ARM: Intel a 'serious competitor', we're still better

[Bateluer]

http://www.xbitlabs.com/news/cpu/di...Enough_for_Smartphones_but_ARM_Is_Better.html

Chief executive officer of ARM Holdings, a leading developer of mobile microprocessor technologies, said that Intel Corp.’s latest Medfield system-on-chip is “good enough” for smartphones and the company considers Intel a “serious competitor”. But ARM claims that Intel will never be a leader of the mobile chip market since power consumption of its chips is just too high.

Their statements read like an adult talking to an upstart child. 😛 Let the pissing contest begin. 😀

As a tech enthusiast, I volunteer my time to personally test the latest devices from each. For science.

 

[theeedude]

Medfield is not going to make much progress in phones, most of its sales are going to come from Windows 8 tablets cannibalizing Atom netbooks and low end "Ultra"books. I guess cannibalization is better than having more of those sales eaten into by ARM powered devices. The phone partners I see are weak players who are glad to have Intel do the design and spend marketing dollars pushing it.

 

[quest55720]

The cannibalization is already happing because of the Ipad and Android tablets. If the second gen medfield later this year can run win 8 86 well intel could make a killing. Right now ARM tablets have a huge advantage with battery, weight, thickness and touch interface. If win 8 86 tablets are compareable in those areas arm tablets are in serious trouble. Be no reason to buy an android tablet if I can get the same thing with a win 8 86 tablet and be able to run win 86 programs.

 

[theeedude]

The cannibalization is already happing because of the Ipad and Android tablets. If the second gen medfield later this year can run win 8 86 well intel could make a killing. Right now ARM tablets have a huge advantage with battery, weight, thickness and touch interface. If win 8 86 tablets are compareable in those areas arm tablets are in serious trouble. Be no reason to buy an android tablet if I can get the same thing with a win 8 86 tablet and be able to run win 86 programs.

You want the ARM programs, those are going to be designed and optimized for touch and mobile computing. x86 programs are going to be bloated and designed for workstations with backwards compatibility back to windows 95.

 

[Bateluer]

You want the ARM programs, those are going to be designed and optimized for touch and mobile computing. x86 programs are going to be bloated and designed for workstations with backwards compatibility back to windows 95.

Would MS's tight control of Metro apps prevent/mitigate that?

 

[pcslookout]

You want the ARM programs, those are going to be designed and optimized for touch and mobile computing. x86 programs are going to be bloated and designed for workstations with backwards compatibility back to windows 95.

Are you so sure about that?

 

[theeedude]

Would MS's tight control of Metro apps prevent/mitigate that?

Maybe, but then it would undo the "advantage" of being able to run legacy desktop x86 apps. With ARM you are guaranteed to get a clean sheet design not burdened with ensuring compatibility with Wintel programs and drivers going back years.

 

[zsdersw]

AMD was overconfident, too... and look where that ultimately got them.

 

[Oyeve]

You want the ARM programs, those are going to be designed and optimized for touch and mobile computing. x86 programs are going to be bloated and designed for workstations with backwards compatibility back to windows 95.

I find that highly unlikely. I am confident win8 apparently will be optimized appropriately.

 

[poofyhairguy]

I am confident win8 apparently will be optimized appropriately.

Sure, Windows 8 will have that Metro tablet mode, but Windows isn't the problem. The problem is the Windows' ecosystem.

Is Office touch optimized? Not any version I have already paid for.

Is Photoshop touch optimized? Quickbooks? Heck, is Firefox touch optimized?

Nope, not one bit on x86. And it is not like just loading Windows 8 will magically fix the problem.

If MS was smart they would start a new certification program (with a sticker or whatever for boxes/websites) for apps that are "Windows 8 Optimized." Any app that gets the certification goes into a decent touch mode when the Metro UI is turned on.

Make it slick and automatic- when docked the tablet's apps go into normal mode. When undocked the "Optimized" apps go into touch mode. Make it so that only "Optimized" apps can be seen by the Metro UI and the problem will solve itself eventually.

Since I have not seen such a push from MS, until I see such a thing I am going to continue to assume that they will fail in the tablet market. Apple would have had something like that worked out day 1.

