Guys. Apple still use Intel in their Mac line. They probably would use Atom now if it worked. They collaborate on Thunderbolt. Intel is no more stabbing Apple in the back with the ultrabook then it stabbed MS in the back when it wooed Appke to move from PPC - and thank God they did.
That's nonsense. Apple has a very lucrative market which none of the competition has been able to approach. Rather than leaving it for the market to decide, Intel decided to pump $300 M into Apple's competition with the stated goal of creating a product which would compete directly with Apple's product. AND, there's little stated reason to do it - Intel sells the overwhelming majority of CPUs used in computers today - and that's not likely to change. By subsidizing Apple's competition, Intel is probably not selling any more chips (and maybe getting lower margins on the ones they DO sell if rumors of a 20% price cut are true). So there's no real benefit to Intel and a great disadvantage to Apple. Apple's competitors are being subsidized to attack Apple.
In the case of Apple moving to Intel, the situation is different. First, Intel did not have the business and switching Apple to Intel did increase Intel's volumes. Second, it was a major PR advantage for Intel. Finally, Intel did not pay Apple to create a competitor to Windows - Apple already had a competitor to Windows. At MOST, Intel may have helped with the cost of the transition, but even that isn't clear - and it was certainly not $300 M.
Introducing Googorolla Xoom 2 using Atom CPU. Battery life - 2 hours, but you can run Windows 8 on it!
Those who said that Intel is stabbing apple, not true. Intel has no place to go. Microsoft stabbed them, and apple is using competing designs for its mobile business. So where else but android can they go if they wish to stay relevant? WebOS? Bada? I think Intel is making the right move, though I don't think atom can compete with Arm at this point.
Apple uses ARM for one reason. The technology is better for mobile computing. Intel's offerings use more power, and runs hotter.
Quote:
Originally Posted by addicted44
2) The x86 architecture is an inefficient and inferior architecture. Its success lay in how cheap and widespread it was. However, in the mobile space, the backward compatibility benefits don't exist anymore, so Intel is essentially starting from scratch. There is no reason to believe that Medfield can catch up with ARM anytime soon.
This has been true of Intel's prior implementations of x86 for the mobile space. There's not much evidence that this is a fundamental issue. Once apon a time Transmeta built x86 chips that were top-notch for the mobile platforms of the day. Intel now owns Transmeta's IP and most of their researchers. Given that they (Intel) has consistently had the best process technology, and has renewed its focus on mobility, I'd be surprised if they couldn't create x86 implementations that are both competitive and cheaper than ARM.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
2) The x86 architecture is an inefficient and inferior architecture. Its success lay in how cheap and widespread it was. However, in the mobile space, the backward compatibility benefits don't exist anymore, so Intel is essentially starting from scratch. There is no reason to believe that Medfield can catch up with ARM anytime soon.
Its not so much an architecture problem as Atoms weren't originally designed to compete against ARM in power consumption. IIRC Atom was originally designed with a TDP of 1 to 10 watts, where as ARM are designed to be sub 1 watt.
With Intel targeting lower TDPs, and qualcomm, samsung, TI et al. increasing the performance of their ARM SoCs, eventually the two are going to converge. When and where we don't know yet, but I wouldn't count intel out yet. They have the best process tech (pretty much a full node ahead of everyone else), and some amazing R&D.
If you look at actual efficiency via performance per watt, Atom has ARM beat by a longshot. Atoms benches at 1000+ MFLOPS for 10watts of power used vs something like 30 MFLOPS for 0.5watts for most 1Ghz ARM processors; you do the math. Granted there's a lot of optimizations in play here, but its still a huge margin nonetheless.
And finally keep in mind that atom has not adopted intel's tick-tock development cycle yet. They will starting with the next architecture change, I think, which is when things should really start heating up. We're looking at a 4 year old architecture at the moment.
I think it ridiculous for Intel to try to undercut a customers market. Intel is trying to get Intel's other customers to enter a market Apple (another one of its customers) essentially created by copying Apple.
Intel didn't do anything to Microsoft. First, while they might be partners, Microsoft is not an real Intel customer. Intel partners with Microsoft to help Intel's customers (the PC makers) build computers using chips Intel sells. HP, Dell, and the like are Intel's customers. Adding Apple as a customer didn't undermine any partnership Microsoft and Intel may have together, which is to facilitate the sale of Intel chips to hardware manufacturers that use Intel products. Second, it would be anti-competitive for Intel to turn down Apple as a customer.
