Personally I’m more impressed with the silicon quantum chips IBM and Intel are developing.
There are tons of real physical challenges to quantum computing. One of them is the extremely short-lived and unstable lifetime of operation. There yet has to be found a solution for error correction for the mess that comes out of it. We’re years away from a commercial product - maybe decades.
Kind of like self-driving cars. It's gonna take a lot longer to reach practical reality than people think.
I think this is simplifying a bit too much. If Microsoft was truly dominant, they would have successfully moved to ARM years ago. Instead, the dominance is the combination of Windows + x86, and while Microsoft might be influencing it they're not really controlling it. Maybe in the future, once the Qualcomm deal ends, if Microsoft puts more might into ARM.
Microsoft also doesn’t have control over developers in the same way as Apple. Apple has always been very aggressive at deprecating and removing old technologies while Microsoft has attempted to maintain compatibility for as long as possible.
Thus a lot of successful Windows apps have very old code bases and therefore wouldn’t be compatible with a Microsoft’s equivalent of Rosetta 2. And without Rosetta 2, Apple’s migration to ARM would have been dead in the water.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
I think having your own operating system is a part of this equation.
Apple languished for decades under the "control" of Motorola, IBM, Intel & others. Jobs coming back with NeXT & helping push the end of the "Classic" Macintosh operating system is not to be underestimated. In the same way having a decade or so to see the 68k to PowerPC switch as well as the PowerPC to Intel switch means that Apple as a company had a handle on the intricacies of large platform switches & maintaining support for older software along the way.
I think you've hit on the key reasons why Apple viewed the majority of the Apple Silicon transition as a manageable risk. Unlike Microsoft they didn't have to worry about the impact their change would have on a huge base of hardware OEMs and partners. They also had successful experience with mitigating the application layer to operating system layer dependencies on the software side using their own technology like Rosetta to allow for a smoother transition. There were obviously some aspects of the migration that they viewed as challenging, but I don't think they were fearful of any big unknowns that could prove to be show stoppers.
The greatest risk for Apple, in my opinion, had nothing at all to do with the traditional platform change challenges that they faced with PPC and Intel. The greatest challenge for Apple was to secure a world-class chip fabrication partner like TSMC who they could rely on over an extended period of time to stay on the leading edge and deliver in sufficient quantity and quality on very tight schedules. If Apple didn't have TSMC to depend on they'd be taking on a far greater challenge regardless of their own internal chip design prowess. So even though Apple has for the most part weaned themselves off of Intel and some other chipset makers, they aren't absolutely self-sufficient and still need to play nicely with critical partners to sustain the hardware side of their business.
From a risk management perspective, Apple has not removed risk entirely, they've only moved the risk to a different place/level. Hopefully, world events will not unfold in a way that impacts their risk before it can be properly mitigated through domestic investment in the critical technologies and manufacturing capabilities that Apple is dependent on today from TSMC. Apple is not alone in this regard, but they are very highly exposed. We as a nation should all recognize that this risk needs to be addressed and mitigated for the sake of much more than Apple and its silicon strategy. It's an existential national risk to US security and prosperity.
Apple should have added AMD models, not used souped-up phone chips in computers.
Actually you are saying: Even when Apple is using phone chips in their computers, they are still able to beat the computers with "real computer chips" from the competition
Apple should have added AMD models, not used souped-up phone chips in computers.
So the primary reason to switch to in-house development — independence from other companies — should be remedied by switching to another second-party architecture?
Personally I’m more impressed with the silicon quantum chips IBM and Intel are developing.
And yet Apple is CURRENTLY FIELDING their chips, and aggressively iterating and improving upon them. By the time their mythical quantum computing chips actually function, I'd imagine Apple will have surpassed them, too.
Apple Silicon could not compete with quantum chips. Quantum tech is a whole new ball game and Apple are not in it. Quantum chips in Apple product would be yet another transition. Maybe Quantum tech is the replacement for Apple Silicon?
And quantum chips can't compete with Apple Silicon as these can't be mass produced for consumer user so what's your point?
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Personally I’m more impressed with the silicon quantum chips IBM and Intel are developing.
And yet Apple is CURRENTLY FIELDING their chips, and aggressively iterating and improving upon them. By the time their mythical quantum computing chips actually function, I'd imagine Apple will have surpassed them, too.
