Sim free Samsung Galaxy S4 in Argos £599.95 although it does have £50 off at the moment making it only 95 pence more expensive than the 5s.
Nice try, sunshine.
For those not familiar with the UK market - Argos is a rather expensive outlet, so by picking them, Crosslad has tried to mislead. Amazon.co.uk has the SG4 16Gb model at £428 SIM free. An HTC One 32 Gb can be had for £375. The iPhone 5s is very expensive.
Galbi, you're funny, as in funny-strange. Apple knows it's making a faster mobile processor and a faster microprocessor in general, that's what. To start with, the A7 has twice as many general purpose registers as the A6 or even Intel's X64 processor architecture. (Oh, wouldn't Intel and AMD like to have 16 more registers to work with? " src="http://forums-files.appleinsider.com/images/smilies//lol.gif" />)
What makes these forums colorful are all the folks who post an uninformed opinion before they know much about the topic. Almost too easy to pick them off. What's amusing is that people who do understand the meaning of 64-bit, like Anand Shimpi from Anandtech, are ignored by these uninformed people. Galbi must think he's smarter than Anand. And that's funny.
Is exynos really that bad? I know people were disappointed that it wasn't a true octa-core processor (basically 2 quad-core processors glued together), but it sounds good in theory. Lower power consumption when the slower processor "is good enough" and switches to the faster processor when the user would value speed over battery life.
No. Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos. Its the high end processor. It suffers from basically the same thing Apple is now being 'accused' of. It was released to much fanfare as being an 'eight core processor' but couldn't actually use all 8 at once. It was also expensive to make. It has *huge* potential for the future.... but that's the future =P
It's a little like putting a huge/fast engine in a race car and then not giving it a means to deliver enough gas. It is going to be expensive, but you won't get the optimal performance boost from it.
Apple is being hit with a little bit of the same. They built a 64-bit phone, but there are plenty of other limiting factors that will prevent it from reaching its potential. One of the legit ones is Apple only gave the 5s 1Gig of memory. With a 32-bit bus you can address up to 4GB of memory. So if you have only 1GB of memory there really isn't a driving need for 64-bit since 32-bit really wasn't limiting you.
To use the prior analogy, there really wouldn't be a benefit to a race car to come up with a way to deliver more gas if the engine really didn't have a means to consume it and translate it to power delivered.
I think in both cases the pundits are wrong... If the 5s is delivering twice the performance of the 5 as claimed, that right there is a win either way you slice it- and the fact that is 64bit means it has a ton of potential in future iterations once Apple fixes the 'limiting factors' the pundits are currently focusing on.
Same is true for the Exynos too. If they get all 8 cores firing (and they recently have), then shore up their limiting factors, the thing has a lot of potential for growth.
Focusing on the 'fastest' part of any computer is always a mistake because Amdahl's Law will always come into effect. It happened in PC's when everyone was so focused on building the fastest processors that improving processors started having limited effect- they had to focus on the buses and other limiting factors. The strength of a chain is determined by its weakest link. The speed of a computer (or phone) is determined by its slowest bottleneck.
Either way its good news for consumers. The 5s being twice as fast as the 5 is something any Apple fan should be happy with, and any pundit that misses that is missing the forest for focusing on a tree.
You gain a performance benefit, even if the memory barrier is not an issue.
The A7 has more general purpose registers (31 instead of 14), and these are all 64-bits wide. More registers means less use of the stack or RAM to hold temporary variables such as pointers, counters, etc. If you have a large block of data to process, a 64-bit program can do it in 64-bit chunks at a time, instead of looping through twice as often, processing it 32-bits at a time. There are a few other novelties in the 64-bit ARM design, like PC-relative addressing, guaranteed 128-bit SIMD operations, etc., none of which are about memory barriers.
For most people, the A7 will mean "faster."