Quote:
Originally Posted by asdasd
Guys. Apple still use Intel in their Mac line. They probably would use Atom now if it worked. They collaborate on Thunderbolt. Intel is no more stabbing Apple in the back with the ultrabook then it stabbed MS in the back when it wooed Appke to move from PPC - and thank God they did.
If you ever want to see what a x86 iPhone app runs like ask a dev to run the simulator.
Well of course. Because if Intel and Apple were in a partnership to explore Atom-based iPhone prototypes, Intel wouldn't be allowed to announce anything. ANYTHING.
But I don't think that's happening. Intel's just trying to make Microsoft jealous. They've broken up so many times, you just know they can't quit each other.
I'd be curious about where your information concerning Transmeta is coming from (not saying it is wrong). Transmeta sued Intel. Intel had to pay Transmeta 200 million, plus 20 million a year for five years. Intel obtained a limited license, and some patents. Transmeta is still in business though.
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
Quote:
Originally Posted by jukes
This has been true of Intel's prior implementations of x86 for the mobile space. There's not much evidence that this is a fundamental issue. Once apon a time Transmeta built x86 chips that were top-notch for the mobile platforms of the day. Intel now owns Transmeta's IP and most of their researchers. Given that they (Intel) has consistently had the best process technology, and has renewed its focus on mobility, I'd be surprised if they couldn't create x86 implementations that are both competitive and cheaper than ARM.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
The price of arrogance is abject failure. In this day and age, this is a desperate move by Intel and this was instantly hundreds of millions right down the drain the moment this "partnership" was announced.
Unlike Intel's RubbishGPU-BundleGate they're not going to be able to engage in illegal practices like when they pushed out AMD, ATI and Nvidia.
There's two things Intel has failed to do, despite all their other admirable expertise and the absolutely kick-ass Core series CPUs they've come out with. That's graphics and ARM-level low power SoCs.
Steve Jobs said when he got "fired" from Apple it freed him to be "less sure of himself" and that the burden of success was replaced by "the lightness of being a beginner again".
Intel is hanging on to its dogma of x86 and this will cost them dearly in the mobile and tablet space over the next five years. With Apple's weight (and all their competitors thusly) behind ARM and PowerVR you can bet your bottom dollar ARM and PowerVR is reinvigorated enough to brush off the x86 challenge.
Intel, love ya trucks, but a lot of people want sedans or hatchbacks instead.
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
What has been astonishing is the quick pace at which Apple was able to produce the A5 and the distinct performance gains over the A4, not to mention the ridiculously low power and thermal envelope of the A4 and A5 in the iPad given the apps and multimedia it's able to drive.
This certainly made Intel quiver in their boots. However trying to "fight back" with Atom is IMO a terrible, terrible mistake.
I think that Intel has great chips, with their i7 and I also think that they have pretty good SSDs, I use both of those things, but when I hear mention of the word "Atom", I immediately think of real crappy and weak netbooks.
Atom + Android sounds like a match made in hell, a horrible OS running on a horrible chip. I'll stick with my kickass iPad, thank you very much.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
Well, AMD did say that there A series APU's scale...
interesting to see how AMD's fusion works in the future in regards to all this.
I for one think it will destroy Intel's chances (mobile device = consume media, AMD's graphics = more efficient that Intel's (for now))
Well, AMD did say that there A series APU's scale...
interesting to see how AMD's fusion works in the future in regards to all this.
I for one think it will destroy Intel's chances (mobile device = consume media, AMD's graphics = more efficient that Intel's (for now))
I think AMD shows much promise but 2012 will be their make-or-break year in terms of whether they get any real traction.
Their APUs are a good idea but in 2012 they might find themselves wedged in between CoreULV and ARM. Also, their APUs have an excellent GPU component but it looks like GPGPU has not taken off as we might expect.
And finally keep in mind that atom has not adopted intel's tick-tock development cycle yet. They will starting with the next architecture change, I think, which is when things should really start heating up. We're looking at a 4 year old architecture at the moment.