Apple Silicon could not compete with quantum chips. Quantum tech is a whole new ball game and Apple are not in it. Quantum chips in Apple product would be yet another transition. Maybe Quantum tech is the replacement for Apple Silicon?
a) your assumption that Apple isn’t closely following development in that field, or perhaps conducting research of their own, is based upon what?
b) whatever advantages are there for the product ranges Apple chooses to offer are still decades out. You’re not going to see quantum chips in a music production laptop, let alone a cellphone, anytime soon.
Apple is far from established and is coming up light on its promises. It has a significant problem & it isn’t the ISA. Until product managers allow significant forks for macOS-specific frameworks (Accelerate & Metal/TBDR) and commit to UMA architecture, as Serif has, Apple Silicon will struggle to demonstrate its value. It demands a mindset shift away from CPU OR GPU benchmarks to APU or Application benchmarks & preferably not those from an x86 vendor like Puget Systems.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Just a toe dip as that design was aimed at business customers in China but Apple wasn't alone in its thinking here.
Pray tell, where is Huawei going to fab its 5nm designs? I see a lot of propaganda about China building its own 5nm lithography machines, which I don't believe exist at all in production, so color me more than skeptical.
More to the point, nothing that Huawei has done is anything close to the disruptive architecture that Apple has created for M series silicon. It is just more of the same PC main board architecture, but now with more SOC!
The first is that Kirin 9006C is a chip previously produced by TSMC ;
There is one thing to say that the appearance of the Kirin 9006C at this time is really a bit abrupt. After all, after the Huawei 9000 chip, Huawei has not launched a new Kirin chip, and there is currently no foundry that can produce 5nm chips for Huawei.
Given that Huawei's current chip inventory is still a mystery, it is not clear how many Kirin chips Huawei has, and how many Kirin chips are in stock. Therefore, the Kirin 9006C is likely to be a chip produced by TSMC and Huawei during the cooperation.
Apple should have added AMD models, not used souped-up phone chips in computers.
That would have worked well for the Mac Pro. And maybe for the iMac (particularly if they wanted to make another iMac Pro). But it wouldn't have worked for anything else.
He's just trolling. If they had gone with AMD Apple would be worse off and he'd be making some other claim about what they should've done to be a more successful company. He knows full well that he's posting stupid comments.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Just a toe dip as that design was aimed at business customers in China but Apple wasn't alone in its thinking here.
Pray tell, where is Huawei going to fab its 5nm designs? I see a lot of propaganda about China building its own 5nm lithography machines, which I don't believe exist at all in production, so color me more than skeptical.
More to the point, nothing that Huawei has done is anything close to the disruptive architecture that Apple has created for M series silicon. It is just more of the same PC main board architecture, but now with more SOC!
The first is that Kirin 9006C is a chip previously produced by TSMC ;
There is one thing to say that the appearance of the Kirin 9006C at this time is really a bit abrupt. After all, after the Huawei 9000 chip, Huawei has not launched a new Kirin chip, and there is currently no foundry that can produce 5nm chips for Huawei.
Given that Huawei's current chip inventory is still a mystery, it is not clear how many Kirin chips Huawei has, and how many Kirin chips are in stock. Therefore, the Kirin 9006C is likely to be a chip produced by TSMC and Huawei during the cooperation.
Barking up distant trees as usual.
I was challenging the OP's notion that only Apple was doing the homegrown thing on desktops and laptops.
I wasn't talking about fabrication.
Two clearly different things.
I could have gone further but refrained from doing so and then you go and blow it with your 'disruptive Apple Silicon architecture claims'. So allow me to put your feet back on the ground.
There is zero disruptive going on here. Zero.
Apple has moved back to a non-Intel solution to processing. That is it. The only real difference with the old days is that they can pull all the strings now and don't have to depend on the likes of IBM or Motorola for product road maps. That is it at the moment.
So much for disruption!
Is it producing communications hardware or developing the technologies that go into that? Nope.
Is it producing high end cluster inference systems? Nope.
Personally I’m more impressed with the silicon quantum chips IBM and Intel are developing.
And yet Apple is CURRENTLY FIELDING their chips, and aggressively iterating and improving upon them. By the time their mythical quantum computing chips actually function, I'd imagine Apple will have surpassed them, too.
Apple Silicon could not compete with quantum chips. Quantum tech is a whole new ball game and Apple are not in it. Quantum chips in Apple product would be yet another transition. Maybe Quantum tech is the replacement for Apple Silicon?