 

[ViRGE]

AMD was overconfident, too... and look where that ultimately got them.

Agreed; never, ever underestimate Intel. Always assume they have their best people working on something and are targeting you first. In spite of Atom's poor showing in smartphones so far this has more to do with Intel's desires to maintain their 65% margins than it does their capabilities. So should Intel decide to sacrifice short-term margins for long-term gains, they have a ton of resources to throw at the problem.

ARM should always be afraid of being on the receiving end of Intel's signature move. Hit a competitor with a process technology advantage (HKMG, etc), and then follow it up with a new architecture for a knockout blow. Specifically in ARM's case this means they need to be very careful of what the combination of tri-gate transistors and an OoO Atom could bring.

 

[alent1234]

Agreed; never, ever underestimate Intel. Always assume they have their best people working on something and are targeting you first. In spite of Atom's poor showing in smartphones so far this has more to do with Intel's desires to maintain their 65% margins than it does their capabilities. So should Intel decide to sacrifice short-term margins for long-term gains, they have a ton of resources to throw at the problem.

ARM should always be afraid of being on the receiving end of Intel's signature move. Hit a competitor with a process technology advantage (HKMG, etc), and then follow it up with a new architecture for a knockout blow. Specifically in ARM's case this means they need to be very careful of what the combination of tri-gate transistors and an OoO Atom could bring.

so far Intel has shown zero inclination in doing any of these

AMD never had more than a tiny part of the market. ARM is shipping hundreds of millions of CPU's a year by their licensees and there are now huge software catalogs for ARM software.

3-5 years from now will see ARM based laptops running iOS and Android. These will be good enough for 90% of computer users as a primary computer, they will have huge software catalogs and Intel will be in real trouble still clinging to their margins and Xeons

intel's problem aren't engineers, it's management

 

[sontin]

The fun part is that there is no ARM on the market. There are hundreds of companies which use their IP and designs to ship SoCs.

Look at nVidia and Tegra 3. They have 3 different ARM processors:
ARM(v)7, one Low-Power A9, four High-Performance A9 cores. Everyone is independent. Qualcomm is designing their own Processors based on ARM's IP.

So and one year later Intel will bring one Atom-Core to the market. They are so much behind the ARM partners that they need a wonder to catch up with them.

 

[JeffreyLebowski]

so far Intel has shown zero inclination in doing any of these

AMD never had more than a tiny part of the market. ARM is shipping hundreds of millions of CPU's a year by their licensees and there are now huge software catalogs for ARM software.

3-5 years from now will see ARM based laptops running iOS and Android. These will be good enough for 90% of computer users as a primary computer, they will have huge software catalogs and Intel will be in real trouble still clinging to their margins and Xeons

intel's problem aren't engineers, it's management

That's some ignorant comments right there. iOS and Android systems suck for content creation. Do you really think Apple is going to push for everyone to use $600-$800 tablets, when people happily spend $2000 on a MacBook Pro AND also buy an iPad, and an iPhone?

The problem with "ARM" based systems, ie iOS or Android, is there doesn't seem to be a defacto Gold Standard for apps. Word Processing? Spreadsheet? ect, M$ still rules the roost. And when you get into the corporate world, mobile apps just don't do enough. Wintel currently has more, refined, usable apps. Mobile apps might begin to outnumber desktop apps, but the quality and sheer power of those apps won't compare to desktop versions.
We deal with Excel spreadsheets with over 100k rows in them. A mobile device would choke on that.

 

[alent1234]

People like my wife, mom and others I know don't care. Internet, email, basic photo editing, hard drive to hold photos and movies is enough

 

[poofyhairguy]

We deal with Excel spreadsheets with over 100k rows in them. A mobile device would choke on that.

So would an Atom netbook that can technically run the software.

The issue isn't ARM inherently, it is that the mobile device world is about seven years behind the desktop computer world when it comes to raw power.

The benefit for ARM is that a year ago it was ten years behind and next year it will be five years behind. At the rate ARM is moving within five years we should have mobile devices that can take any spreadsheet you want- software compatibility will be the bigger issue and not raw power.

 

[alent1234]

My work core i5 laptop rarely sees more than 10% CPU. For most people a dumbed down version of office is more than enough. Most of the windows bloat is running services that a lot of users never use.

