Yep, and he is not the only one. Not until there is a teardown by a reputable entity announcing the Samsung-manufactured A9 should anyone believe that Samsung won. That company lies and will steal at every turn in an attempt to make itself look good. Thankfully, Apple has been able to raise the bar high enough and move forward fast enough to blunt the lies and thefts of that shameless company.
So true Samsung has a different lower of ethics than Apple and most others.
Combine that with underhanded Android and you have a dumb phone.
Apple is believed to have begun solidifying partners for its anticipated "A9" processor in late 2014, ahead of anticipated September 2015 launch of a new "iPhone 6s" series.
Rumors have suggested the "A9" chip might use a 14-nanometer manufacturing process. Smaller processors are more efficient, resulting in power savings that can allow devices like the iPhone to see improved battery life, even as the chips themselves become more powerful and capable.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
Intel announced it will delay the 10nm process until the middle of 2017. That is NOT a good sign.
Intel is delaying a lot of chips lately and I am suspecting that Apple is losing its patience and will probably release a Mac with its ARM chips which has shown to be extremely efficient and powerful.
Oh come on! Apple doesn't lose its patience and veer off a well-thought-out plan. I'm sure every plan Apple is following have many contingency options, but to make a non-intel Mac just doesn't even make sense.
We may well see touch iPads ship with detachable keyboards but they won't carry the Mac moniker.
Apple is believed to have begun solidifying partners for its anticipated "A9" processor in late 2014, ahead of anticipated September 2015 launch of a new "iPhone 6s" series.
Rumors have suggested the "A9" chip might use a 14-nanometer manufacturing process. Smaller processors are more efficient, resulting in power savings that can allow devices like the iPhone to see improved battery life, even as the chips themselves become more powerful and capable.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
It's been my understanding that the CPU clock speed is not as much an issue as the need for more parallel GPU capability. The energy saving of thinner manufacturing processing allows adding more GPU capacity without overheating the A9. Whatever Apple does will keep the HW and SW balanced as well as keeping the major elements of the HW balanced for maximum performance of CPU and GPU, along with heat buildup and battery load...
Every time the manufacturing process allows for a smaller/thinner transistor Apple seems to reexamines how to best use the extra capability for the best user experience.
In chip technology, hype is part of life. Intel's bread and butter is designing,making and selling next generation chips from desktop/laptop to servers to internet of connectivity so Intel will make sure it stays ahead of game. IBM just announced breakthrough for 7nm width.
Apple is making TSMC and Samsung to compete for it's business and in process make them invest billions to push the chip technology up. But, all this valid when their first chip comes out of foundry working at 10nm node. So, until than take with the grain of salt. This means iphone 7 processor A10 will not be on 10nm in 2016.
Just to clarify, Samsung/Global Foundries are partners with IBM on 7nm fab.
You both are correct in what you say.
Most desktops and laptops can easily do most of what most people need done. The pressure now is to make mobile computing even more mobile by lowering weight and battery consumption.
Intel's process is more space efficient. Both Samsung and TSMC use two different process nodes on the same chip. The process for the transistors is the new node, but the wiring between the chips is a half, or a full node behind.
Considering how much behind everyone was 10 years ago, I think Intel SHOULD be worried. Process advantage was they're big ace in the hole and its now fading. They charge a lot more for they're processors. Being one shrink ahead may not be worth it for many users eventually.
It's been my understanding that the CPU clock speed is not as much an issue as the need for more parallel GPU capability. The energy saving of thinner manufacturing processing allows adding more GPU capacity without overheating the A9. Whatever Apple does will keep the HW and SW balanced as well as keeping the major elements of the HW balanced for maximum performance of CPU and GPU, along with heat buildup and battery load...
Every time the manufacturing process allows for a smaller/thinner transistor Apple seems to reexamines how to best use the extra capability for the best user experience.
One thing people don't notice is how much of the SOC is used to process image/video outside the GPU. That's why Apple's camera despite losing on absolute picture quality to the S6, is still the best all rounder. Being to handle the data from the sensors quickly is part of the photo/video experience.
Intel announced it will delay the 10nm process until the middle of 2017. That is NOT a good sign.
Intel is delaying a lot of chips lately and I am suspecting that Apple is losing its patience and will probably release a Mac with its ARM chips which has shown to be extremely efficient and powerful.
That's brain dead. Apple would more likely release an APU or wait for Zen from AMD than go with an ARM set up.
