All these things (and even more) you describe is due to Apple's ability to design their own chips in custom detail. It makes me wonder if most competitors would find that an impossibility to do? Apple compounds their custom advantage by buying the final product is such huge quantities that their price is likely lower then an off-the-shelf chip.
Very few manufacutre have the product range that would really benefits rom going custom. That is why so many got to TI, Qualcom and others.
Another thing that Apple's SoC does is communicate with the M7/8 chip which handles all the needed finger print ID and encryption storage.
Actually I thought that was another chip, M8 does however handle a lot of sensor I/O so who knows.
Intel, like Microsoft, didn't anticipate the mobile computer/device/phone wave coming. Jobs even tried to clue Intel into fabbing the ARM chips for Apple. But the profits per ARM chip was a fraction of what Intel was getting for each x86 chip they sold. So when the wave hit, Intel and Microsoft were caught flat-footed and the scramble took until now to have a viable product to offer. Intel's now-shipping power-sipping x86 chip needs to be marketing at far less then the old x86 chips because, like this forum shows, ARM is now approaching viability as a alternative option.
ARM is getting a lot of traction in industries not even related to tablets and light laptops. Intel has shot themsleves in the foot more than one way.
So, to answer your question directly, Intel put themselves in their own weak position. However, Apple fueled a faster change to a need for mobile chips than we would have seen if Apple hadn't become a player in portable phones/tablets/computers. Imagine, for a moment, where we'd be today if Apple were not part of the equation...
How little intelligence would a phone have? Would it still have a qwerty keyboard?
Would Microsoft have ever released a tablet one could carry around? Would we still read our Amazon eBooks on our desktop computer screen?
FInally, would laptop computers still weight about what they did in the early years of this century and cost about as much too?
Considering your list:
actually I can remember open source efforts to do such a cell phone before iPhone hit the market. I can also remember Nokia just not getting a handle on software for their PDA. In other words the effort was inplace to change the world but nobody had the resources to get it right.
i consider MS to be a failure as a company so obviously can't say much about them. However reading books on a computer screen is just as viable as reading on a tablet. Really depends upon your goals but I like to have my books equally available on both devices.
certanly not, the industry trend was obvious. The only real difference here was that Apple pushed harder.
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
It has always seemed to me that Apple’s partners consistently drop the ball at critical times. Moto and IBM didn’t advance the PPC to keep up with Intel X86 development. Nvidia, ATI, Radius all screwed up on graphics cards and hurt Apple. Even today we have third party graphics causing trouble, namely the MacBook Pro lawsuits. So it would seem both logical and good business for Apple to develop critical hardware chips in house. But then the question of who does the fabrication arises. I’m sure Apple is not interested in getting into the fab business itself but maybe they will have to one day.
Apple will end up getting into Fab, because the market will force them to, the slowness of Motorola, I.B.M. and Intel has made Apple rollup it's sleeves and get dirty and in the future Apple will even more dirty.
ARM is getting a lot of traction in industries not even related to tablets and light laptops. Intel has shot themsleves in the foot more than one way.
Considering your list:
actually I can remember open source efforts to do such a cell phone before iPhone hit the market. I can also remember Nokia just not getting a handle on software for their PDA. In other words the effort was inplace to change the world but nobody had the resources to get it right.
i consider MS to be a failure as a company so obviously can't say much about them. However reading books on a computer screen is just as viable as reading on a tablet. Really depends upon your goals but I like to have my books equally available on both devices.
certanly not, the industry trend was obvious. The only real difference here was that Apple pushed harder.
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
The industry trend was ponderous, slow to act and mired in ennui when it came to thinking differently about mobility.
They would never have given up Ethernet ports in laptops had the MacBook Air not shown a thin chassis trumps a plethora of ports.
Same for removable batteries.
Even today, "the industry" continues to produce products designed for what the consumer wanted yesterday.
To say "If you build it, they will come" is cloyingly trite, but in the case of the Air it is quite true. The same is and will be true for the 64-bit chip. It may be another year before consumers see the advantage, but see it they will.
The reason why the A8X has so many transistors is because it's an SoC, which means "system on a chip". All of the support functions that would be carried by external chips are on that. In addition, as has been said, Apple includes a camera processing chip which most (or all) other SoC's don't.
We also don't know what approximately 35% of the chip area is being used for. That's possibly a billion transistors.
Isn't the fingerprint tech in there somewhere? Imagine if Apple had to wait for Intel to get it's thumb out of it's ass and produce that type of tech. In addition the new smart watch device will have a brand new in house cpu chip too (Intel need not apply).