And quantum chips can't compete with Apple Silicon as these can't be mass produced for consumer user so what's your point?
You have to understand that people like @Manwithnoname will always come up with something to throw shade at Apple. Quantum SOCs are likely two decades away from the consumer market but they prove Apple’s SOCs are crap. In twenty years we don’t even know if consumer computing will be with personal devices and not just terminals. And then we have @AvonB7 coming in with his usual “Apple wasn’t first” bullshit. In the real world Apple is shipping consumer and pro products with their in house designed SOCs around the world.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Apple is the only one that could do it. Remember that Hp has had a long, and proud, history of designing and making it’s own chips for its minicomputers. But when most everything went to Windows, the chips had to be compatible with the it, so since Microsoft declined to support anything other than x86, there was nothing they could do.
Apple, with its own OS, didn’t have to worry about that. They could support any chip they wanted to, as long as they could get third party developers to go along.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Just a toe dip as that design was aimed at business customers in China but Apple wasn't alone in its thinking here.
Pray tell, where is Huawei going to fab its 5nm designs? I see a lot of propaganda about China building its own 5nm lithography machines, which I don't believe exist at all in production, so color me more than skeptical.
More to the point, nothing that Huawei has done is anything close to the disruptive architecture that Apple has created for M series silicon. It is just more of the same PC main board architecture, but now with more SOC!
The first is that Kirin 9006C is a chip previously produced by TSMC ;
There is one thing to say that the appearance of the Kirin 9006C at this time is really a bit abrupt. After all, after the Huawei 9000 chip, Huawei has not launched a new Kirin chip, and there is currently no foundry that can produce 5nm chips for Huawei.
Given that Huawei's current chip inventory is still a mystery, it is not clear how many Kirin chips Huawei has, and how many Kirin chips are in stock. Therefore, the Kirin 9006C is likely to be a chip produced by TSMC and Huawei during the cooperation.
Barking up distant trees as usual.
I was challenging the OP's notion that only Apple was doing the homegrown thing on desktops and laptops.
I wasn't talking about fabrication.
Two clearly different things.
I could have gone further but refrained from doing so and then you go and blow it with your 'disruptive Apple Silicon architecture claims'. So allow me to put your feet back on the ground.
There is zero disruptive going on here. Zero.
Apple has moved back to a non-Intel solution to processing. That is it. The only real difference with the old days is that they can pull all the strings now and don't have to depend on the likes of IBM or Motorola for product road maps. That is it at the moment.
So much for disruption!
Is it producing communications hardware or developing the technologies that go into that? Nope.
Is it producing high end cluster inference systems? Nope.
Laptops and desktops. That is it. There is no disruption to be found there. Zero.
Maybe Huawei should actually focus on a few niches, instead of attempting to be all things to all industries, after they got whacked in the handset market.
From the looks of it Huawei is making great strides in the surveillance industry, so there's that.
Oh, and as far as Apple Silicon being disruptive, yeah, it actually is.
Never before has a company’s entire product range, from tablet to all-in-one, been spread across a single processor. It is simply unimaginable in the PC market, which has instead built its complexity on a million different variations and declinations. Every year when Intel presents new versions of its processors, it does so with dozens and dozens of variations, designed on the basis of the activations of different areas of the chip, the number of cores and cores, the range of memory. An incredible puzzle made up of extremely articulated price lists, and with product ranges (i3, i5, i7, i9) with different consumption and frequencies, of which one wonders what actually are the differences that justify different price points and which should guide consumer choice on the basis of presumed differences in performance.
Out of all the desktop/laptop OEMs out there Apple was the only one with the balls to strike out on its own. As the article points, with tremendous risk and challenges. The rest remain locked down to whatever Intel or AMD does or does not have to offer. But since the debut of the M1 suddenly other OEMs are investigating producing their own SOCs.
One thing is certain for those of us who keep riding the Apple log flume, it’s a thrilling ride and you never know what’s around the corner, just like an Indiana Jones movie.
Just a toe dip as that design was aimed at business customers in China but Apple wasn't alone in its thinking here.
Pray tell, where is Huawei going to fab its 5nm designs? I see a lot of propaganda about China building its own 5nm lithography machines, which I don't believe exist at all in production, so color me more than skeptical.
