Recent

% | $
Quotes you view appear here for quick access.

ARM Holdings plc Message Board

  • intelphonechips intelphonechips Mar 26, 2012 2:49 PM Flag

    Intel roadmap spells trouble for ARMH

    http://www.slashgear.com/intel-roadmap-teases-14nm-mobile-chips-in-2014-27215924/

    Intel showed off its mobile chip roadmap today at MWC 2012, featuring an ambitious plan that Intel CEO Paul Otellini believes will “light a fire under” Apple, Samsung, and LG. For this year, the company has set 32nm architecture chips with its Medfield Atom Z2460 processors. (Make sure to check out our hands-on with the Intel Orange “Santa Clara” Medfield phone). But next year, it’s shrinking its chips down to a new 22nm process and then further to 14nm by 2014.

    At this pace, Intel is moving faster than Moore’s Law, according to which the number of transistors that can fit into an integrated circuit doubles about every two years. At the same time, the chips will either have the same performance but half the size or have the same size but double the performance. With 14nm, Intel is getting very close to reaching the theoretical limit of silicon, which is around 9nm to 11nm. However, the company has mentioned before that it expects to eventually jump to 8nm by 2015.

    Intel believes that the mobile market moves at a far faster pace than desktops and hence the company will be going be shrinking down the slicon aggressively. One major reason Intel will be able to go at such a rapid pace is its adoption of FinFETs next year, which will allow a single transistor to act as a multi-gate device. In this way, one transistor can perform much more calculations than in a traditional system and hence reduce the number of electrodes and consequently the size of chips. Furthermore, Otellini affirmed that Intel had plenty of manufacturing capacity for both PC and non-PC devices.

    SortNewest  |  Oldest  |  Most Replied Expand all replies
    • we need to replace the big wheels at intel for allowing a termite size company to outmaneuver them in the mobile venture. under their watch. now, it is playing catch-up. no excuse. heads should roll off.

    • I think you are going to be crushed when your ARM fanboy fantasy is ruined when real world benchmarks show how very wrong you have been all along. Javascript matters a lot in mobile performance, it powers the web.
      -----

      I doubt it:) Why?

      http://browse.geekbench.ca/geekbench2/search?q=Atom

      http://browse.geekbench.ca/geekbench2/search?q=ARM

      Javascript is a very important benchmark but it varies so much from browser to browser to OS to OS. Try it on your own machine and see.
      -----
      One more thing, Medfield is not the game changer, that comes next year with Valley View which will have 22nm Tri-gate trasnsitors. The whole world will see just how much of a difference Tri-gate makes in April 2012. This bloated pig of a stock will be under $20 by year end, maybe sooner.
      -----

      Arh, Tri-gate - yes, we'll see how much difference it makes in high performance processes...

    • gregory.lynn@rocketmail.com gregory.lynn Mar 31, 2012 7:40 AM Flag

      She'll get crushed by S4/A15 at 28/32nm when they arrive which maybe sooner than you think...
      _______________

      I think you are going to be crushed when your ARM fanboy fantasy is ruined when real world benchmarks show how very wrong you have been all along. Javascript matters a lot in mobile performance, it powers the web.

      One more thing, Medfield is not the game changer, that comes next year with Valley View which will have 22nm Tri-gate trasnsitors. The whole world will see just how much of a difference Tri-gate makes in April 2012. This bloated pig of a stock will be under $20 by year end, maybe sooner.

    • Early benchmarks showed 32nm Medfield better than 45/40nm ARM processors in performance as well as battery life. I would guess roughly 20% better performance per watt. More recent unofficial benchmarks indicate Medfield is overall on par with 32/28nm next-gen ARM processors with some positives and some negatives. And Medfield is market ready whereas comparable next-gen ARM processors are few months away and won't be in big volume this year.
      -----

      Er, not quite. Intel's benchmarks only highlighted javascript based tests...

      ...in other tests medfield was on par (slighly ahead) or beaten by EXISTING 40/45nm based phones.

      Medfield will get beaten by higher clocked A9@40/45nm which will be in the market by the time she arrives (actually, I'de put money on it that your phone will beat medfield in most performance benchmarks).

      She'll get crushed by S4/A15 at 28/32nm when they arrive which maybe sooner than you think...

    • As I recall, a Medfield based phone was found to have performance comparable to an iPhone 4s.

      • 1 Reply to mellors.john
      • Pegatron undertaking exclusive ODM production of 15-inch Acer ultrabook, say Taiwan makers

        Aaron Lee, Taipei; Adam Hwang, DIGITIMES [Wednesday 28 March 2012]




        Pegatron has undertaken exclusive ODM production of a 15-inch ultrabook for Acer since mid-March, with monthly shipments of 100,000 units, according to Taiwan-based supply chain makers.

        The ultrabook features a fiberglass case and is expected to be launched in April-May, the sources indicated. Viewing that monthly shipments range between 30,000 and 50,000 units for most ultrabooks, Acer's planned monthly shipment volume is relatively large, the sources noted.

        Pegatron aims to ship 18 million notebooks in 2012, and may upward adjust the target due to orders from Acer, the sources noted.

        While most first-generation ultrabooks launched since the fourth quarter of 2011 have up to 13-inch screen-sizes, 14- and 15-inch are expected to be mainstream sizes for second-generation models launched beginning the second quarter, the sources said.

        Global sales of ultrabooks are expected to significantly increase in the first wave in the second quarter of 2012 due to Intel's launch of Ivy Bridge processors, and in the second wave in the third or fourth quarter when Microsoft launches Windows 8, the sources indicated.

    • i hope so. this giant has been asleep. it is time to wake up. armh is way too far ahead. with intel's resources, there is no
      excuse for playin catch up to ant size armh.

      • 1 Reply to mercedlade
      • Ironically, playin catch up has rewarded Intel handsomely in the past, be it 64-bit ISA or servers markets (twice). Same will happen with smartphones/tablets markets.

        It is not that Intel was lucky in the past. If you have superior technology and enough resources but didn't focus on some markets, it is not difficult to catch up in those markets. And you save on market research. And you minimize the risk of failed products.

        Regarding "ARM is too far ahead"; Yes, it is in mobile and embedded market share, but not too far ahead in technology. It is a misperception.

    • At that rate Intel should garner 5% of mobile market IN 2015.

      Intel investors should also be aware of whatever mobile market share they gain, they will lose 2-3X as much in earnings.

      Look up phyrric victory for reference.

      • 1 Reply to lackeygarrett
      • "At that rate Intel should garner 5% of mobile market IN 2015."

        Markets often have dramatic turns over three years period. The present rate of change may accelerate or it may totally turn around. Look at what happened in past 3 years.

        "Intel investors should also be aware of whatever mobile market share they gain, they will lose 2-3X as much in earnings."

        This is not the first thing for Intel investors to worry about. Intel is going to lose 2-3X in earnings anyway. The best thing they can do is to grab 1X mobile market share instead of yielding it to rivals. That is what investors will be looking for. And regarding 2-3X loss, the market dynamics and expected overall growth will ensure that that will not happen.

 
ARMH
43.13+0.22(+0.51%)May 26 4:00 PMEDT