One thing I am pretty sure of is that they have better solved the battery consumption problem. The other that you were right on about is heat, maybe they could go the way of the fifties and convince everyone bigger is better and large , really large tail fins are cool (heats sinks that is). LOL
What has been astonishing is the quick pace at which Apple was able to produce the A5 and the distinct performance gains over the A4, not to mention the ridiculously low power and thermal envelope of the A4 and A5 in the iPad given the apps and multimedia it's able to drive.
This certainly made Intel quiver in their boots. However trying to "fight back" with Atom is IMO a terrible, terrible mistake.
Well, whatever intel fights back with, its going to be named Atom. What's important is the architecture behind the name.
Yeah, currently its not a good idea since atom as is isn't designed to be used in smartphones, but with a new architecture coming out, I think it holds promise:
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
In terms of absolute performance, I don't think intel has a ways to go; its hard to directly compare the performance of atom and ARM, they play in different fields. The video shows that both are capable of running a browser, but I don't think we can draw too much meaningful information beyond that. The best article I've seen that attempts to compare the performance of atom with ARM is: http://www.brightsideofnews.com/news....aspx?pageid=0 and atom actually does a lot better than expected.
I can't decide whether I want an Android on ARM, Android on Atom, Chrome on X86, Android on Itanium, Chrome on dome, Android on moon, or Ice Cream Sandwich on Schmidt.
That's nonsense. Apple has a very lucrative market which none of the competition has been able to approach. Rather than leaving it for the market to decide, Intel decided to pump $300 M into Apple's competition with the stated goal of creating a product which would compete directly with Apple's product. AND, there's little stated reason to do it - Intel sells the overwhelming majority of CPUs used in computers today - and that's not likely to change. By subsidizing Apple's competition, Intel is probably not selling any more chips (and maybe getting lower margins on the ones they DO sell if rumors of a 20% price cut are true). So there's no real benefit to Intel and a great disadvantage to Apple. Apple's competitors are being subsidized to attack Apple.
In the case of Apple moving to Intel, the situation is different. First, Intel did not have the business and switching Apple to Intel did increase Intel's volumes. Second, it was a major PR advantage for Intel. Finally, Intel did not pay Apple to create a competitor to Windows - Apple already had a competitor to Windows. At MOST, Intel may have helped with the cost of the transition, but even that isn't clear - and it was certainly not $300 M.
There is nothing stopping Apple from grabbing that $300M and incorporating Intel's cpu in tablets.
Comments
Guys. Apple still use Intel in their Mac line. They probably would use Atom now if it worked. They collaborate on Thunderbolt. Intel is no more stabbing Apple in the back with the ultrabook then it stabbed MS in the back when it wooed Appke to move from PPC - and thank God they did.
That's nonsense. Apple has a very lucrative market which none of the competition has been able to approach. Rather than leaving it for the market to decide, Intel decided to pump $300 M into Apple's competition with the stated goal of creating a product which would compete directly with Apple's product. AND, there's little stated reason to do it - Intel sells the overwhelming majority of CPUs used in computers today - and that's not likely to change. By subsidizing Apple's competition, Intel is probably not selling any more chips (and maybe getting lower margins on the ones they DO sell if rumors of a 20% price cut are true). So there's no real benefit to Intel and a great disadvantage to Apple. Apple's competitors are being subsidized to attack Apple.
In the case of Apple moving to Intel, the situation is different. First, Intel did not have the business and switching Apple to Intel did increase Intel's volumes. Second, it was a major PR advantage for Intel. Finally, Intel did not pay Apple to create a competitor to Windows - Apple already had a competitor to Windows. At MOST, Intel may have helped with the cost of the transition, but even that isn't clear - and it was certainly not $300 M.
Those who said that Intel is stabbing apple, not true. Intel has no place to go. Microsoft stabbed them, and apple is using competing designs for its mobile business. So where else but android can they go if they wish to stay relevant? WebOS? Bada? I think Intel is making the right move, though I don't think atom can compete with Arm at this point.
Apple uses ARM for one reason. The technology is better for mobile computing. Intel's offerings use more power, and runs hotter.
2) The x86 architecture is an inefficient and inferior architecture. Its success lay in how cheap and widespread it was. However, in the mobile space, the backward compatibility benefits don't exist anymore, so Intel is essentially starting from scratch. There is no reason to believe that Medfield can catch up with ARM anytime soon.