More to the point, nothing that Huawei has done is anything close to the disruptive architecture that Apple has created for M series silicon. It is just more of the same PC main board architecture, but now with more SOC!
The first is that Kirin 9006C is a chip previously produced by TSMC ;
There is one thing to say that the appearance of the Kirin 9006C at this time is really a bit abrupt. After all, after the Huawei 9000 chip, Huawei has not launched a new Kirin chip, and there is currently no foundry that can produce 5nm chips for Huawei.
Given that Huawei's current chip inventory is still a mystery, it is not clear how many Kirin chips Huawei has, and how many Kirin chips are in stock. Therefore, the Kirin 9006C is likely to be a chip produced by TSMC and Huawei during the cooperation.
Barking up distant trees as usual.
I was challenging the OP's notion that only Apple was doing the homegrown thing on desktops and laptops.
I wasn't talking about fabrication.
Two clearly different things.
I could have gone further but refrained from doing so and then you go and blow it with your 'disruptive Apple Silicon architecture claims'. So allow me to put your feet back on the ground.
There is zero disruptive going on here. Zero.
Apple has moved back to a non-Intel solution to processing. That is it. The only real difference with the old days is that they can pull all the strings now and don't have to depend on the likes of IBM or Motorola for product road maps. That is it at the moment.
So much for disruption!
Is it producing communications hardware or developing the technologies that go into that? Nope.
Is it producing high end cluster inference systems? Nope.
Laptops and desktops. That is it. There is no disruption to be found there. Zero.
Maybe Huawei should actually focus on a few niches, instead of attempting to be all things to all industries, after they got whacked in the handset market.
From the looks of it Huawei is making great strides in the surveillance industry, so there's that.
Oh, and as far as Apple Silicon being disruptive, yeah, it actually is.
Never before has a company’s entire product range, from tablet to all-in-one, been spread across a single processor. It is simply unimaginable in the PC market, which has instead built its complexity on a million different variations and declinations. Every year when Intel presents new versions of its processors, it does so with dozens and dozens of variations, designed on the basis of the activations of different areas of the chip, the number of cores and cores, the range of memory. An incredible puzzle made up of extremely articulated price lists, and with product ranges (i3, i5, i7, i9) with different consumption and frequencies, of which one wonders what actually are the differences that justify different price points and which should guide consumer choice on the basis of presumed differences in performance.
You found another tree!
Can't you see anything wrong with that quote? What is Apple's 'entire' product range?
Why would companies want all their products on one Soc (which is untrue anyway) if they already have successful business lines using varied SoCs?
But seeing as you bring the subject up, do you have any idea at all of the Ascend line of processors and how they scale from earbuds up to AI inference clusters? Passing through wearables, data centers etc. Do you know what part Mindspore or CANN play in the whole thing? This is years before Apple announced Apple Silicon.
As for Huawei as a business, well, nothing new to see, is there? It is moving ahead with its product goals in many fields and none of them are 'niches'. The cloud business is booming. The handset market is just one of them and obviously SoC fabrication is a problem - at the moment.
Surveillance industry? HiSilicon has been a leading player in camera SoCs for many years. At one point, it was claimed that as much as 70% of camera SoCs were HiSilicon. Yes. All over the US too.
That's right, basically everywhere you go, there are cameras watching you. London, New York, Paris, Milan, Barcelona and many smaller towns and cities. Government installed. Then there are private homes, business etc. Transportation. This isn't limited to China but you label it surveillance.
Fine, but surveillance is basically everywhere.
IoT. Is that in Apple's 'entire' product range? Another nope. Yet again, HiSilicon is there and offering solutions to industry partners with built in Open Harmony support.
So now you can look up Ascend Max, Mini, Tiny, Nano plus CANN/Mindspore etc and see how true the OP's statement was.
Comments
Kind of like self-driving cars. It's gonna take a lot longer to reach practical reality than people think.
Thus a lot of successful Windows apps have very old code bases and therefore wouldn’t be compatible with a Microsoft’s equivalent of Rosetta 2. And without Rosetta 2, Apple’s migration to ARM would have been dead in the water.
The greatest risk for Apple, in my opinion, had nothing at all to do with the traditional platform change challenges that they faced with PPC and Intel. The greatest challenge for Apple was to secure a world-class chip fabrication partner like TSMC who they could rely on over an extended period of time to stay on the leading edge and deliver in sufficient quantity and quality on very tight schedules. If Apple didn't have TSMC to depend on they'd be taking on a far greater challenge regardless of their own internal chip design prowess. So even though Apple has for the most part weaned themselves off of Intel and some other chipset makers, they aren't absolutely self-sufficient and still need to play nicely with critical partners to sustain the hardware side of their business.
