back to article Intel and Micron's XPoint: Is it PCM? We think it is

Does 3D XPoint memory use phase change memory (PCM) technology or not? After IMTF co-chair Guy Blalock said XPoint uses a chalcogenide material, like Phase-Change Memory (PCM), your storage correspondent sent a mail to Intel and Micron saying: “According to my understanding phase-change memory is a non-volatile, chalcogenide …

  1. Mage Silver badge


    Sounds like a better duck?

    1. Charles 9 Silver badge

      Re: Duck

      How about a miniature goose with a duck call?

    2. Anonymous Coward
      Anonymous Coward

      Re: Duck

      Sounds like nebulous sales puff with subtle hints of whalesong and wheatgrass smoothie.

      1. Destroy All Monsters Silver badge

        Re: Duck

        Sounds like O.J. Simpson is not admitting to killing his wife.

  2. John Savard Silver badge

    Simple Enough

    It undergoes a bulk change, but not a bulk change in phase, but a bulk change in other properties. So it never turns from solid to liquid, say.

    1. Ian Johnston Silver badge

      Re: Simple Enough

      Many materials have more than one solid phase.

  3. Lee Mulcahy

    Unlike != Not?

    Although I agree that it sounds like phase change memory, I disagree with the author's contentions. The phrase "Unlike Phase change memory" strongly implies that it is NOT phase change memory.

    1. Anonymous Coward

      Re: Unlike != Not?




      Implies != States

      It would have been *VERY* easy for the Intel/Micron pact to *STATE* whether the stuff they're peddling involves phase changes. They DID NOT. Even in reply to the correspondent's direct, single-point follow-up question. INSTEAD they appear to have had marketing (or, in this case, it smells like LEGAL) splaff out a carefully contrived EVASION.

      So I feel we're left having to ask ourselves WHY the Intel/Micron corporate lawyers are CHOOSING to evade the simple question. What are they trying to hide, and WHY?

      My GUESS would be one of two explanations:

      • "XPoint" IS phase change memory but they're trying to avoid confirming that. Because their innovative archetecturizing impinges all over someone else's patents or somesuch?
      • "XPoint" is NOT PCM but the pact is deliberately trying to IMPLY that it is. Possibly because whatever it REALLY is impinges all over someone else's patents, or uses some other physics which is already public domain and the vendors are trying to "borrow" PCM patent protection for long enough to secure market domination before "generics" pop up?

      Unlike our scribe, I'm more inclined towards something resembling the latter, but whatever their reasons it's certainly piqued my interest and it would be nice to not have to guess what's really going on!

      Nice work Reg :)

      Maybe our trusty scribe could try a re-followup? Something like this perhaps:

      Yes, yes, of course your heroic product architecture engineers have heroically architected a brave new "cross point" product architecture proactively reimagined to leverage out of the box innovation, thusly realining the playingfield into a vertical trajectory to reverse the neuron flow into a whole new paradigm... But that's not what I'm asking about. What I'm asking is HOW IT WORKS. By what physical mechanism(s)/propert(y/ies) is the binary data recorded within the physical material?


    2. Anonymous Coward
      Anonymous Coward

      Re: Unlike != Not?

      > Implying

  4. FrogInABlender

    quod erat demonstrandum !

    1. Warm Braw Silver badge

      or we are dummies


  5. Alistair

    Any physicists around that care to comment

    on chalcogenide behaviours *other* than the phase change induced by electrical stimulation?

    I've an engineering text that might have the answer, but that text is buried amongst other such dead tree objects I've not touched in ............ Well, long enough that I'm not gonna admit it.

  6. Flocke Kroes Silver badge

    Sounds like ...

    ... the difference between PCM and XPoint is the same as the difference between software patent and computer implemented invention.

  7. ksb1972

    Trade secrets perhaps?

    Why share hints/tips/pointers with your competitors and let them catch up/overtake you quicker?

    Which multi $billion corporation wouldn't want to keep a sharp eye on the competition? Especially when it announces The Next Best Thing since sliced bread?

    1. Anonymous Coward
      Anonymous Coward

      Re: Trade secrets perhaps?

      That's what patents are for.

      ..unless you're suggesting what's really going on in there is something unpatentable?

      1. Anonymous Coward
        Anonymous Coward

        Re: Trade secrets perhaps?

        Elon Musk has allowed some of his patents to be licensed freely and doesn't patent other aspects of his technology because that would simply allow China to copy it.

  8. Geoff Johnson


    Sounds to me like someone has a patent on phase change memory, so they want to pretend they're not infringing.

  9. itzman

    Pulse Code Modulation?

    ??? Er what? YATLATLLATLA*

    * yat-lat-lattler Yet Another Three Letter Acronym That Looks Like Another Three Letter Acronym.

  10. Len Goddard

    False Dichotomy

    "Our conclusion is that 3D XPoint uses phase change memory. Quod ergo demonstrandum - or we are dummies?"

    It is quite possible - even likely - that 3D XPoint is phase change memory AND you are dummies.

    1. Anonymous Coward
      Thumb Up

      Re: False Dichotomy

      Indeed. Equally possible - even likely - that 3D XPoint is not phase change memory AND they are - in fact - not dummies.

    2. allthecoolshortnamesweretaken

      Re: False Dichotomy

      We won't know until we open the box.

  11. Numen


    I wonder if some of the chalcogenide magnetic properties are being used. Doesn't look like it from the diagram, though.

  12. luis river


    Where is 3d xpoint? All words, all speculations, all vaporware, industry has spent decades trying to create new type NVM, all fault, Hewlett Packard (memristor) has promised for one long time something similar and What do we have? Nothing

  13. P0l0nium

    Why speculate .....???

    Its CMOx. It relies on oxygen ion migration in oxide and Intel/Micron bought it from Unity Semiconductor.

  14. Joerg

    All this bashing against 3D XPoint... SSD competitors?

    All this bashing against 3D XPoint...are SSD competitors behind it ?

    They surely must be scared because the unreliable TLC,MLC,SLC technologies will become obsolete and useless as soon as 3D XPoint hits the market.

    Now many sites have started bashing the 3D XPoint claiming that Intel and Micron lied and it's all a facade and just another old technology and nothing new on the horizon...

    Who would benefit from such a nonsense bashing then?

    Clearly TLC, MLC, SLC manufacturers and so current SSD competitors indeed.

    So is 3D XPoint really going to be pure crap and all the claims by Intel and Micron just fake? I really don't think so. If that was ever the case all this bashing wouldn't have ever appeared on the 'net all of a sudden like this....

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2021