The ARM inherent efficiency myth

[krumme]

Yes, but even in this useage scenario using MS Office as an example, x86 compatibility is not required, because MS Excel/Word/etc...

Sorry - forgot that. I have to remember to myself Atom is in phones also, in a way that makes AMD server share look like a giant cash cow.

 

[ams23]

No, it actually doesn't. I only proves that, right now, the advantages of x86 are viewed as less favorable than the advantages that other products offer.

If the only difference was x86 or not, I guarantee you a large percentage of customers would go with x86. It's been that way for decades.

The vast majority of consumers do not even know what x86 is. Most consumers use tablets for watching video and movies, web browsing, social networking, ebook reading, emailing, apps, and casual gaming. x86 compatibility is completely irrelevant. The situation is different with desktop and laptop computers with disc drives, where the OS and software built for the OS has historically supported only x86 processors for many years.

 

[simboss]

It's sort of like saying Porsche isn't capable of making cars that are fuel-efficient, because none of their cars get great mileage. But I have every confidence that they could design an incredibly fuel-efficient car if that was their priority. It isn't.

Why?

What makes you think that someone that can design an extremely powerful and power hungry car/cpu can design equally well an extremely cheap and power efficient car/cpu?

History proves that this is not the case, neither in CPU or cars, the few attempts from high end manufacturers to get into the low end have mostly failed as have the few attempts made by the mainstream one to get into the luxury market.
This does not mean that Porsche or Cray will disappear, but in the end most of us only dream of those cars/cpus.

 

[krumme]

I can see a potential for x86 phones and tablets for the business market. But in another sense that we think.

Intel have the solutions for it, or can build it in time. And its nothing about the efficiency of the CPU or performance or something like that of minor, if any, importance.

The main point for Intel must be to reduce TCO per employee, and be the main brand in the B2B market. Reduce TCO for all IT involved. And TCO with measures of more hidden time, in using the IT, especially on the client side.

Sharing technical and software platform for all products; servers, desktops, notebooks, tablets, phones, (controllers...) have the potential to reduce central IT administrative cost. But far more cost can be reduced on client side; reduced learning curve, huge transparency for the user.

Intel can add its security IP, cloud expertice, small and midrange server solutions and so on. They have to build on their entire product and technology portfolio here. They have somthing unique here. And its not their dear fabs.

They have to stop thinking like the past - as separated product - and use Atom in Phones as a levarage to get their entire portfolio work together. Atom could be less efficient, its about applying it in another context.

In that way they can also use their selling and marketing machine, and fx. hide cost for phones in whatever - fx. license - solutions they can provide.

The strategic advantage of such a move, is that it secures their server share side even more because the solutions is tied together.

That market is far more valuable and fit for Intel. It doesnt have the organization to fight on the consumer market, its engi thinking is even having a hard time serving and innovating in the B2B market. And selling phones and tabs for kids and fun, is the most dead end ever.

 

[Homeles]

x86 compatibility is completely irrelevant.

It's not irrelevant. It is a rather major feature of the Surface Pro.

 

[TuxDave]

Why?

What makes you think that someone that can design an extremely powerful and power hungry car/cpu can design equally well an extremely cheap and power efficient car/cpu?

History proves that this is not the case, neither in CPU or cars, the few attempts from high end manufacturers to get into the low end have mostly failed as have the few attempts made by the mainstream one to get into the luxury market.
This does not mean that Porsche or Cray will disappear, but in the end most of us only dream of those cars/cpus.

I'm just speculating, but I assume there's some knowledge/talent/manpower overlap between Lexus/Toyota and Acura/Honda.

 

[Charles Kozierok]

The vast majority of consumers do not even know what x86 is.

The vast majority of consumers understand the meaning of: "Oh, I can run MS Word and Outlook and all my other PC applications on my tablet just like on my PC? Cool! Maybe I really don't need a PC then, which will save me a lot of money and hassle."

 

[ams23]

It's not irrelevant. It is a rather major feature of the Surface Pro.