One reason I believe Apple chose to jump to Arm 64-Bit is the extra functionality, including AES cryptography and SHA-1/256 hashing on the chip as well as double precision floating point. Both these can lighten the load the processor does currently via software, as well as speeding up the device
From reading other websites it may also enable Apple to sandbox a lot better which would help in the fight against Viruses/Malware and rogue processes. It could even let Apple implement "users" on the iPhone/iPad, which is something we all would like.
... Also the diagram above for the ARM Cortex-A50 series, indicates that the designs are for 20nm and 14nm processes, so maybe the A7 does actually already use a 20nm process rather than the 28nm or 32nm processes previously suspected...now that would really be a jump on the industry. For once Apple's 'doubling down' on security paid off!
Exactly. Now that Apple has a 64-bit SoC, it's possible that all they need to do for the next two or three years is to shrink the die. That means next year's "iPhone 6" might have a 14nm A8, which would be functionally identical to the A7 but faster. Smaller die -> shorter leads -> higher potential clock speeds and/or lower power consumption. A few years down the road, once the process is small enough, Apple could add more cores. If that amount of computing power is actually necessary in an iPhone. (The real trick, of course, would be replacing Intel's power-hungry hot-running x86 legacy chips in the MacBook Air line. The lack of 64-bit processing was the last technical detail preventing that. Apple may soon be able to avoid the Intel Tax.)
Apple's "tick tock" iPhone development cycle only applies to the enclosure. The 3G/3GS, 4/4S, and now 5/5S had similar if not identical enclosures. So from a consumer standpoint, the even-year iPhones appear to be similar to the previous year's iPhones. But the iPhone internal hardware and iOS advance steadily every year. A fact that some bloggers furiously spin as boring and inconsequential. Not any more. With the 64-bit A7, Apple's engineering has broken clear of the pack. Samsung and all would-be competitors are going to fall further behind every year. Especially with Android, the messy mash-up of a boat anchor holding them back.
So yeah, let's all make some popcorn and see how the haters try to spin 64-bit computing in the iPhone 5S. Some will claim it's a "hoax." Others will claim that there is no technical advantage to the ability to address more than a terabyte of RAM. But there's only a small window of opportunity for them to attempt that kind of negative spin. Once the benchmarks are posted, it will be radio silence with respect to 64-bit computing power. Their only talking point will be screen size, which isn't an issue in developed countries.
Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos.
Why would we wish a pox on all our houses? North America didn't get the Exynos because we have broad access to LTE on this continent and LTE consumes significantly more power. Combine more cores with LTE in a Samsung phone and you get even worse battery life. No, thank you!
I guess Samsung might have a major problem moving to ARMv8 namely Android support.
Apple has both create an adapted iOS release updated xcode too enable 3rd party developers to recompile their apps - something Samsung can't do on their own...
That is like blowing a 10 gallon air compressor through a straw!
Apple what the heck do you think you are doing?
That is nothing more than marketing ploy to get people reinterested in the smartphone as we are already seeing saturation in the market place.
Sure it's makes a good marketing bullet point, but I wouldn't call it a ploy as I see a real benefit to iPhone going 64-bit that no one around the web seems to be mentioning. While some people gargle on DED's balls as he trys to defend a technology that offers little benefit in its current implementation, I think Apple is busy setting up a base for future products (I believe jungmark hinted at this). Having both 32-bit and 64-bit versions of an OS adds a level of fragmentation and introduces the need for backwards compatibility. Once iOS reaches a point where the benefits of having a 64-bit OS can shine, I'm betting that they won't have to be concerned about offering 32-bit compatibility at that point as the legacy devices will then be 64-bit as well. If this is what Apple is planning then I think it's a really smart move on their part.
There are a few other novelties in the 64-bit ARM design, like PC-relative addressing, guaranteed 128-bit SIMD operations, etc., none of which are about memory barriers.
For most people, the A7 will mean "faster."
Yes, the 64bit move is definitely not made to overcome memory issues as the Cortex A15 has LPAE support and could address more than 4GB RAM.
Also I have yet to find a confirmation on Apple's A7 architecture and whether it matches ARMv8 (especially regarding available registers), is there any source for that?