Finally, the analyst said that while the iPad 2’s three-cell battery pack is labeled “assembled in China," this label refers to the whole battery pack and further investigation showed the battery itself was made by Apple Japan, which operates as a subsidiary of Apple.
“Typically, battery cells are made at the site of assembly but because the iPad 2’s lithium-ion polymer battery is unusually thin, it likely requires advanced battery cell manufacturing technologies that reside in Japan,” said Wayne Lam, senior analyst, competitive analysis, at IHS.
The only real difference here was that Apple pushed harder.
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
I agree! Didn't they: invent the mobile trackball, then replace that with the coupled capacitive trackpad (the only ones in the industry that don't suck), built-in batteries, all-flash designs, Retina displays, MagSafe, Thunderbolt, headphone ports that work with in-line remote/mic, backlit keyboards that actually work, laser-cut keys, scissor-action keyboard keys, industry-leading power management, instant wake, FileVault, "transparent aluminum" (i.e. laser cut holes for indicator LEDs), ambient light sensors (that also dim the sleep indicator!!!), 14 hour battery life, Target Disk Mode (SCSI/Firewire/and Thunderbolt!) and... Damn. And that's just laptops.
However, while Intel was interested in selling its new Core x86 chips to Apple for use in Macs (and developing the supporting chipsets for them), it wasn't interested in building mobile chips for Apple's iPhone, at least not at the price Apple wanted to pay and in the quantity Intel expected Apple to buy.
...
Intel's former chief executive Paul Otellini (below) revealed last year that he didn't believe his company would able to earn enough money building mobile chips for Apple's new iPhone to cover its development costs, largely because he couldn't imagine Apple selling iPhones in large quantities.
...
Intel's inability to foresee the potential of Apple's new iPhone may have been colored by its disappointing experiences with XScale, the rebranded StrongARM group it announced plans to acquire from Digital Equipment Corporation in 1997 as part of a patent infringement settlement.
This article mentions several times Intel's inability to see the iPhone selling in large quantities.
However, wouldn't this decision have been based on Apple's order for chips? Apple has admitted several times the response to the iPhone has been beyond what they could have hoped. If they would have had a bigger order, perhaps Intel would have had a better idea of how many chips Apple was really going to buy. It was a risky business to invest a huge sum of R&D for such a small original order.
Apple pre-2005 was a totally different company in terms of scale. Nobody (except for the truest of Apple believers) saw what would happen.
"Can someone explain to me why does A8 chip have 3 billion transistors compared to Intel's (Haswell' i7) 1.4 billion transistors?"
Good question - I wondered the same thing and with so many transistors packed in, how come it doesn't run hot at all compared to x86 chip of similar count?!?
The A8 runs at a far slower speed than the i7. The A8 runs up to 1.5 GHz. The i7 goes up to 4 GHZ.
The A8 uses a lower voltage and lower current than the i7. This is because it has to run off batteries.
The i7 does not have this limitation. It can use 150 watts of power - enough to seriously burn your hand.
The i7 by using so much electricity does computations FAR FASTER than the A8.
If Apple ran the A8 with as much current and speed as the i7, it would run into the same problems of increasing CPU computational speed as the i7. This is because of the limits of physics.
"Intel's former chief executive Paul Otellini... revealed last year that he didn't believe his company would [be] able to earn enough money building mobile chips for Apple's new iPhone to cover its development costs, largely because he couldn't imagine Apple selling iPhones in large quantities."
But it's amazing that AMD's newest chips are still on 32nm. 32nm!! Intels comparable part is 125 watts, while AMD's is 220 watts. So far as x86 goes, Intel is still far ahead.
We'll see what happens four, or so, years from now.
AMD's APUs are on 28nm, the FX is on 32nm as they haven't changed since Vishera [Piledriver]. Steamroller and Excavator FX are scrapped. The replacement, Zen FX is coming in 2016.
Excavator APUs Carrizos are coming this year already being tested in OEMS presently. They are being stamped out by Global Foundries 28nm.
Early 2016 second generation HSA APUs and first generation HSA FX-CPUs from project Zen at 20nm/14nm FinFET arrive that are SMT based cores and the APUs will leverage HBM memory.
Maybe my reading comprehension is not that good, but for me this aren't attacks to Intel as a chip maker, it's more about how Apple managed to get it right with their own designs. It's also about the failed business decision taken years ago when Intel was presented with the opportunity to power the original iPhone and passed it.
Given the success of the ARM, I am curious as to why Intel doesn't join the fray? I can imagine a number of reasons, but the main ones is fear of diluting their offering and branding?
