. The bandwidth for read isn’t too horrible, if the seeks can be kept within reason. This is somewhere between tapes and actual pressed optical media (not dyed /re/writeable). Should seek way faster than tape (maybe even on par with BluRay) and 30Mbps read is manageable for doomsday scenarios.
Long term databanks. Libraries. GitHub’s archive bunkers. Microfilm replacements.
ycui1986
at theoretical perspective, if the X-Y plane can be addressed with 16-bit DAC by controlling laser deflection. then to seek any data with in a 4GB address space will have typical latency of 300us with the latest laser scanning technology.
I am not aware any laser scanning technology that can do 16-bit accuracy that has no moving part. so, fundamentally, this is a storage technology with mechanical addressing.
laser can be scanned by acoustic wave, but that itself lack the beam pointing accuracy. the ultrasonic drive frequency will limit how fast is can deflects the laser beam.
netsharc
Had to go back to the article to see the actual speed units. Above commenter meant MBps (Megabytes/second), not millibits/seconds nor megabits/second.
"According to SPhotonix, its current prototypes achieve write speeds of around 4 MBps and read speeds of roughly 30 MBps."
Assuming I did the math right, that means it'd take almost 3 years at max write speed to fill up the 360TB drive. So yeah, not quite ready for public consumption just yet.