By the way, 128bit SIMD operations are present on todays 32bit SoCs as well:
The Cortex A8 (A9 as well afaik) provides NEON which has instructions and does 128bit operations in 2 cycles.
Cortex A15 comes with the successor of NEON and can accomplish 128bit operations in a single cycle per core.
My guess would be that most performance improvements come from optimizations in their AArch32.
Quote:
Originally Posted by snitkjaer
I guess Samsung might have a major problem moving to ARMv8 namely Android support.
Apple has both create an adapted iOS release updated xcode too enable 3rd party developers to recompile their apps - something Samsung can't do on their own...
ARMv8 support is in the linux kernel since 2012 and Kernel 3.7. So the move to ARMv8 should be flawless.
All apps should be backwards compatible (Java apps are ovious, native 32bit code runs on AArch32).
It will be interesting to see if google can improve Dalvik's (virtual machine) performance for AArch64 - that would give all android apps a small performance boost without having to recompile them.
"very few apps that will take advantage of the processor"
The apps that count will!
And let's not forget the A7X variants for the next iPads and next AppleTV. In one move, Apple can:
differentiate itself from the competition in tablets
become a major player in the console game market with AppleTV
become the leader in a 4K capable AppleTV
create a new market in "personal TV" -- The AppleTV concurrently streams the same (or different) live or recorded videos to multiple iPads *
* Imagine the possibilities: in the home; in the meeting room; in the board room; in the classroom; in the operating room; on set (movie and broadcast/cable TV)
What's holding Apple back from opening up the Apple TV to the AppStore? Instantly have a library of hundreds of thousands of games, game controller compatibility, console-level graphics with the new A7 processors ... As well as other types of media, consumption, and even productivity apps ...
I am willing to bet my right arm that Apple isn't developing an iwatch, but will partner with other companies (such as Nike) to make wearable tech. I think iwatch trademarking is a red herring, and while some trip over themselves (Samsung) to make smartwatches, Apple will take over the living room, the car, and who knows what else.
No. Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos. Its the high end processor. It suffers from basically the same thing Apple is now being 'accused' of. It was released to much fanfare as being an 'eight core processor' but couldn't actually use all 8 at once. It was also expensive to make. It has *huge* potential for the future.... but that's the future =P
If you're only thinking about the specs and benchmarks of the Exynos 5, then yes, it's not "that bad" at all. But in the real world, you can't ignore realities (including cost!) when engineering precision devices selling in massive quantities.
Recall that outside of Apple, nobody really saw any immediate potential in AuthenTech either, because (in the SEC filings of the company) its parts and the work required to integrate them were deemed too expensive. Only Apple saw a way to make Touch ID work AND be affordable and profitable. That's engineering, not just churning out a bigger/faster/higher spec'ed device. That's PC-era doodling.
Exynos 5 clearly cost Samsung lots to develop, just as Apple spent lots of resources on its A6. But Apple put its A6 in every iPhone 5 it sold, and then sold more iPhone 5 units that any other smartphone. That's how to drive down costs. And this year, its using the A6 in the 5c, having driven down costs enough to sell its popular phone (which continued to outsell Samsung's flagship) for $100 less.
The A7 is an ambitious, expensive part. But Apple knows it won't have too much difficulty selling huge volumes this year, and in iPads, and in next years mainstream phone. This is the exciting part of technological progress. And Apple is leading it.
Apple did the same thing with lots of other technologies, leveraging economies of scale to take expensive new technologies and make them affordable, while still making a profit. This is not easy to do.
That's why most companies, including all those White Box tablet makers and PC cloners, only ever assembled parts off the shelf and slopped out yesterday's technology at commensurate prices.
Samsung's costs in developing the (several versions of) Exynos 5 are not benefitting from total global shipments of the GS4 and other flagship models than can command a premium (relative to Samsung's other basic phones running Android 2.x). It's missing vast economies of scale, sending its profits to Qualcomm instead.