The Surface Pro is what I would categorize as a convertible. It can function and work just fine as a tablet, but with the optional keyboard attachment it is very close to being a true replacement for a laptop computer (with the exception being that Surface Pro does not come with a built-in DVD/CD disc drive, so one would need to add an external disc drive to install most older software). So in the case of the Surface Pro (and other similar convertible products), x86 compatibility is relevant for those who want a worthy replacement for their x86 laptop. For the majority of consumers today however, x86 compatibility does not appear to be relevant based on consumer purchasing trends and consumer usage patterns.

 

[beginner99]

The success of iPad, Kindle Fire, Nexus 7, and Galaxy tablets compared to all Windows 8 tablets is proof that most consumers do not care about x86 compatibility in a tablet.

Or that Windows 8 just plain sucks.

 

[ams23]

Or that Windows 8 just plain sucks.

Looking back, I think that Microsoft miscalculated in how they positioned Windows RT and Windows 8, and in the effect that MS Office RT would have on Windows RT tablet sales. In order to encourage the growth of an ARM-based ecosystem, Microsoft should have done the following:

1) Create Windows Consumer OS aimed at ARM-based smartphones and tablets, with no desktop mode, with less software bloatware to fill up the hard drive space, and with MS Office RT Home & Student Edition not pre-installed in order to keep cost as low as possible for end user devices.

2) Create Windows Professional OS aimed at x86-based convertibles, laptops and desktops, with full desktop mode, and with MS Office Home & Student Edition pre-installed in order to add productivity for more premium and higher cost end user devices.

Rather than doing this, Microsoft decided to name their consumer OS "Windows RT" (where people don't even know what RT stands for), they decided to bundle MS Office RT Home & Student Edition with Windows RT (which adds cost for end user devices) without bundling it with Windows 8, they used ARM-based hardware that was no longer cutting-edge, and they ended up with a scenario where consumers were given mixed messages when ARM-based Windows RT (Tegra 3) tablets were selling for $499 USD at the same time that x86-based Windows 8 (Clovertrail) tablets were selling for $499 USD (or even less).

 

[Homeles]

For the majority of consumers today however, x86 compatibility does not appear to be relevant based on consumer purchasing trends and consumer usage patterns.

There's been x86 tablets out for, what, a month now? You've got some real hard evidence there.

Or that Windows 8 just plain sucks.

You're really good at broadcasting for the hardware enthusiast hive mind.

 

[ShintaiDK]

Windows 8 will be replaced in August.

One of the obvious failures of Microsoft Surface, specially the gimped RT, is the margins. Microsoft simply wanted the same margins that Apple had. But those are to dissapear and not return.

And for OEMs, there is the cost issue. Microsoft is simply too greedy on its software cost. And Microsoft havent been able to make money from anything outside its old PC OS/Applications core. Microsoft is also trying to force everyone into the rental agreement now.

 

[ams23]

There's been x86 tablets out for, what, a month now? You've got some real hard evidence there.

Put yourself in the shoes of most average consumers. Of all the things they are looking for in a tablet (including price, screen size, screen resolution, weight, aesthetics, battery life, reliability, responsiveness, web browsing speed, hard drive space, application and gaming performance, etc.), x86 compatibility surely must be at or near the bottom of the list (assuming these consumers know what x86 is in the first place, which is unlikely). I am not saying that an x86 tablet cannot be a great and popular device, I am just saying that x86 compatibility is not a concern for most people in my opinion.

 

[ams23]

Windows 8 will be replaced in August.

One of the obvious failures of Microsoft Surface, specially the gimped RT, is the margins. Microsoft simply wanted the same margins that Apple had. But those are to dissapear and not return.

And for OEMs, there is the cost issue. Microsoft is simply too greedy on its software cost. And Microsoft havent been able to make money from anything outside its old PC OS/Applications core. Microsoft is also trying to force everyone into the rental agreement now.

Microsoft still has an opportunity to right the ship to some extent, but the only way I can see that happening is to clearly distinguish between a lower cost consumer-focused ARM-compatible OS (which would include ARM processors and Intel processors using emulation, which would not include any software bloat to add to cost or complexity, and which would target both smartphone and tablet products) and a higher cost professional-focused x86 compatible OS (which would include Intel and AMD processors, which would include at least some variant of MS Office, and which would target convertibles, laptops, and desktops). Microsoft desperately needs to take advantage of both ARM SoC's and x86 SoC's, and with their current path, there is very little room for a Windows ARM ecosystem to grow.