At the same time where are Intel's serious competitors to the ARM's benefit of more processing capability per megawatt. Even datacenters are going to want to cut their power usage at some point, while maintaining processing capability.
AMD's APUs are on 28nm, the FX is on 32nm as they haven't changed since Vishera [Piledriver]. Steamroller and Excavator FX are scrapped. The replacement, Zen FX is coming in 2016.
Excavator APUs Carrizos are coming this year already being tested in OEMS presently. They are being stamped out by Global Foundries 28nm.
Early 2016 second generation HSA APUs and first generation HSA FX-CPUs from project Zen at 20nm/14nm FinFET arrive that are SMT based cores and the APUs will leverage HBM memory.
They're still two steps behind intel. Even moving to 28nm on some of their chips doesn't change the fact that foundry wise, and therefor chip design wise, they are years behind. I have no faith that anything AMD says about upcoming designs are going to match the actual product. They haven't, for years. I really don't care who is in charge of the company. They are circling the drain.
Given the success of the ARM, I am curious as to why Intel doesn't join the fray? I can imagine a number of reasons, but the main ones is fear of diluting their offering and branding?
At the same time where are Intel's serious competitors to the ARM's benefit of more processing capability per megawatt. Even datacenters are going to want to cut their power usage at some point, while maintaining processing capability.
It's similar to the problem Microsoft is having in offering their wares to non Windows OSs. While they had to support the Mac, for various reasons, that's not true for anything else. As long as they see this Windows product as being central to the companys future, they couldn't offer product anywhere else. That's now changed, to a certain extent. The market has opened their eyes.
This is the problem Intel has. ARM SoCs are much less expensive than any x86 design, even the Atom line, which is why they've lost about $4 billion last year in their mobile division subsidizing prices to below production cost. But they need to counteract ARM now matter what, they believe. It must be with x86 compatible chips too. Because if it isn't x86 compatable, then why stick with intel? Again, that Microsoft's problem too.
Wall Street has never given Apple stock a premium value for making the effort of developing their own chips in-house. Ax processors are considered to be no better than any other smartphone or tablet processor on the market. In fact, Wall Street and the rest of the industry sees them as being inferior because they don't run as high a clock speed or have as many cores as rival chips do. In the industry, having greater specs on paper means everything. Apple was laughed at for having the first 64-bit mobile processor and Apple has gotten no respect for having it all this time while rivals laggardly introduce their own 64-bit mobile processors. Apple gets no respect now and will likely get even less respect in the future for possibly offering other mobile products using their own Ax processors.
Despite the claims that Intel has lost about $7 billion dollars over the last few years with their mobile processors, if you lay Apple's share price and Intel's share price gains over the past 52 weeks, you'll see that Intel's share price gains are certainly greater than Apple's with Intel handily having a higher P/E ratio than Apple. That's how Wall Street investors see Apple as an inferior company even to Intel. It doesn't matter how pro-Apple people and Apple shareholders see Apple, Wall Street sees Apple in a totally different light. Apple would seem to be on fire but it makes no difference to Wall Street that sees Apple as a has-been company and always loses market share to everyone. Intel is seen as having a bright future with much potential growth and Apple not so much of either.
The main difference between Apple and Intel as an investment, is Apple gets most of it's income from the iPhone, which is very dependent on consumers continuing to prefer the iPhone. Intel keeps collecting money churning out PC CPU, which company IT Depts stick to for job security.
So naturally Intel's income is considered more secure and the reward is a higher PE.
Comments
I'm still sticking to my early last year prediction that Apple goes to piezo electric fans, which are flat, super quiet, and very power efficient.
I hadn't seen that. It could be very interesting. Thanks.
Very few manufacutre have the product range that would really benefits rom going custom. That is why so many got to TI, Qualcom and others.
Actually I thought that was another chip, M8 does however handle a lot of sensor I/O so who knows.
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
It has always seemed to me that Apple’s partners consistently drop the ball at critical times. Moto and IBM didn’t advance the PPC to keep up with Intel X86 development. Nvidia, ATI, Radius all screwed up on graphics cards and hurt Apple. Even today we have third party graphics causing trouble, namely the MacBook Pro lawsuits. So it would seem both logical and good business for Apple to develop critical hardware chips in house. But then the question of who does the fabrication arises. I’m sure Apple is not interested in getting into the fab business itself but maybe they will have to one day.
Apple will end up getting into Fab, because the market will force them to, the slowness of Motorola, I.B.M. and Intel has made Apple rollup it's sleeves and get dirty and in the future Apple will even more dirty.
ARM is getting a lot of traction in industries not even related to tablets and light laptops. Intel has shot themsleves in the foot more than one way.