This has been true of Intel's prior implementations of x86 for the mobile space. There's not much evidence that this is a fundamental issue. Once apon a time Transmeta built x86 chips that were top-notch for the mobile platforms of the day. Intel now owns Transmeta's IP and most of their researchers. Given that they (Intel) has consistently had the best process technology, and has renewed its focus on mobility, I'd be surprised if they couldn't create x86 implementations that are both competitive and cheaper than ARM.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
2) The x86 architecture is an inefficient and inferior architecture. Its success lay in how cheap and widespread it was. However, in the mobile space, the backward compatibility benefits don't exist anymore, so Intel is essentially starting from scratch. There is no reason to believe that Medfield can catch up with ARM anytime soon.
Its not so much an architecture problem as Atoms weren't originally designed to compete against ARM in power consumption. IIRC Atom was originally designed with a TDP of 1 to 10 watts, where as ARM are designed to be sub 1 watt.
With Intel targeting lower TDPs, and qualcomm, samsung, TI et al. increasing the performance of their ARM SoCs, eventually the two are going to converge. When and where we don't know yet, but I wouldn't count intel out yet. They have the best process tech (pretty much a full node ahead of everyone else), and some amazing R&D.
If you look at actual efficiency via performance per watt, Atom has ARM beat by a longshot. Atoms benches at 1000+ MFLOPS for 10watts of power used vs something like 30 MFLOPS for 0.5watts for most 1Ghz ARM processors; you do the math. Granted there's a lot of optimizations in play here, but its still a huge margin nonetheless.
And finally keep in mind that atom has not adopted intel's tick-tock development cycle yet. They will starting with the next architecture change, I think, which is when things should really start heating up. We're looking at a 4 year old architecture at the moment.
Presumably just a recompile?
But if you have hand coded SIMD stuff in there it could be more. But it also greatly increases regression testing.
Intel didn't do anything to Microsoft. First, while they might be partners, Microsoft is not an real Intel customer. Intel partners with Microsoft to help Intel's customers (the PC makers) build computers using chips Intel sells. HP, Dell, and the like are Intel's customers. Adding Apple as a customer didn't undermine any partnership Microsoft and Intel may have together, which is to facilitate the sale of Intel chips to hardware manufacturers that use Intel products. Second, it would be anti-competitive for Intel to turn down Apple as a customer.
Guys. Apple still use Intel in their Mac line. They probably would use Atom now if it worked. They collaborate on Thunderbolt. Intel is no more stabbing Apple in the back with the ultrabook then it stabbed MS in the back when it wooed Appke to move from PPC - and thank God they did.
If you ever want to see what a x86 iPhone app runs like ask a dev to run the simulator.
But I don't think that's happening. Intel's just trying to make Microsoft jealous. They've broken up so many times, you just know they can't quit each other.
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
This has been true of Intel's prior implementations of x86 for the mobile space. There's not much evidence that this is a fundamental issue. Once apon a time Transmeta built x86 chips that were top-notch for the mobile platforms of the day. Intel now owns Transmeta's IP and most of their researchers. Given that they (Intel) has consistently had the best process technology, and has renewed its focus on mobility, I'd be surprised if they couldn't create x86 implementations that are both competitive and cheaper than ARM.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
Unlike Intel's RubbishGPU-BundleGate they're not going to be able to engage in illegal practices like when they pushed out AMD, ATI and Nvidia.
There's two things Intel has failed to do, despite all their other admirable expertise and the absolutely kick-ass Core series CPUs they've come out with. That's graphics and ARM-level low power SoCs.
Steve Jobs said when he got "fired" from Apple it freed him to be "less sure of himself" and that the burden of success was replaced by "the lightness of being a beginner again".
Intel is hanging on to its dogma of x86 and this will cost them dearly in the mobile and tablet space over the next five years. With Apple's weight (and all their competitors thusly) behind ARM and PowerVR you can bet your bottom dollar ARM and PowerVR is reinvigorated enough to brush off the x86 challenge.
Intel, love ya trucks, but a lot of people want sedans or hatchbacks instead.
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
What has been astonishing is the quick pace at which Apple was able to produce the A5 and the distinct performance gains over the A4, not to mention the ridiculously low power and thermal envelope of the A4 and A5 in the iPad given the apps and multimedia it's able to drive.