 

[ShintaiDK]

Microsoft shoot themselves in the foot with Windows RT. Its their biggest flop of the lot currently.

 

[Homeles]

Put yourself in the shoes of most average consumers. Of all the things they are looking for in a tablet (including price, screen size, screen resolution, weight, aesthetics, battery life, reliability, responsiveness, web browsing speed, hard drive space, application and gaming performance, etc.), x86 compatibility surely must be at or near the bottom of the list (assuming these consumers know what x86 is in the first place, which is unlikely). I am not saying that an x86 tablet cannot be a great and popular device, I am just saying that x86 compatibility is not a concern for most people in my opinion.

Right now there are more important factors, but once Intel (and perhaps AMD) reach parity with ARM tablets, I'd imagine x86 would tip the edge. I don't disagree that that it's not at the top of everyone's list, but I can assure you that it is most certainly on the list.

Really, this topic has not yet reached maturity for discussion. We're only just beginning to see x86 in the tablet and smartphone world, and it's too early to be jumping to conclusions either way. When there are truly competitive x86 processors for tablets and smartphones, then we will be able to see how compatibility plays a role.

 

[ams23]

Microsoft shoot themselves in the foot with Windows RT. Its their biggest flop of the lot currently.

The idea of creating a Windows operating system that is compatible with ARM processors is a good one, as ARM mobile processors are the fastest growing CPU processors in history. The problem for Microsoft was in the way that they implemented this strategy.

 

[ShintaiDK]

The idea of creating a Windows operating system that is compatible with ARM processors is a good one, as ARM mobile processors are the fastest growing CPU processors in history. The problem for Microsoft was in the way that they implemented this strategy.

ARM segment is not growing. It went backwards the same percentage as x86 in 2012. There was simply canibalism inside the ARM segment.

And to fracture a ecosystem so you got Windows that is incompatible with Windows was imbecile. Specially when Mcirosoft did a double release. One for each uarch.

 

[ams23]

Right now there are more important factors, but once Intel (and perhaps AMD) reach parity with ARM tablets, I'd imagine x86 would tip the edge. I don't disagree that that it's not at the top of everyone's list, but I can assure you that it is most certainly on the list.

Really, this topic has not yet reached maturity for discussion. We're only just beginning to see x86 in the tablet and smartphone world, and it's too early to be jumping to conclusions either way. When there are truly competitive x86 processors for tablets and smartphones, then we will be able to see how compatibility plays a role.

ARM is still by far the most prevalent CPU processor type in mobile handheld [smartphone and tablet] devices, and that will not change anytime soon. Moving forward, as ARM-based devices become more and more powerful, software developers will increasingly develop PC software applications for ARM processors. Consumers are relying less and less on legacy PC software. Software is increasingly delivered through the cloud, and modern day tablets do not even include a built-in DVD/CD disc drive. Windows, Android, and iOS app stores all feature software that is compatible with ARM processors. The general theme and trend is that, as time goes on, x86 compatibility becomes less important.

 

[Olikan]

Windows 8 will be replaced in August.

WHAT?
wow....M$ is really in trouble

 

[ShintaiDK]

WHAT?
wow....M$ is really in trouble

Windows "Blue". Might be named R2. Windows Server 2012 will also be replaced. Kernel will change version as well to 6.3.

Its actually half on purpose. More releases, more sales, more revenue. But its rushed abit due to the disasterous Windows 8 sales numbers.

Unfortunately Microsoft didnt learn the lesson yet. So no start menu, no boot to desktop.

 

[Fjodor2001]

Windows "Blue". Might be named R2. Windows Server 2012 will also be replaced. Kernel will change version as well to 6.3.

Its actually half on purpose. More releases, more sales, more revenue. But its rushed abit due to the disasterous Windows 8 sales numbers.

Unfortunately Microsoft didnt learn the lesson yet. So no start menu, no boot to desktop.

So what will be the difference between Windows 8 vs "Blue"/R2? Any diff list?

