• @the_q
    link
    English
    7
    edit-2
    1 year ago

    deleted by creator

    • @[email protected]
      link
      fedilink
      English
      -3
      edit-2
      1 year ago

      They probably need the capacity for more lucrative products. I seriously doubt they will not be manufacturing something. Supply and demand sets the price of course so increasing prices are a result of reducing supply, of course.

      • @the_q
        link
        English
        2
        edit-2
        1 year ago

        deleted by creator

        • @[email protected]
          link
          fedilink
          English
          -21 year ago

          There is only so much capacity to produce these things, and it’s shared with other products like CPUs and other chipsets. I don’t know for sure but probably they either could not pay asking price to produce in same quantities as before (due to increased demand of the fabrication facilities) or Samsung is choosing to product more of another product that shares the same fab.

  • AutoTL;DRB
    link
    fedilink
    English
    21 year ago

    This is the best summary I could come up with:


    According to a recent report via TrendForce, the NAND flash market will soon break away from its trend of decreasing prices.

    Samsung’s solution is to induce a supply-side shock by cutting NAND output by as much as 50% (mainly on processes under 128 layers).

    When supply is artificially capped, the higher demand will slowly push stocks lower, increasing the sales price.

    That’s the mechanism Samsung is leveraging to stem the bleeding prices of NAND, which has seen a severe decline in the last year.

    In its current state, AI workloads require somewhat disproportionate compute as GPU and CPU power is much more critical to performance than having a quick, hot-storage solution.

    The gold rush for Nvidia GPUs (and their cost, which has led Nvidia to lock-in some 823% in profit percentage) also brings opportunity cost with it: spending money on GPUs leaves less money available for more “classical” supercomputing systems.


    The original article contains 418 words, the summary contains 150 words. Saved 64%. I’m a bot and I’m open source!