Eh? TSMC is fabbing millions of 20nm chips for Apple every month. That's why nVidia and AMD don't have new GPU's; Apple bought out the available capacity.
That's Low Power only. If you know anything about Fabs you know 20nm LPE won't work in CPUs or GPGPUs.
Samsung is 2 years ahead of tsmc and has already produced and tested 10nm.
They are about to release there full epop soc in the note 5 that has everything in the soc.
GPU,CPU,modem,ram and nand all on a 14nm chip.
Even Intel has no answer to the 14nm exynos 7420 in the gs6 and the note 5 will take them to a full 3 years ahead of anyone one else.
It's funny how apple teases about its hardware and software design videos and forgets Samsung spends 20 billion a year on fab r&d and has liscenced its 14nm to global industries and is working with IBM pushing 7nm forward.
They leaked exynos 10nm exynos soc scored 2300 single scrore in geek bench and I'm pretty sure Intel is starting to worry about Samsung in the server market too
Intel announced it will delay the 10nm process until the middle of 2017. That is NOT a good sign.
Intel is delaying a lot of chips lately and I am suspecting that Apple is losing its patience and will probably release a Mac with its ARM chips which has shown to be extremely efficient and powerful.
I...don't think so.
While ARM chips are becoming more powerful more quickly than x86 chips are, even Apple's A8x just barely makes it to the bottom of the x86 line, just above the low end Atom.
While I have hopes that at some point, Apple's chips can compete on the low end with OS X, it will never replace a medium or high end x86 chip.
Eh? TSMC is fabbing millions of 20nm chips for Apple every month. That's why nVidia and AMD don't have new GPU's; Apple bought out the available capacity.
This is not entirely true. Apple attempted to pay TSMC for the entire production from at least one fab, but TSMC declined. That was pretty smart of TSMC, unless they lose too many sales from Apple directly as a result, which is something we won't know, unless in their financial report they make an oblique reference to it.
These companies don't want to become dependent on one major customer. More often than people know, a company will actually turn business away from a large customer rather than get trapped by that customer's demands.
With TSMC being the world's largest foundry, that is, an independent manufacturer of chips for other companies, rather than designs of their own, they have some leeway in what they do. Smaller foundries don't.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
The battery life of the 6 series is really quite good, particularly the 6+. So while more battery life would be nice, it's certainly not a necessity. More processing power, on the other hand, is always welcomed.
In addition to that, there is the fact that you can always buy a battery case to double the life of the charge, or an external battery to quadruple the life of the charge, if you really need it. But there is no way to improve upon the processing power, so that is more important to most people.
Considering how much behind everyone was 10 years ago, I think Intel SHOULD be worried. Process advantage was they're big ace in the hole and its now fading. They charge a lot more for they're processors. Being one shrink ahead may not be worth it for many users eventually.
I agree completely. I've been warning about that very problem. Once 10nm is reached, there will be years for everyone to catch up in process tech until 7 nm is achieved. And then, well, who knows? Experts in that field still think that 5nm may not be possible. If that's true, and I believe it is, despite Intel showing 5nm on its roadmap, which I think they're doing for marketing and financial reasons, then until a new technology is ready, we won't see major advances for some time. Everyone who has a good process technology will be on an even footing, and it will just be efficiency of production, and yield that will separate the good from the bad. And there, Intel holds the lead now.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
The battery life of the 6 series is really quite good, particularly the 6+. So while more battery life would be nice, it's certainly not a necessity. More processing power, on the other hand, is always welcomed.
In addition to that, there is the fact that you can always buy a battery case to double the life of the charge, or an external battery to quadruple the life of the charge, if you really need it. But there is no way to improve upon the processing power, so that is more important to most people.
I had an iPhone 4, which had great battery life and after 4.5 years still about 50% of the original life.
My current iPhone5c has lots of power, but it's battery life is definitely less than the 4 was when new. Perhaps 65% of the 4.
I've considered getting a case which doubles the battery life, but that spoils the thin light form factor.
I had an iPhone 4, which had great battery life and after 4.5 years still about 50% of the original life.
My current iPhone5c has lots of power, but it's battery life is definitely less than the 4 was when new. Perhaps 65% of the 4.
I've considered getting a case which doubles the battery life, but that spoils the thin light form factor.