Considering your list:
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
The industry trend was ponderous, slow to act and mired in ennui when it came to thinking differently about mobility.
They would never have given up Ethernet ports in laptops had the MacBook Air not shown a thin chassis trumps a plethora of ports.
Same for removable batteries.
Even today, "the industry" continues to produce products designed for what the consumer wanted yesterday.
To say "If you build it, they will come" is cloyingly trite, but in the case of the Air it is quite true. The same is and will be true for the 64-bit chip. It may be another year before consumers see the advantage, but see it they will.
The reason why the A8X has so many transistors is because it's an SoC, which means "system on a chip". All of the support functions that would be carried by external chips are on that. In addition, as has been said, Apple includes a camera processing chip which most (or all) other SoC's don't.
We also don't know what approximately 35% of the chip area is being used for. That's possibly a billion transistors.
Isn't the fingerprint tech in there somewhere? Imagine if Apple had to wait for Intel to get it's thumb out of it's ass and produce that type of tech. In addition the new smart watch device will have a brand new in house cpu chip too (Intel need not apply).
i wish they would get into the battery business.
Actually, they are in the battery business!
http://appleinsider.com/articles/11/03/28/japan_disaster_causes_shortage_of_lithium_ion_batteries_for_apple
“Typically, battery cells are made at the site of assembly but because the iPad 2’s lithium-ion polymer battery is unusually thin, it likely requires advanced battery cell manufacturing technologies that reside in Japan,” said Wayne Lam, senior analyst, competitive analysis, at IHS.
I love my Apple products but I think we need to be careful and not give them more credit than they have actually earned.
I agree! Didn't they: invent the mobile trackball, then replace that with the coupled capacitive trackpad (the only ones in the industry that don't suck), built-in batteries, all-flash designs, Retina displays, MagSafe, Thunderbolt, headphone ports that work with in-line remote/mic, backlit keyboards that actually work, laser-cut keys, scissor-action keyboard keys, industry-leading power management, instant wake, FileVault, "transparent aluminum" (i.e. laser cut holes for indicator LEDs), ambient light sensors (that also dim the sleep indicator!!!), 14 hour battery life, Target Disk Mode (SCSI/Firewire/and Thunderbolt!) and... Damn. And that's just laptops.
However, while Intel was interested in selling its new Core x86 chips to Apple for use in Macs (and developing the supporting chipsets for them), it wasn't interested in building mobile chips for Apple's iPhone, at least not at the price Apple wanted to pay and in the quantity Intel expected Apple to buy.
...
Intel's former chief executive Paul Otellini (below) revealed last year that he didn't believe his company would able to earn enough money building mobile chips for Apple's new iPhone to cover its development costs, largely because he couldn't imagine Apple selling iPhones in large quantities.
...
Intel's inability to foresee the potential of Apple's new iPhone may have been colored by its disappointing experiences with XScale, the rebranded StrongARM group it announced plans to acquire from Digital Equipment Corporation in 1997 as part of a patent infringement settlement.
This article mentions several times Intel's inability to see the iPhone selling in large quantities.
However, wouldn't this decision have been based on Apple's order for chips? Apple has admitted several times the response to the iPhone has been beyond what they could have hoped. If they would have had a bigger order, perhaps Intel would have had a better idea of how many chips Apple was really going to buy. It was a risky business to invest a huge sum of R&D for such a small original order.
Apple pre-2005 was a totally different company in terms of scale. Nobody (except for the truest of Apple believers) saw what would happen.
Can someone explain to me why does A8 chip have 3 billion transistors compared to Intel's (Haswell' i7) 1.4 billion transistors?
NOTE that the A8 Chip has 2 Billion transistors so is closer to the Intel i7 chip in transistor count. It is the A8X which has 3 billion transistors.
The A8 chip and the A8X do far more specialized processing than Intel's i7 chip in hardware.
For example
1. The A8/A8X has hardware routines to analyze the sensor and touch information in 3 dimensions. Haswell does not.
2. The A8/A8X has a TrustZone component that does the computation, encryption and storage for TouchID. Haswell does not.
3. The A8/A8X has a large SRAM Cache Memory in addition to each core's cache memory to reduce fetches from RAM
4. The A8/A8X has specialized hardware routines for mapping, image processing, etc. Haswell does not.
Exactly! This is why I've been saying for months now that the SoC is the printed circuit board of the 1980's.
Very few manufacutre have the product range that would really benefits rom going custom. That is why so many got to TI, Qualcom and others.
Actually I thought that was another chip, M8 does however handle a lot of sensor I/O so who knows.