 

[ams23]

ARM segment is not growing. It went backwards the same percentage as x86 in 2012. There was simply canibalism inside the ARM segment.

The number of mobile handheld devices that use ARM processors far outnumbers the number of mobile handheld devices that use x86 processors, and Microsoft has only a very very small piece of that pie. So there was certainly potential upside for Microsoft to expand the reach of Windows to include ARM processors.

And to fracture a ecosystem so you got Windows that is incompatible with Windows was imbecile. Specially when Mcirosoft did a double release. One for each uarch.

It would make little sense for Microsoft to not support ARM processors considering the prevalence of ARM processors in mobile handheld devices. Now, it would be ideal to have one Windows operating system that would somehow support both x86 and ARM processors, but how exactly would that be done in an efficient way (especially considering that Intel will not even allow most companies to emulate an x86 processor)? The only sensible thing to do was to create a "Consumer" Windows OS that is compatible with ARM and a "Professional" Windows OS that is compatible with x86. Microsoft tried to do that, but they did not provide good focus and clarity about differences between these OS's, nor did they provide good product and price segmentation between these OS's.

 

[Homeles]

The idea of creating a Windows operating system that is compatible with ARM processors is a good one, as ARM mobile processors are the fastest growing CPU processors in history. The problem for Microsoft was in the way that they implemented this strategy.

Not to mention how much of an underwhelming piece of garbage Tegra 3 is. If Nvidia thinks they're going to win the ARM war with unmodified ARM cores, they're out of their mind. Thankfully this isn't the case, and we'll see more competitive products from them in the future.

Right now, I feel that the market is simply saturated with hacked solutions to consumer demands. Windows 8's tablet interface, while a step in the right direction, lacks polish. Intel's Ivy Bridge certainly isn't an ideal tablet solution. Ultrabooks are great in concept, but lacking in execution. Tegra so far has been a rather unremarkable series of products. AMD is the epitome of providing consumers with lackluster products. 5 year old Atom is laughable.

There are few exceptions: Apple, some of Samsung's products, and Qualcomm. Coincidentally, these three are also the ones capitalizing most on the current market.

Beginning this year, this will change. Haswell will likely make ultrabooks a far more enticing product, while simultaneously providing a more worthwhile power-tablet solution. I still don't see it as being ideal for tablets, but it should be a significant improvement. Silvermont will likely bring caesura to Atom being the butt of everyone's jokes. AMD's Temash and Kabini look promising. Tegra 4 appears to be still lacking, but we will see once it arrives shortly. We still have Project Denver to look forward to.

Apple, Samsung and Qualcomm are still on the pedestal. I see so much potential in Samsung in particular, but they just aren't leveraging their advantage of being a chip designer, manufacturer, and OEM. Hopefully they can resolve their internal conflicts and become a better version of Apple.

Apple made the interesting move of being a full-blown custom ARM designer, and they already had the advantage of being an OEM with their own OS. As much as dislike their litigation, they are certainly a force to be reckoned with.

Qualcomm makes great chips, but like AMD, they're a one-trick pony. It will be interesting to see how they fare in the coming years. The route they are currently taking leads them to a dead end.

Google, I suppose, is another company to look out for. I personally see Microsoft as being too much of a dinosaur to be able to turn around and become supreme overlord of the mobile world.

The future looks interesting. The present sure isn't.

WHAT?
wow....M$ is really in trouble

Switching to a more frequent release cadence puts them in trouble? I do hope that you're on drugs.

Windows development has always been so slow and predictable... becoming more nimble is exactly what they need.

 

[ams23]

Not to mention how much of an underwhelming piece of garbage Tegra 3 is. If Nvidia thinks they're going to win the ARM war with unmodified ARM cores, they're out of their mind. Thankfully this isn't the case, and we'll see more competitive products from them in the future.

The design of Tegra 3 is not the cause of Microsoft's issues (note that the Google Nexus 7 with Tegra 3 is generally well regarded). Microsoft's timing in releasing Windows RT at the same time as Windows 8 and towards the end of Tegra 3's life cycle was not very good, and their messaging and segmentation between Windows RT and Windows 8 was not very good either.