"Beat Intel"...In marketing terms to laymen, maybe. Other fab plants processes are almost always bigger than when Intel advertises the same nanometers. And Intel uses the same process for front end of line and back end of line transistors, while other fabs only use the smallest transistors for front end of line, and use the previous generation transistors for back end. Meaning Intels transistors are not only smaller and more efficient, but they use the smallest ones more, whereas other fabs have a mix of their smallest and larger transistors.
People with a bit of technical knowledge like the author keep looking at the numbers on paper and thinking Intels lead is shrinking. It has, but not as much as you would think with just the numbers. I'd say they still have about an 18 month lead right now. Even with the 10nm delay in the future, If Intel, with their huge R&D budget is having problems with the physics getting funky down at that size, I'm sure everyone else will too.
Or, they'll just advertise something equivalent to Intels 14nm process as a "10nm" process at that time. Like they always have.
Comments
So, do you think TSMC will be making the A9 then?
Yep, and he is not the only one. Not until there is a teardown by a reputable entity announcing the Samsung-manufactured A9 should anyone believe that Samsung won. That company lies and will steal at every turn in an attempt to make itself look good. Thankfully, Apple has been able to raise the bar high enough and move forward fast enough to blunt the lies and thefts of that shameless company.
So true Samsung has a different lower of ethics than Apple and most others.
Combine that with underhanded Android and you have a dumb phone.
Apple is believed to have begun solidifying partners for its anticipated "A9" processor in late 2014, ahead of anticipated September 2015 launch of a new "iPhone 6s" series.
Rumors have suggested the "A9" chip might use a 14-nanometer manufacturing process. Smaller processors are more efficient, resulting in power savings that can allow devices like the iPhone to see improved battery life, even as the chips themselves become more powerful and capable.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
(You mean feature size?) Intel 14nm Apple (ARM) 22nm.
The already huge gap between ARM and Intel will widen even further.
Intel will eventually lose the whole chip market because the near future will hold tablets only.
Utter nonsense.
How many iPads do you see?
Oh come on! Apple doesn't lose its patience and veer off a well-thought-out plan. I'm sure every plan Apple is following have many contingency options, but to make a non-intel Mac just doesn't even make sense.
We may well see touch iPads ship with detachable keyboards but they won't carry the Mac moniker.
It's been my understanding that the CPU clock speed is not as much an issue as the need for more parallel GPU capability. The energy saving of thinner manufacturing processing allows adding more GPU capacity without overheating the A9. Whatever Apple does will keep the HW and SW balanced as well as keeping the major elements of the HW balanced for maximum performance of CPU and GPU, along with heat buildup and battery load...
Every time the manufacturing process allows for a smaller/thinner transistor Apple seems to reexamines how to best use the extra capability for the best user experience.
You both are correct in what you say.
Most desktops and laptops can easily do most of what most people need done. The pressure now is to make mobile computing even more mobile by lowering weight and battery consumption.
Intel's process is more space efficient. Both Samsung and TSMC use two different process nodes on the same chip. The process for the transistors is the new node, but the wiring between the chips is a half, or a full node behind.
Considering how much behind everyone was 10 years ago, I think Intel SHOULD be worried. Process advantage was they're big ace in the hole and its now fading. They charge a lot more for they're processors. Being one shrink ahead may not be worth it for many users eventually.
It's been my understanding that the CPU clock speed is not as much an issue as the need for more parallel GPU capability. The energy saving of thinner manufacturing processing allows adding more GPU capacity without overheating the A9. Whatever Apple does will keep the HW and SW balanced as well as keeping the major elements of the HW balanced for maximum performance of CPU and GPU, along with heat buildup and battery load...
Every time the manufacturing process allows for a smaller/thinner transistor Apple seems to reexamines how to best use the extra capability for the best user experience.
One thing people don't notice is how much of the SOC is used to process image/video outside the GPU. That's why Apple's camera despite losing on absolute picture quality to the S6, is still the best all rounder. Being to handle the data from the sensors quickly is part of the photo/video experience.
That's brain dead. Apple would more likely release an APU or wait for Zen from AMD than go with an ARM set up.
That's Low Power only. If you know anything about Fabs you know 20nm LPE won't work in CPUs or GPGPUs.
That's Low Power only. If you know anything about Fabs you know 20nm LPE won't work in CPUs or GPGPUs.
You also forgot to mention the massive size difference(transistors and thus area) between an AX chip and a high end GPU.
They are about to release there full epop soc in the note 5 that has everything in the soc.
GPU,CPU,modem,ram and nand all on a 14nm chip.
