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

Intel Corporation Message Board

  • maximum_probability maximum_probability Feb 7, 2013 8:35 PM Flag

    How hard is it for intel to start another architecture?

    I am trying to understand the possibilities available to intel here given its vast resources and cash availability vs some small opponent like ARM.

    How hard is it for intel to keep improving its own core architecture but at the same time research the possibility of designing a completely new architecture similar to eg risc of ARM or at least in a direction other than x86 that will strike aRM at its own area and eliminate the need for people to licence from ARM and instead see in intel a new threat alternative at all levels both the mobile and the PC /server.

    Intel must make sure it recognizes that ARM is not the problem. The real problem is allowing some other companies like say QCOM to grow enough earnings that they then become able to attack intel in other areas. To do that it needs to challenge those companies directly where it hurts by outdoing the ARM+others architecture game with its own alternative that can be superior.

    Is that direction considered? Why do we need to worry about the mobile transition or adapting x86 to it and not try to not only defeat them at their own game since we can fund research there to outperform their designs but also to move the market into directions of new applications of PCs that will require super massive computing power that mobile devices will never be able to provide forcing the server and cloud markets to rely insanely more on the traditional intel architecture. You need to corner the market in many levels. I would imagine intel with the economy of scale can launch such multiple wars and win them all if it wanted. Who can finance such efforts better than intel? However if this is not done pretty soon this will be indeed some other company not ARM etc but the big players that are now relying on ARM and growing their earnings on their designs...

    In fact if intel wants to completely kill everyone they need to consider a particular architecture and hardware direction that will prove very expensive and complicated and yet very efficient for nobody else to be able to chase them there and the applications to depend on it. If you develop technology that takes dozens of billion to build factories for, nobody will be able to chase them there but they will be able to chase them there and better if intel doesnt take the market in that direction very soon.

    Sentiment: Strong Buy

    SortNewest  |  Oldest  |  Most Replied Expand all replies
    • There is no problem with the x86 architecture, per se. There is (was?) a problem with the x86 implementations. Intel developed the microprocessor lines from the start, when they were too slow and user productivity continued to expand as Intel made the CPU faster and faster.

      Intel recognized when they reached the point where the CPU was fast enough but the "killer" app was not to be found. They had reached that point several times before but increased memory size and software requirements had pushed that performance level out. Intel pushed in the direction of low power and portability but Apple took a huge leap that Intel could not bring itself to take.

      The x86 implementation will be dramatically changed with Haswell. Haswell was taped out twice. The second time was to drop the power envelop by 50%. The fruit of that design will be seen in 2013 and beyond. Power reduction circuitry, FMA floating point improvements and especially TSX transaction memory (for threaded code to manage locks).

      The battery life benchmarks will be the club used to beat the analysts over their heads.

    • Perhaps you should pose a different question: how hard is it for you to pick a different stock? Apparently VERY difficult as you are wedded to failure.

      • 2 Replies to longtalltxn
      • Your wife knows about being wedded to a failure, living in Texas.

      • Thats kind a bs cheap gutless answer buddy. I can and do make money with intel even in flat or down markets by selling puts and buying leaps at -2sd and then selling the leaps at +2sd covering puts or letting expire or getting exercised and selling calls until taken away indefinitely eg sold leaps before earnings and then rebuying them 50% cheaper like say near here that we are now and so on.

        I would of course make more if the stock started a move to 40 or 100 that i am very confident eventually will happen when all the bs clears. But until then i am ok with making 5%-10% per trade cycle here and there rather than having up 30-40% up months which will be the case if an uptrend begins.

        The problem with finding new stocks is that their derivatives are insanely priced and have massive spreads and they are mostly overvalued bubbles ready to crash. The only other stock i like long here is apple at these PE levels. And of course i would enter only in severe down days and only by shorting puts on it as well.

        I am not going to switch the core of the trading portfolio out of intel the most undervalued company in the universe and into some 2x the PE type stock like TXN or ARMH or QCOM lol . I am patient enough to wait until intel and the idiot analysts get it right at the same time.

        The harder you train to make money the easier the money will come when the bs ends and the uptrend starts. A tough stock that you know is better than some casino garbage. with uptrend about to crash.

        If you are so happy to trash intel and not answer my question then please offer the other stocks that you like better.

        Sentiment: Strong Buy

37.44+0.26(+0.70%)Sep 28 4:00 PMEDT