Sat, Dec 20, 2014, 2:51 PM EST - U.S. Markets closed

Recent

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

Intel Corporation Message Board

  • paul.ottelini paul.ottelini Jul 6, 2011 10:22 PM Flag

    Can Intel destroy ARM?

    In a word, YES!

    How?

    If Intel decides to use its trigate technology on nothing but x86 architecture. It's that simple.

    SortNewest  |  Oldest  |  Most Replied Expand all replies
    • gregory.lynn@rocketmail.com gregory.lynn Aug 29, 2011 2:01 PM Flag

      Hahaha, Yahoo is pretty lame.

    • What happened to the post you were replying to?

    • gregory.lynn@rocketmail.com gregory.lynn Aug 29, 2011 11:00 AM Flag

      Not an alias, real name, location, age, etc.

    • gregory.lynn@rocketmail.com gregory.lynn Aug 29, 2011 8:24 AM Flag

      Not an alias, my real name, age, and location. Thanks for playing.

    • This is your third post with the same info in it.

      You're worse that the investment spammers...

    • you are applying consumer thinking in the server space. That is why the theoritical possibility of ARM servers is good enough for a starbucks chat...

      Server selection has to do with reliability, cost of maintenance, etc. Server selection does not include criteria "mm let's try this:)" it's exactly the opposite.

      -----

      You are talking to the converted:) I agree with what you are saying.

      The ARM crowd know their "solution" is only suitable for certain work loads. The work loads that are currently being virtualized (ie, one's with low cpu utilization) and web servers that are sitting on the front edge of the cloud (server farms).

      We'll see more from ARM next year when the A15 appears in devices. Full virtualization, 40 bit addressing blah blah blah.

    • there are existing Atom processors running at 0.65W. You go figure what a server's total TDP would be right here, right now. And that's real-ware not thought-ware. But this is not the point. The point is that today one engineer who selects servers en masse can go and get twice as many "good enough" cores for the same dollar. And they don't do it. The reason behind ARM hypothetical failure (they dont even have a product to begin with now) in the server space is who does the selection of the product. The server crowd is not thinking the same way a consumer thinks at the mall. They don't care if the product looks cool or if everyone else at starbucks has it. Server selecting engineers do not step on dollars to save pennies.

      you are applying consumer thinking in the server space. That is why the theoritical possibility of ARM servers is good enough for a starbucks chat...

      Server selection has to do with reliability, cost of maintenance, etc. Server selection does not include criteria "mm let's try this:)" it's exactly the opposite.

      A server selcting enginneer will know exactly what performance he is getting for the dollar, will not accept offers to the tune of "look at iPad it does most of the things you need for $600 so you can save a $100 for the mall" while it has no ports, no keyboard, cannot be upgraded, will be obsolete next year etc etc etc ...

      Even if ARM comes with a product around 2015, it would be another 5-6 years before it has meaningful adoption. Now, in between, guess what INTC will be doing...

      To close my position on the server space here, there is a champion and a punching bag today. They could be two punching bags down the road :)

      The consumer fight though, is a whole different animal. With the M2M upcoming as well there are many aspects and many posibilities to consider. And ARM is the champion but the contender is getting stronger day by day.

    • 1. processor to processor core ARM has a small power advantage over the last year's 45nm Atom cores. The problem with INTC versus ARM is in the SOC power envelope not in the processing core. Simpy the INTC "SOC" sucks, at the moment. And this makes a big difference in mobile SOCs , I don't see the relevance in server chips. There are 4W atom server designs since 2009.

      2. AMDs current offerings have nothing wrong in either architecture or manufacturing. The Barca problem is history now. Current AMD offerings have very competitive if not better performance per wattage. And market share *still* erodes on a daily basis.

      3. ARM server chips capable of 64 bit processing are paper-ware at the moment.

      Let's combine now 1,2,3 ...

      "as they know they are the ones with the economies of scale and are lusting at the margins on Intel's server chips"

      ok, lets laugh now. Assume that ARM comes with a chip that has some demand. Pretty long shot but lets say(I would bet my farm on that mind you)

      why on earth the customers would tolerate anywhere the same price gouging they tolerate from intc? didnt you say yourself the ARM SOC margins in mobile are atrocious for INTC standards ?

      ARM can lust all day long. It's a free country. But even if it wins a design or two, it would be a freebie for INTC profit margin standards...
      -----

      Like I said, I don't think ARM will get far in servers. To answer some of your points:

      That 5W figure was the entire server, including all the support chips and memory. The 4W atom figure is just the cpu (20-60W for the full server). I used that as an example to show what some firms are doing in the ARM eco-system. Unless ARM address the floating point problem, all this is pointless.

      I disagree with some of the numbers when you compare SoC CPU power usage. The typical power difference in the 45/40nm range is about 4 times.

      You do bring up an important point in that the SoC design is becoming more important especially as the CPU isn't the largest single component any more. In this regard Intel are up against the design skills of Qualcomm and TI and not ARM.

      To be honest I haven't see performance/watt numbers for the latest AMD/Intel chips. I do know that in raw performance AMD haven't kept up.

      Next to nothing is known about the ARM 64 bit processors and at this stage it's not worth talking about.

      Customers wouldn't tolerate Intel levels of pricing. The idea that the upfront cost would be cheaper as well as cost of ownership (due to power and cooling costs).

      To repeat I think ARM will struggle in the server space. The 'fear' (if that is the right word!) for Intel is that these ARM servers cause an erosion of margin until the threat is killed off. We're talking years (2015) before this threat appears in any great force and that is a long, long time in this game.

    • He has been recognized as self-promoting on other sites as well. Here is one where they actually first thought he was a spam-bot. In a way I suppose he is simply a spam-person.

      http://www.semiaccurate.com/forums/archive/index.php/t-1811.html

    • 1. processor to processor core ARM has a small power advantage over the last year's 45nm Atom cores. The problem with INTC versus ARM is in the SOC power envelope not in the processing core. Simpy the INTC "SOC" sucks, at the moment. And this makes a big difference in mobile SOCs , I don't see the relevance in server chips. There are 4W atom server designs since 2009.

      2. AMDs current offerings have nothing wrong in either architecture or manufacturing. The Barca problem is history now. Current AMD offerings have very competitive if not better performance per wattage. And market share *still* erodes on a daily basis.

      3. ARM server chips capable of 64 bit processing are paper-ware at the moment.

      Let's combine now 1,2,3 ...

      "as they know they are the ones with the economies of scale and are lusting at the margins on Intel's server chips"

      ok, lets laugh now. Assume that ARM comes with a chip that has some demand. Pretty long shot but lets say(I would bet my farm on that mind you)

      why on earth the customers would tolerate anywhere the same price gouging they tolerate from intc? didnt you say yourself the ARM SOC margins in mobile are atrocious for INTC standards ?

      ARM can lust all day long. It's a free country. But even if it wins a design or two, it would be a freebie for INTC profit margin standards...

    • View More Messages
 
INTC
36.37-0.65(-1.76%)Dec 19 4:00 PMEST

Trending Tickers

i
Trending Tickers features significant U.S. stocks showing the most dramatic increase in user interest in Yahoo Finance in the previous hour over historic norms. The list is limited to those equities which trade at least 100,000 shares on an average day and have a market cap of more than $300 million.
3M Company
NYSEFri, Dec 19, 2014 4:04 PM EST
ImmunoGen, Inc.
NASDAQFri, Dec 19, 2014 4:00 PM EST
Southwestern Energy Co.
NYSEFri, Dec 19, 2014 4:05 PM EST