This certainly made Intel quiver in their boots. However trying to "fight back" with Atom is IMO a terrible, terrible mistake.
Atom + Android sounds like a match made in hell, a horrible OS running on a horrible chip. I'll stick with my kickass iPad, thank you very much.
ARM's main advantage is the flexibility to source manufacturing with anyone in the fab industry. Until AMD really gets a mobile x86 out there, committing to x86 means committing to Intel.
Well, AMD did say that there A series APU's scale...
interesting to see how AMD's fusion works in the future in regards to all this.
I for one think it will destroy Intel's chances (mobile device = consume media, AMD's graphics = more efficient that Intel's (for now))
Well, AMD did say that there A series APU's scale...
interesting to see how AMD's fusion works in the future in regards to all this.
I for one think it will destroy Intel's chances (mobile device = consume media, AMD's graphics = more efficient that Intel's (for now))
I think AMD shows much promise but 2012 will be their make-or-break year in terms of whether they get any real traction.
Their APUs are a good idea but in 2012 they might find themselves wedged in between CoreULV and ARM. Also, their APUs have an excellent GPU component but it looks like GPGPU has not taken off as we might expect.
And finally keep in mind that atom has not adopted intel's tick-tock development cycle yet. They will starting with the next architecture change, I think, which is when things should really start heating up. We're looking at a 4 year old architecture at the moment.
One thing I am pretty sure of is that they have better solved the battery consumption problem. The other that you were right on about is heat, maybe they could go the way of the fifties and convince everyone bigger is better and large , really large tail fins are cool (heats sinks that is). LOL
What has been astonishing is the quick pace at which Apple was able to produce the A5 and the distinct performance gains over the A4, not to mention the ridiculously low power and thermal envelope of the A4 and A5 in the iPad given the apps and multimedia it's able to drive.
This certainly made Intel quiver in their boots. However trying to "fight back" with Atom is IMO a terrible, terrible mistake.
Well, whatever intel fights back with, its going to be named Atom. What's important is the architecture behind the name.
Yeah, currently its not a good idea since atom as is isn't designed to be used in smartphones, but with a new architecture coming out, I think it holds promise:
http://www.anandtech.com/show/4333/i...m-architecture
If this video is any indictor of the competition, Intel still has a ways to go. That also is assuming ARM stands still. ARM currently has an advantage because companies like Apple can custom tailor ARM's specification to met individual needs. That is one of the reasons it will be hard to beat the iPad, as Apple is using a chip custom tailored to Apple's needs.
In terms of absolute performance, I don't think intel has a ways to go; its hard to directly compare the performance of atom and ARM, they play in different fields. The video shows that both are capable of running a browser, but I don't think we can draw too much meaningful information beyond that. The best article I've seen that attempts to compare the performance of atom with ARM is: http://www.brightsideofnews.com/news....aspx?pageid=0 and atom actually does a lot better than expected.
Eh, screw'em all.
I'll just take an iPhone 5.
That's nonsense. Apple has a very lucrative market which none of the competition has been able to approach. Rather than leaving it for the market to decide, Intel decided to pump $300 M into Apple's competition with the stated goal of creating a product which would compete directly with Apple's product. AND, there's little stated reason to do it - Intel sells the overwhelming majority of CPUs used in computers today - and that's not likely to change. By subsidizing Apple's competition, Intel is probably not selling any more chips (and maybe getting lower margins on the ones they DO sell if rumors of a 20% price cut are true). So there's no real benefit to Intel and a great disadvantage to Apple. Apple's competitors are being subsidized to attack Apple.
In the case of Apple moving to Intel, the situation is different. First, Intel did not have the business and switching Apple to Intel did increase Intel's volumes. Second, it was a major PR advantage for Intel. Finally, Intel did not pay Apple to create a competitor to Windows - Apple already had a competitor to Windows. At MOST, Intel may have helped with the cost of the transition, but even that isn't clear - and it was certainly not $300 M.
There is nothing stopping Apple from grabbing that $300M and incorporating Intel's cpu in tablets.
There is nothing stopping Apple from grabbing that $300M and incorporating Intel's cpu in tablets.
That will never happen until 2013 at the very earliest. And again as people point out that's assuming ARM and PowerVR stand still.