Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Tuesday September 19 2017, @05:42PM   Printer-friendly
from the it'll-cost-you-to-remember dept.

IC Insights has predicted that DRAM prices will continue to increase this year:

According to IC Insights, DRAM prices will continue to increase even though they have more than doubled (+111%) over the last 12 months. IC Insights predicts that by the end of the calendar year DRAM's price per bit will have jumped a record 40% (or more).

[...] Of course, the record pricing levels are great for our friends at the major foundries. Samsung, Micron, and Sk Hynix are also raking in their own record profits and enjoying healthy margins. We have both DRAM and NAND shortages occurring at the same time, which is great for the foundries, and unless a player breaks ranks to gain market share, we can expect more foot-dragging before any of the foundries increases output.

The booming mobile industry and server markets are exacerbating the issue, so you would expect that the fabs would boost DRAM output. Unfortunately, the three primary fabs (Micron bought Elpida, reducing the number of players) don't share the same vision.

IC Insights indicates that Micron will not increase production capacity, instead relying upon improvements in yields and shrinking down to smaller nodes to boost its DRAM bit output. Sk Hynix has expressed its desire to boost DRAM output but hasn't set a firm timeline for fab expansion (unlikely to occur in the near term). Samsung is as tight-lipped as usual, so we aren't sure of its intentions.

In the 1980s there were 23 major DRAM suppliers, but cutthroat pricing and continual oversupplies eventually led to the wave of consolidation that left us with the current three suppliers.

Previously:

December 2015: DDR4 Memory Prices Declined 40% in 6 Months

May 2017:
DRAM Price Surge Continues
Samsung Set to Outpace Intel in Semiconductor Revenues

July 2017:
Micron Temporarily Suspends Operation of DRAM Production Facility
Samsung Increases Production of 8 GB High Bandwidth Memory 2.0 Stacks

August 2017:
DRAM Prices Continue to Climb
Samsung & SK Hynix Graphics Memory Prices Increase Over 30% In August


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This discussion has been archived.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: 1, Interesting) by Anonymous Coward on Tuesday September 19 2017, @06:41PM (6 children)

    by Anonymous Coward on Tuesday September 19 2017, @06:41PM (#570297)

    This is why competition is so important; it's an essential element of evolution by variation (supplier competition) and selection (consumer choice).

    However, such competition can be a lot more complex than the obvious: Maybe now, programmers will have an incentive once again to make their programs less memory intensive; improvements in software can make negate the undue gains being sought by these colluding manufacturers. Decentralized solutions FTW!

    Starting Score:    0  points
    Moderation   +1  
       Interesting=1, Total=1
    Extra 'Interesting' Modifier   0  

    Total Score:   1  
  • (Score: 0) by Anonymous Coward on Tuesday September 19 2017, @06:55PM

    by Anonymous Coward on Tuesday September 19 2017, @06:55PM (#570304)

    Since most programmers can't actually program, I'm not sure they even know how to make their programs less memory intensive. Why not just use up every last resource available, after all? Who on Earth would want to run countless programs simultaneously!?

  • (Score: 0) by Anonymous Coward on Tuesday September 19 2017, @07:53PM (2 children)

    by Anonymous Coward on Tuesday September 19 2017, @07:53PM (#570324)

    Oligopolies almost also suck. Always have, always will. Telecoms, the big-3 in Detroit before Japanese competition, Wintel, etc. are examples of self-protecting crap. They use "economies of scale" as an excuse, but the value of competition is usually greater than the benefit of economies of scale.

    • (Score: 3, Interesting) by Gault.Drakkor on Tuesday September 19 2017, @09:13PM (1 child)

      by Gault.Drakkor (1079) on Tuesday September 19 2017, @09:13PM (#570363)

      They use "economies of scale" as an excuse, but the value of competition is usually greater than the benefit of economies of scale.

      Economies of scale do help the corporation as an individual corporation. Competition hurts the individual corporation in the short term. So from corporations point of view its a win-win to have fewer players.

      Whether or not they share those benefits of economies of scale is a different issue. As you suggest, sharing is strongly influenced by how much competition there is.

      My problem with what you said is that you seem to be merging it to being one point of view, that oligopolies, and monopolies suck. Which is false, where very clearly some parties benefit.

      I do agree with you in that once again shows; if there are large barriers to entry, which chip fabs represent, capitalism doesn't work so well. So this implies there should be market regulation.

      • (Score: 0) by Anonymous Coward on Wednesday September 20 2017, @12:14AM

        by Anonymous Coward on Wednesday September 20 2017, @12:14AM (#570445)

        I meant from a consumers' standpoint. Tim Cook and Mark Zuckerberg obviously loves oligopolies.

  • (Score: 2) by richtopia on Tuesday September 19 2017, @08:40PM (1 child)

    by richtopia (3160) on Tuesday September 19 2017, @08:40PM (#570348) Homepage Journal

    DRAM is an example where barriers to entry slow market corrections. It takes A LOT of money to build a modern fab, and a long time. And that assumes it is one of the major players who already know how to make memory; if a new player is entering the industry then you have to learn how to yield too.

    Ultimately, memory is still affordable, so unfortunately I don't think this will result in programmers being more memory conscious. The one exception would be a mobile operating system: if Google prioritizes memory management in future builds of Android it would help the lower end devices. However, I'm skeptical that will happen, as features get promotions and sell phones, not better memory management.