Even Intel has no answer to the 14nm exynos 7420 in the gs6 and the note 5 will take them to a full 3 years ahead of anyone one else.
It's funny how apple teases about its hardware and software design videos and forgets Samsung spends 20 billion a year on fab r&d and has liscenced its 14nm to global industries and is working with IBM pushing 7nm forward.
They leaked exynos 10nm exynos soc scored 2300 single scrore in geek bench and I'm pretty sure Intel is starting to worry about Samsung in the server market too
I...don't think so.
While ARM chips are becoming more powerful more quickly than x86 chips are, even Apple's A8x just barely makes it to the bottom of the x86 line, just above the low end Atom.
While I have hopes that at some point, Apple's chips can compete on the low end with OS X, it will never replace a medium or high end x86 chip.
This is not entirely true. Apple attempted to pay TSMC for the entire production from at least one fab, but TSMC declined. That was pretty smart of TSMC, unless they lose too many sales from Apple directly as a result, which is something we won't know, unless in their financial report they make an oblique reference to it.
These companies don't want to become dependent on one major customer. More often than people know, a company will actually turn business away from a large customer rather than get trapped by that customer's demands.
With TSMC being the world's largest foundry, that is, an independent manufacturer of chips for other companies, rather than designs of their own, they have some leeway in what they do. Smaller foundries don't.
The battery life of the 6 series is really quite good, particularly the 6+. So while more battery life would be nice, it's certainly not a necessity. More processing power, on the other hand, is always welcomed.
In addition to that, there is the fact that you can always buy a battery case to double the life of the charge, or an external battery to quadruple the life of the charge, if you really need it. But there is no way to improve upon the processing power, so that is more important to most people.
I agree completely. I've been warning about that very problem. Once 10nm is reached, there will be years for everyone to catch up in process tech until 7 nm is achieved. And then, well, who knows? Experts in that field still think that 5nm may not be possible. If that's true, and I believe it is, despite Intel showing 5nm on its roadmap, which I think they're doing for marketing and financial reasons, then until a new technology is ready, we won't see major advances for some time. Everyone who has a good process technology will be on an even footing, and it will just be efficiency of production, and yield that will separate the good from the bad. And there, Intel holds the lead now.
Hopefully Apple doesn't use up that increased efficiency by pushing the A9 chip even faster for more processing power.
It would be nice to have a slightly slower than max speed A9 model, giving longer battery life than current iPhones.
The battery life of the 6 series is really quite good, particularly the 6+. So while more battery life would be nice, it's certainly not a necessity. More processing power, on the other hand, is always welcomed.
In addition to that, there is the fact that you can always buy a battery case to double the life of the charge, or an external battery to quadruple the life of the charge, if you really need it. But there is no way to improve upon the processing power, so that is more important to most people.
I had an iPhone 4, which had great battery life and after 4.5 years still about 50% of the original life.
My current iPhone5c has lots of power, but it's battery life is definitely less than the 4 was when new. Perhaps 65% of the 4.
I've considered getting a case which doubles the battery life, but that spoils the thin light form factor.
Why not just replace the battery?
I had an iPhone 4, which had great battery life and after 4.5 years still about 50% of the original life.
My current iPhone5c has lots of power, but it's battery life is definitely less than the 4 was when new. Perhaps 65% of the 4.
I've considered getting a case which doubles the battery life, but that spoils the thin light form factor.
Why not just replace the battery?
Because the new 5c is only one week old, that battery life is obviously it's norm.
"Beat Intel"...In marketing terms to laymen, maybe. Other fab plants processes are almost always bigger than when Intel advertises the same nanometers. And Intel uses the same process for front end of line and back end of line transistors, while other fabs only use the smallest transistors for front end of line, and use the previous generation transistors for back end. Meaning Intels transistors are not only smaller and more efficient, but they use the smallest ones more, whereas other fabs have a mix of their smallest and larger transistors.
See here
http://www.extremetech.com/wp-content/uploads/2014/12/Cell-SizeComparison.png ;
People with a bit of technical knowledge like the author keep looking at the numbers on paper and thinking Intels lead is shrinking. It has, but not as much as you would think with just the numbers. I'd say they still have about an 18 month lead right now. Even with the 10nm delay in the future, If Intel, with their huge R&D budget is having problems with the physics getting funky down at that size, I'm sure everyone else will too.
Or, they'll just advertise something equivalent to Intels 14nm process as a "10nm" process at that time. Like they always have.