Stories
Slash Boxes
Comments

SoylentNews is people

posted by takyon on Thursday June 20 2019, @09:05PM   Printer-friendly
from the real-deal? dept.

Submitted via IRC for Bytram

Discovery of a 'holy grail' with the invention of universal computer memory

[An] electronic memory device -- described in research published in Scientific Reports -- promises to transform daily life with its ultra-low energy consumption. [...] The device is the realisation of the search for a "Universal Memory" which has preoccupied scientists and engineers for decades.

Physics Professor Manus Hayne of Lancaster University said: "Universal Memory, which has robustly stored data that is easily changed, is widely considered to be unfeasible, or even impossible, but this device demonstrates its contradictory properties."

A US patent has been awarded for the electronic memory device with another patent pending, while several companies have expressed an interest or are actively involved in the research.

The inventors of the device used quantum mechanics to solve the dilemma of choosing between stable, long-term data storage and low-energy writing and erasing. The device could replace the $100bn market for Dynamic Random Access Memory (DRAM), which is the 'working memory' of computers, as well as the long-term memory in flash drives.

[...] Professor Hayne said: "The ideal is to combine the advantages of both without their drawbacks, and this is what we have demonstrated. Our device has an intrinsic data storage time that is predicted to exceed the age of the Universe, yet it can record or delete data using 100 times less energy than DRAM."

Room-temperature Operation of Low-voltage, Non-volatile, Compound-semiconductor Memory Cells (open, DOI: 10.1038/s41598-019-45370-1) (DX)


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Interesting) by DannyB on Thursday June 20 2019, @09:46PM (5 children)

    by DannyB (5839) Subscriber Badge on Thursday June 20 2019, @09:46PM (#858239) Journal

    Existing OSes are organized around having separate durable storage and fast volatile memory.

    Having the best of both as a single form of storage would invite research into radical design changes to the OS.

    You would no longer have "boot", just merely "rest low power mode".

    Maybe you no longer have conventional "files". Still some concept of folders and "files", but a file may now be a data structure instead of an array of bytes. A word processor or spreadsheet "document" may be the actual in memory data structure. It may have a serializable form for transport across systems. But imagine your office suite program taking no time to open documents because documents don't need to be "parsed".

    This "file" of data structure is durable across restarts of the OS. Memory and "disk" are now one thing. The overall behavior of the system might seem more like a Palm Pilot once was. There is no "off" state, just resting. (it's not dead, it's resting)

    Next I start thinking about languages like Clojure with immutable persistent data structures in memory. No object can be modified, only a new modified object can be created. If I have an array of ten million elements, and I change a[5], then I get an entire new array of ten million elements with the fifth element changed. The old array is still there and holders of references to that old array still see the fifth element unchanged. Yet the performance expectations of changing a[5] are upheld -- eg, you don't experience the delay of copying all ten million elements to a new array. (implementation detail: everything is 32-way b-trees, not really an array underneath. Your new array simply gets a new leg of nodes up to the new root node. Your new array simply has a new root node into a new tree, but the old and new tree structures share the vast majority of their tree structure.)

    Now thinking about how those kinds of language features work now, and in such a new type of OS, it would make every possible Undo change of any document become cheap, and could always be part of the "file" containing the document. The user could decide to trim the undo history so that it is no longer to undo past 30 days ago, so that a lot of structure gets discarded. Such a thing is easy in memory now with languages like Clojure, but suppose that the memory is also the disk, eg, durable storage.

    --
    People today are educated enough to repeat what they are taught but not to question what they are taught.
    Starting Score:    1  point
    Moderation   +3  
       Interesting=3, Total=3
    Extra 'Interesting' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   5  
  • (Score: 0) by Anonymous Coward on Thursday June 20 2019, @11:28PM (1 child)

    by Anonymous Coward on Thursday June 20 2019, @11:28PM (#858295)

    Having the best of both as a single form of storage would invite research into radical design changes to the OS.
    Yeah we would have to segment our memory into two spaces. Storage and usage. We could do that today if we really wanted to. 64bits is a huge memory space. I have not looked too deeply but I suspect nvme does it.

    Now thinking about how those kinds of language features work now, and in such a new type of OS, it would make every possible Undo change of any document become cheap,
    Space limitations are still in play. Journal filesystems are fairly common. You can turn this feature on windows today if you want. You are thinking something like this would have unlimited space. We are not there yet.

    • (Score: 2) by DannyB on Friday June 21 2019, @03:35PM

      by DannyB (5839) Subscriber Badge on Friday June 21 2019, @03:35PM (#858578) Journal

      No, I'm not thinking it magically has unlimited space. I'm thinking that some things can be done more efficiently and differently.

      Maybe to state it a bit differently, I'm thinking that "files" are the in-memory representation of things as they are used by the applications that work on them once loaded into the application. There is no "on disk" format. Or more properly "serialized" format of the document -- except for transfer between machines, or for external out-of-the-live-memory backup.

      No journaling file system. Just the illusion of a file system. Folders and "files". But files are really documents within their running application, in some sense.

      This is just thinking out loud. It's not as though I have some fully baked rethinking of the OS. But I have thought about this before. What happens if memory and disk were one and the same thing.

      --
      People today are educated enough to repeat what they are taught but not to question what they are taught.
  • (Score: 0) by Anonymous Coward on Friday June 21 2019, @01:10AM

    by Anonymous Coward on Friday June 21 2019, @01:10AM (#858359)

    Just run the code in place. Makes a big difference.

  • (Score: 1, Interesting) by Anonymous Coward on Friday June 21 2019, @03:38AM (1 child)

    by Anonymous Coward on Friday June 21 2019, @03:38AM (#858416)

    Linux has supported a form of this for a very long time. XIP support was added in 2005:

    https://lwn.net/Articles/135472/ [lwn.net]

    Executables are executed directly from flash and not copied to RAM.

    • (Score: 2) by DannyB on Friday June 21 2019, @03:36PM

      by DannyB (5839) Subscriber Badge on Friday June 21 2019, @03:36PM (#858581) Journal

      That is part of what I'm thinking of. But also that "files" are the in-memory data structures as an application would use them after "loading" them from disk.

      --
      People today are educated enough to repeat what they are taught but not to question what they are taught.