Without knowing anything my guess is they are making layers of some graphene sheets with vapor deposition coating and making a roll or a sandwhich of that. I would imagine it will be extremely expensive.
I wonder how this compares to Microsoft’s Project Silica. It seems like it is much more dense in terms of data storage, but it also sounds like the data is etched into the surface of the ceramic vs Silica’s data being stored within the glass. Obviously still miles more durable than most modern data storage methods.
Initially, there will be no need for particle beams/electron microscopy
It sounds like this might never be viable in the consumer space. Pity. The world is more cloud-centered than I’d like as it is. I want 1PB in my pocket!
Yes, the CeraByte web site is suspiciously devoid of any mention of rewrite-ability (or lack thereof). They just mention reading and writing. www.cerabyte.com/how-it-works/
More questions than answers, looking through their web site.
After a few minutes of duckduckgoing it, it looks like they are a new company still in the funding phase. They are due to present at the Storage Developer Conference later this month, but I’m not sure they actually have a product yet. storagedeveloper.org/events/agenda/session/527
In the abstract they focus on cold storage, but also mention “the ability to fully recycle the media”.
Depending on necessary speed, Tape Drives fit that use case pretty well. This feels like it could be a slight improvement on that format, but it’ll entirely depend on density
For sure. Also, if the density is sufficiently high compared to alternatives, it could be objectively better anyway. For instance, a typical SSD is rated for less than 1000 full write cycles. So if I have a write-once media with more than 1000 times the space, I’ll be able to write more to it even in the worst-case scenario.
In the film industry, long term storage of digital films is a real issue. Disney still creates a technicolor (3strip) copy of their films as digital data isn’t as reliable as good old silver emulation.
It makes sense from a layman’s perspective on why it can’t be rewritten, but this is really just a single write storage mechanism. May be really good for archival data, but anything that needs active use is probably far from feasible
They said the same thong about ssds for mass storage. Granted this tech is dramatically different. But we went from coffee table sized laser disks in the 90s to psp disc’s in the 2000s. That’s like a 1:100 size reduction and at a 1000 x data storage rate. Who knows what the future stores.
Write with laser or particle beam. Read with microscope or electron microscope.
Read will be taking a snapshot of the area with the electron beam so it will depend on the speed of said microscope. The idea isn’t for ultra fast write and read access. It’s long term storage with low volume. Currently they’re using tape for long term storage but the longevity is likely less than that of ceramic.
According to the startup, these new formats are poised to address density, performance, and access paradigms, as well as the cost and sustainability demands of datacenters.
Here, for the first time, it will detail how it will introduce CeraMemory with inorganic nanolayers, using 50-100 atoms thick ceramics to store information.
Data reading can be done with equipment using high-resolution microscopic imaging techniques or electron beam microscopy.
Another seemingly excellent inherent property of ceramic storage is the touted media durability and longevity.
Additionally, it is boasted that CeraMemory is resistant to corrosive, acidic, radioactive environments and EMP disruption.
Its presentation at the 2023 Storage Developer Conference, which takes place between September 18 and 21, looks set to be one to watch.
The original article contains 475 words, the summary contains 120 words. Saved 75%. I’m a bot and I’m open source!