Essentially, the difficult position where Apple found itself, buying components from (and enriching) its main rival, is now being inherited by Samsung, which has to pay Qualcomm for finished chips that are designed by Qualcomm (and apparently built by Samsung, at very little profit if you look at the commercial performance of System LSI). That's not how you drive your costs down. That's how you drive your competitor's costs down, as LG and Microkia everyone else now benefits from your economies of scale (but only slightly).
Looking at things from that perspective, the design of Exynos 5 Octa was a huge, grievous error. It would be like Steve Jobs waiting for IBM to deliver the PPC G6 and then being forced to put Intel Core chips in its low end Macs, failing to gain any leverage on either chip platform.
What's holding Apple back from opening up the Apple TV to the AppStore? Instantly have a library of hundreds of thousands of games, game controller compatibility, console-level graphics with the new A7 processors ... As well as other types of media, consumption, and even productivity apps ...
Apple doesn't want to open a failure. There are already SmartTVs and their app stores that aren't really offering anything interesting or useful. It would be a huge resource-sapping distraction for Apple to launch a TV store that nobody cared about, rather than focusing those resources on things that matter.
Congrats to Daniel on a thoughtful article that blends insights from both a business and an engineering perspective.
Having designed integrated circuits I can attest to the complexity in developing a new design. If you add the complexities of moving to a new vendor, modifying your entire OS to take advantage of a 64-bit architecture, modifying the apps that run on that OS, and offloading all of the sensor management to a new M7 processor, it becomes clear that the architectural strategy behind iOS 7 and the A7 chip was far more complicated than it initially appeared.
Daniel also correctly stated the supply chain issues surrounding this level and speed of innovation. The ability of other mobile device manufacturers to replicate Apple's accomplishment will be hindered and delayed for years due to their dependency on OTS parts that don't exist.
I think everyone in the technology media and the mobile industry was simply distracted by the modifications to the look and feel of iOS7, and prior to the Sept. 10 announcement of the iPhone 5s failed to appreciate that Apple has taken a huge leap ahead of the rest of the mobile industry and that there devices will be in a class of their own for the next several years.
"Anyone who thinks they know any of this better than Apple is either wrong or wasting their lives if they are not already generating billions of dollars in value with their grand expertise. Anyone who thinks they know any of this better than Apple is either wrong or wasting their lives if they are not already generating billions of dollars in value with their grand expertise."
This bears repeating. Over and over.
Basically the ultimate rebuke to every single troll nonsense comment made here since the 10th.
It's bad enough that sections of the Android community stereotype us as sheep, without you handing them a smoking gun.
Comments
Yes.
Sim free iPhone 5s in UK Apple Store £549.00
Sim free Samsung Galaxy S4 in Argos £599.95 although it does have £50 off at the moment making it only 95 pence more expensive than the 5s.
Nice try, sunshine.
For those not familiar with the UK market - Argos is a rather expensive outlet, so by picking them, Crosslad has tried to mislead. Amazon.co.uk has the SG4 16Gb model at £428 SIM free. An HTC One 32 Gb can be had for £375. The iPhone 5s is very expensive.
Galbi, you're funny, as in funny-strange. Apple knows it's making a faster mobile processor and a faster microprocessor in general, that's what. To start with, the A7 has twice as many general purpose registers as the A6 or even Intel's X64 processor architecture. (Oh, wouldn't Intel and AMD like to have 16 more registers to work with?
" src="http://forums-files.appleinsider.com/images/smilies//lol.gif" />)
Is exynos really that bad? I know people were disappointed that it wasn't a true octa-core processor (basically 2 quad-core processors glued together), but it sounds good in theory. Lower power consumption when the slower processor "is good enough" and switches to the faster processor when the user would value speed over battery life.
No. Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos. Its the high end processor. It suffers from basically the same thing Apple is now being 'accused' of. It was released to much fanfare as being an 'eight core processor' but couldn't actually use all 8 at once. It was also expensive to make. It has *huge* potential for the future.... but that's the future =P
It's a little like putting a huge/fast engine in a race car and then not giving it a means to deliver enough gas. It is going to be expensive, but you won't get the optimal performance boost from it.