If someone can make an arm pc with just enough software for a large percentage of users but at a lower cost, watch out intel. Microsoft seems to be a little ahead of intel in this but not by much

 

[MrX8503]

I think Medfield is pretty impressive if the benchmarks are to be true. It beats ARM in power and efficiency. The only thing Intel needs now is x86 support. I think Intel knows very well that efficiency is the future, which is why they're pushing so hard with trigate and ultrabooks.

so far Intel has shown zero inclination in doing any of these

AMD never had more than a tiny part of the market. ARM is shipping hundreds of millions of CPU's a year by their licensees and there are now huge software catalogs for ARM software.

3-5 years from now will see ARM based laptops running iOS and Android. These will be good enough for 90% of computer users as a primary computer, they will have huge software catalogs and Intel will be in real trouble still clinging to their margins and Xeons

intel's problem aren't engineers, it's management

You go on to say there's no inclination of Intel doing any of those things....yet you go on to say that iOS/Android laptops are a sure thing.

LOL

 

[alent1234]

Acer, Asus and other computer makers barely break even on their low end computers and apple rules the high end. $500 laptop a good estimate is something like $200 goes to Intel/MS. and most people who buy these barely use the power of the CPU's in there.

If PC makers could make a laptop at the $500 price level but with higher margins they will do it. especially now that Android has a big software catalog. Google/Samsung are already experimenting with Chrome laptops.

With Apple they will soon run out of people willing to shell out $2000 for a laptop. A $500 Macbook where they can still get 30% margins will happen

 

[quest55720]

Acer, Asus and other computer makers barely break even on their low end computers and apple rules the high end. $500 laptop a good estimate is something like $200 goes to Intel/MS. and most people who buy these barely use the power of the CPU's in there.

If PC makers could make a laptop at the $500 price level but with higher margins they will do it. especially now that Android has a big software catalog. Google/Samsung are already experimenting with Chrome laptops.

With Apple they will soon run out of people willing to shell out $2000 for a laptop. A $500 Macbook where they can still get 30% margins will happen

You don't think the same thing is happing in the tablet world? Apple runs this high end for the most part. Content sellers like amazon and B&N ruling the low end by selling tablets near cost to sell content. Wait till amazon gets that 10 inch tablet out you can kiss the rest of the android tablet makers good bye expect maybe 1 or 2 to sell to the high end niche market.

 

[zsdersw]

You don't think the same thing is happing in the tablet world? Apple runs this high end for the most part. Content sellers like amazon and B&N ruling the low end by selling tablets near cost to sell content. Wait till amazon gets that 10 inch tablet out you can kiss the rest of the android tablet makers good bye expect maybe 1 or 2 to sell to the high end niche market.

Content does not flow from only Amazon or B&N.

 

[Bateluer]

You don't think the same thing is happing in the tablet world? Apple runs this high end for the most part. Content sellers like amazon and B&N ruling the low end by selling tablets near cost to sell content. Wait till amazon gets that 10 inch tablet out you can kiss the rest of the android tablet makers good bye expect maybe 1 or 2 to sell to the high end niche market.

Content does not flow from only Amazon or B&N.

B&N and Amazon both offer Nook/Kindle apps for most platforms as well. You can get your Amazon/Nook content on any Android tablet, Android phone, or iOS device. Plus your desktop and laptops as well.

 

[alent1234]

Content does not flow from only Amazon or B&N.

the kindle and nook are the best selling android tablets

 

[zsdersw]

the kindle and nook are the best selling android tablets

A pound of butter weighs a pound.

Now that we've dispensed with the irrelevant facts...

They're the best-selling because of the low price point, not because of the exclusive content.

 

[alent1234]

lots of people i see on the train have kindles because reading is the only feature they care about. internet and email are not that big a deal to them on a tablet and some casual games are good as well. $79 to $249 are the right price points for the features they want.

they don't care if their tablet doesn't have the latest GPU to run infinity blade or dead space.

if it wasn't for the content there would be nothing to do with them and this market is not going to spend $400 to $500 on a 10" tablet. instead of pushing specs most android tab makers should have made lists of what people want to do, how much they want to spend and then make a product