The M8 handles sensor I/0. But the A8 handles the processing of the information when the iPhone or iPad is turned on.
"Can someone explain to me why does A8 chip have 3 billion transistors compared to Intel's (Haswell' i7) 1.4 billion transistors?"
Good question - I wondered the same thing and with so many transistors packed in, how come it doesn't run hot at all compared to x86 chip of similar count?!?
The A8 runs at a far slower speed than the i7. The A8 runs up to 1.5 GHz. The i7 goes up to 4 GHZ.
The A8 uses a lower voltage and lower current than the i7. This is because it has to run off batteries.
The i7 does not have this limitation. It can use 150 watts of power - enough to seriously burn your hand.
The i7 by using so much electricity does computations FAR FASTER than the A8.
If Apple ran the A8 with as much current and speed as the i7, it would run into the same problems of increasing CPU computational speed as the i7. This is because of the limits of physics.
Ouch!
AMD's APUs are on 28nm, the FX is on 32nm as they haven't changed since Vishera [Piledriver]. Steamroller and Excavator FX are scrapped. The replacement, Zen FX is coming in 2016.
Excavator APUs Carrizos are coming this year already being tested in OEMS presently. They are being stamped out by Global Foundries 28nm.
http://www.sisoftware.eu/rank2011d/show_system.php?q=cea598aa98af9ba391b7d0edc0f1d7a598a88ee7daebcda598ad8bf3ceffd9bcd9e4d4f281bc84&l=en
Early 2016 second generation HSA APUs and first generation HSA FX-CPUs from project Zen at 20nm/14nm FinFET arrive that are SMT based cores and the APUs will leverage HBM memory.
It's also about the failed business decision taken years ago when Intel was presented with the opportunity to power the original iPhone and passed it.
At the same time where are Intel's serious competitors to the ARM's benefit of more processing capability per megawatt. Even datacenters are going to want to cut their power usage at some point, while maintaining processing capability.
They're still two steps behind intel. Even moving to 28nm on some of their chips doesn't change the fact that foundry wise, and therefor chip design wise, they are years behind. I have no faith that anything AMD says about upcoming designs are going to match the actual product. They haven't, for years. I really don't care who is in charge of the company. They are circling the drain.
It's similar to the problem Microsoft is having in offering their wares to non Windows OSs. While they had to support the Mac, for various reasons, that's not true for anything else. As long as they see this Windows product as being central to the companys future, they couldn't offer product anywhere else. That's now changed, to a certain extent. The market has opened their eyes.
This is the problem Intel has. ARM SoCs are much less expensive than any x86 design, even the Atom line, which is why they've lost about $4 billion last year in their mobile division subsidizing prices to below production cost. But they need to counteract ARM now matter what, they believe. It must be with x86 compatible chips too. Because if it isn't x86 compatable, then why stick with intel? Again, that Microsoft's problem too.
Superb history of the mobile processor, Mr. Dilger.
The only thing I would say is that I would like to see a deeper look into the history of ARM and how it might tie in with UNIX and OS X pre-Apple.
Wall Street has never given Apple stock a premium value for making the effort of developing their own chips in-house. Ax processors are considered to be no better than any other smartphone or tablet processor on the market. In fact, Wall Street and the rest of the industry sees them as being inferior because they don't run as high a clock speed or have as many cores as rival chips do. In the industry, having greater specs on paper means everything. Apple was laughed at for having the first 64-bit mobile processor and Apple has gotten no respect for having it all this time while rivals laggardly introduce their own 64-bit mobile processors. Apple gets no respect now and will likely get even less respect in the future for possibly offering other mobile products using their own Ax processors.
Despite the claims that Intel has lost about $7 billion dollars over the last few years with their mobile processors, if you lay Apple's share price and Intel's share price gains over the past 52 weeks, you'll see that Intel's share price gains are certainly greater than Apple's with Intel handily having a higher P/E ratio than Apple. That's how Wall Street investors see Apple as an inferior company even to Intel. It doesn't matter how pro-Apple people and Apple shareholders see Apple, Wall Street sees Apple in a totally different light. Apple would seem to be on fire but it makes no difference to Wall Street that sees Apple as a has-been company and always loses market share to everyone. Intel is seen as having a bright future with much potential growth and Apple not so much of either.
The main difference between Apple and Intel as an investment, is Apple gets most of it's income from the iPhone, which is very dependent on consumers continuing to prefer the iPhone. Intel keeps collecting money churning out PC CPU, which company IT Depts stick to for job security.
So naturally Intel's income is considered more secure and the reward is a higher PE.