Apple is being hit with a little bit of the same. They built a 64-bit phone, but there are plenty of other limiting factors that will prevent it from reaching its potential. One of the legit ones is Apple only gave the 5s 1Gig of memory. With a 32-bit bus you can address up to 4GB of memory. So if you have only 1GB of memory there really isn't a driving need for 64-bit since 32-bit really wasn't limiting you.
To use the prior analogy, there really wouldn't be a benefit to a race car to come up with a way to deliver more gas if the engine really didn't have a means to consume it and translate it to power delivered.
I think in both cases the pundits are wrong... If the 5s is delivering twice the performance of the 5 as claimed, that right there is a win either way you slice it- and the fact that is 64bit means it has a ton of potential in future iterations once Apple fixes the 'limiting factors' the pundits are currently focusing on.
Same is true for the Exynos too. If they get all 8 cores firing (and they recently have), then shore up their limiting factors, the thing has a lot of potential for growth.
Focusing on the 'fastest' part of any computer is always a mistake because Amdahl's Law will always come into effect. It happened in PC's when everyone was so focused on building the fastest processors that improving processors started having limited effect- they had to focus on the buses and other limiting factors. The strength of a chain is determined by its weakest link. The speed of a computer (or phone) is determined by its slowest bottleneck.
Either way its good news for consumers. The 5s being twice as fast as the 5 is something any Apple fan should be happy with, and any pundit that misses that is missing the forest for focusing on a tree.
One reason I believe Apple chose to jump to Arm 64-Bit is the extra functionality, including AES cryptography and SHA-1/256 hashing on the chip as well as double precision floating point. Both these can lighten the load the processor does currently via software, as well as speeding up the device
From reading other websites it may also enable Apple to sandbox a lot better which would help in the fight against Viruses/Malware and rogue processes. It could even let Apple implement "users" on the iPhone/iPad, which is something we all would like.
Refernces:
http://en.wikipedia.org/wiki/ARM_architecture#ARMv8_and_64-bit
... Also the diagram above for the ARM Cortex-A50 series, indicates that the designs are for 20nm and 14nm processes, so maybe the A7 does actually already use a 20nm process rather than the 28nm or 32nm processes previously suspected...now that would really be a jump on the industry. For once Apple's 'doubling down' on security paid off!
Exactly. Now that Apple has a 64-bit SoC, it's possible that all they need to do for the next two or three years is to shrink the die. That means next year's "iPhone 6" might have a 14nm A8, which would be functionally identical to the A7 but faster. Smaller die -> shorter leads -> higher potential clock speeds and/or lower power consumption. A few years down the road, once the process is small enough, Apple could add more cores. If that amount of computing power is actually necessary in an iPhone. (The real trick, of course, would be replacing Intel's power-hungry hot-running x86 legacy chips in the MacBook Air line. The lack of 64-bit processing was the last technical detail preventing that. Apple may soon be able to avoid the Intel Tax.)
Apple's "tick tock" iPhone development cycle only applies to the enclosure. The 3G/3GS, 4/4S, and now 5/5S had similar if not identical enclosures. So from a consumer standpoint, the even-year iPhones appear to be similar to the previous year's iPhones. But the iPhone internal hardware and iOS advance steadily every year. A fact that some bloggers furiously spin as boring and inconsequential. Not any more. With the 64-bit A7, Apple's engineering has broken clear of the pack. Samsung and all would-be competitors are going to fall further behind every year. Especially with Android, the messy mash-up of a boat anchor holding them back.
So yeah, let's all make some popcorn and see how the haters try to spin 64-bit computing in the iPhone 5S. Some will claim it's a "hoax." Others will claim that there is no technical advantage to the ability to address more than a terabyte of RAM. But there's only a small window of opportunity for them to attempt that kind of negative spin. Once the benchmarks are posted, it will be radio silence with respect to 64-bit computing power. Their only talking point will be screen size, which isn't an issue in developed countries.
Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos.
Why would we wish a pox on all our houses? North America didn't get the Exynos because we have broad access to LTE on this continent and LTE consumes significantly more power. Combine more cores with LTE in a Samsung phone and you get even worse battery life. No, thank you!
AppleInsider.
Come for the news.
Stay for the editorials.
(Another great piece, by the way.) Thank you!
Apple has both create an adapted iOS release updated xcode too enable 3rd party developers to recompile their apps - something Samsung can't do on their own...
Indeed. By the Left, for the Left. Apple is anathema to them, and must not be praised beyond the perfunctory level.
64 Bit with only 1-2GB of RAM?
That is like blowing a 10 gallon air compressor through a straw!
Apple what the heck do you think you are doing?
That is nothing more than marketing ploy to get people reinterested in the smartphone as we are already seeing saturation in the market place.
Sure it's makes a good marketing bullet point, but I wouldn't call it a ploy as I see a real benefit to iPhone going 64-bit that no one around the web seems to be mentioning. While some people gargle on DED's balls as he trys to defend a technology that offers little benefit in its current implementation, I think Apple is busy setting up a base for future products (I believe jungmark hinted at this). Having both 32-bit and 64-bit versions of an OS adds a level of fragmentation and introduces the need for backwards compatibility. Once iOS reaches a point where the benefits of having a 64-bit OS can shine, I'm betting that they won't have to be concerned about offering 32-bit compatibility at that point as the legacy devices will then be 64-bit as well. If this is what Apple is planning then I think it's a really smart move on their part.
There are a few other novelties in the 64-bit ARM design, like PC-relative addressing, guaranteed 128-bit SIMD operations, etc., none of which are about memory barriers.
For most people, the A7 will mean "faster."
Yes, the 64bit move is definitely not made to overcome memory issues as the Cortex A15 has LPAE support and could address more than 4GB RAM.
Also I have yet to find a confirmation on Apple's A7 architecture and whether it matches ARMv8 (especially regarding available registers), is there any source for that?
By the way, 128bit SIMD operations are present on todays 32bit SoCs as well:
The Cortex A8 (A9 as well afaik) provides NEON which has instructions and does 128bit operations in 2 cycles.
Cortex A15 comes with the successor of NEON and can accomplish 128bit operations in a single cycle per core.
My guess would be that most performance improvements come from optimizations in their AArch32.
I guess Samsung might have a major problem moving to ARMv8 namely Android support.
Apple has both create an adapted iOS release updated xcode too enable 3rd party developers to recompile their apps - something Samsung can't do on their own...
ARMv8 support is in the linux kernel since 2012 and Kernel 3.7. So the move to ARMv8 should be flawless.
All apps should be backwards compatible (Java apps are ovious, native 32bit code runs on AArch32).
It will be interesting to see if google can improve Dalvik's (virtual machine) performance for AArch64 - that would give all android apps a small performance boost without having to recompile them.
No. Exynos is not 'that bad' at all and actually kind of rocks. People with the QUALCOMM chips (North American Samsung phones) *wish* they had the Exynos. Its the high end processor. It suffers from basically the same thing Apple is now being 'accused' of. It was released to much fanfare as being an 'eight core processor' but couldn't actually use all 8 at once. It was also expensive to make. It has *huge* potential for the future.... but that's the future =P
If you're only thinking about the specs and benchmarks of the Exynos 5, then yes, it's not "that bad" at all. But in the real world, you can't ignore realities (including cost!) when engineering precision devices selling in massive quantities.
Recall that outside of Apple, nobody really saw any immediate potential in AuthenTech either, because (in the SEC filings of the company) its parts and the work required to integrate them were deemed too expensive. Only Apple saw a way to make Touch ID work AND be affordable and profitable. That's engineering, not just churning out a bigger/faster/higher spec'ed device. That's PC-era doodling.
Exynos 5 clearly cost Samsung lots to develop, just as Apple spent lots of resources on its A6. But Apple put its A6 in every iPhone 5 it sold, and then sold more iPhone 5 units that any other smartphone. That's how to drive down costs. And this year, its using the A6 in the 5c, having driven down costs enough to sell its popular phone (which continued to outsell Samsung's flagship) for $100 less.
The A7 is an ambitious, expensive part. But Apple knows it won't have too much difficulty selling huge volumes this year, and in iPads, and in next years mainstream phone. This is the exciting part of technological progress. And Apple is leading it.
Apple did the same thing with lots of other technologies, leveraging economies of scale to take expensive new technologies and make them affordable, while still making a profit. This is not easy to do.
That's why most companies, including all those White Box tablet makers and PC cloners, only ever assembled parts off the shelf and slopped out yesterday's technology at commensurate prices.
Samsung's costs in developing the (several versions of) Exynos 5 are not benefitting from total global shipments of the GS4 and other flagship models than can command a premium (relative to Samsung's other basic phones running Android 2.x). It's missing vast economies of scale, sending its profits to Qualcomm instead.
Essentially, the difficult position where Apple found itself, buying components from (and enriching) its main rival, is now being inherited by Samsung, which has to pay Qualcomm for finished chips that are designed by Qualcomm (and apparently built by Samsung, at very little profit if you look at the commercial performance of System LSI). That's not how you drive your costs down. That's how you drive your competitor's costs down, as LG and Microkia everyone else now benefits from your economies of scale (but only slightly).
Looking at things from that perspective, the design of Exynos 5 Octa was a huge, grievous error. It would be like Steve Jobs waiting for IBM to deliver the PPC G6 and then being forced to put Intel Core chips in its low end Macs, failing to gain any leverage on either chip platform.
What's holding Apple back from opening up the Apple TV to the AppStore? Instantly have a library of hundreds of thousands of games, game controller compatibility, console-level graphics with the new A7 processors ... As well as other types of media, consumption, and even productivity apps ...
Apple doesn't want to open a failure. There are already SmartTVs and their app stores that aren't really offering anything interesting or useful. It would be a huge resource-sapping distraction for Apple to launch a TV store that nobody cared about, rather than focusing those resources on things that matter.
Priceless!
With that said, if I were Samsung, I would simply switch to adult entertainment business!
Having designed integrated circuits I can attest to the complexity in developing a new design. If you add the complexities of moving to a new vendor, modifying your entire OS to take advantage of a 64-bit architecture, modifying the apps that run on that OS, and offloading all of the sensor management to a new M7 processor, it becomes clear that the architectural strategy behind iOS 7 and the A7 chip was far more complicated than it initially appeared.
Daniel also correctly stated the supply chain issues surrounding this level and speed of innovation. The ability of other mobile device manufacturers to replicate Apple's accomplishment will be hindered and delayed for years due to their dependency on OTS parts that don't exist.
I think everyone in the technology media and the mobile industry was simply distracted by the modifications to the look and feel of iOS7, and prior to the Sept. 10 announcement of the iPhone 5s failed to appreciate that Apple has taken a huge leap ahead of the rest of the mobile industry and that there devices will be in a class of their own for the next several years.
Thank you Daniel for such an interesting read!
"Anyone who thinks they know any of this better than Apple is either wrong or wasting their lives if they are not already generating billions of dollars in value with their grand expertise. Anyone who thinks they know any of this better than Apple is either wrong or wasting their lives if they are not already generating billions of dollars in value with their grand expertise."
This bears repeating. Over and over.
Basically the ultimate rebuke to every single troll nonsense comment made here since the 10th.
It's bad enough that sections of the Android community stereotype us as sheep, without you handing them a smoking gun.
I'd ask you to explain how your statement is related to what he said, but you can't.
Off-topic, I know, but the Right has more than its share of 'by/for idiots' too, starting with Fox News and the Heritage Foundation (in the US).