From a risk management perspective, Apple has not removed risk entirely, they've only moved the risk to a different place/level. Hopefully, world events will not unfold in a way that impacts their risk before it can be properly mitigated through domestic investment in the critical technologies and manufacturing capabilities that Apple is dependent on today from TSMC. Apple is not alone in this regard, but they are very highly exposed. We as a nation should all recognize that this risk needs to be addressed and mitigated for the sake of much more than Apple and its silicon strategy. It's an existential national risk to US security and prosperity.
Even when Apple is using phone chips in their computers, they are still able to beat the computers with "real computer chips" from the competition
While Apple is churning higher performance per watt with ease and doesn't overheat like androids with latest chips.
https://en.wikipedia.org/wiki/List_of_quantum_processors
https://e.huawei.com/en/products/servers/kunpeng/kunpeng-desktop-board
And later from 2021:
https://www.tomshardware.com/news/huawei-dyna-cloud-kirin-9006c
Just a toe dip as that design was aimed at business customers in China but Apple wasn't alone in its thinking here.
It demands a mindset shift away from CPU OR GPU benchmarks to APU or Application benchmarks & preferably not those from an x86 vendor like Puget Systems.
More to the point, nothing that Huawei has done is anything close to the disruptive architecture that Apple has created for M series silicon. It is just more of the same PC main board architecture, but now with more SOC!
Meh.
...and just posted today...
https://inf.news/en/digital/1556de04cbcf47583854a18037c1f59b.html
I finally just blocked him.
I was challenging the OP's notion that only Apple was doing the homegrown thing on desktops and laptops.
I wasn't talking about fabrication.
Two clearly different things.
I could have gone further but refrained from doing so and then you go and blow it with your 'disruptive Apple Silicon architecture claims'. So allow me to put your feet back on the ground.
There is zero disruptive going on here. Zero.
Apple has moved back to a non-Intel solution to processing. That is it. The only real difference with the old days is that they can pull all the strings now and don't have to depend on the likes of IBM or Motorola for product road maps. That is it at the moment.
So much for disruption!
Is it producing communications hardware or developing the technologies that go into that? Nope.
Is it producing high end cluster inference systems? Nope.
https://e.huawei.com/en/products/cloud-computing-dc/atlas/ascend-910#:~:text=Ascend 910 is a high,systems both flexibly and efficiently.
Data centers? Nope.
https://e.huawei.com/en/solutions/hidc
Prefabricated mobile data centers? Nope.
https://e.huawei.com/en/products/network-energy/dc-facilities/ids1000-a
Mobile Data Centres for the automotive industry? Nope
https://e.huawei.com/en/news/it/201810151059
All of the interconnection technologies to keep those systems running? Nope.
Providing industry solutions for manufacturing and management (manufacturing, mining, ports, airports, health and science)? Nope?
https://e.huawei.com/en/solutions/industries/transportation/smart-aviation
Sustainable energy technologies to run all of that? Nope.
https://solar.huawei.com/eu
The OP was talking about CE!
Laptops and desktops. That is it. There is no disruption to be found there. Zero.
Apple, with its own OS, didn’t have to worry about that. They could support any chip they wanted to, as long as they could get third party developers to go along.
From the looks of it Huawei is making great strides in the surveillance industry, so there's that.
Oh, and as far as Apple Silicon being disruptive, yeah, it actually is.
https://techlapse.com/apple/apple-m1-chip-disempowered-the-intel-amd-market/
Can't you see anything wrong with that quote? What is Apple's 'entire' product range?
Why would companies want all their products on one Soc (which is untrue anyway) if they already have successful business lines using varied SoCs?
But seeing as you bring the subject up, do you have any idea at all of the Ascend line of processors and how they scale from earbuds up to AI inference clusters? Passing through wearables, data centers etc. Do you know what part Mindspore or CANN play in the whole thing? This is years before Apple announced Apple Silicon.
As for Huawei as a business, well, nothing new to see, is there? It is moving ahead with its product goals in many fields and none of them are 'niches'. The cloud business is booming. The handset market is just one of them and obviously SoC fabrication is a problem - at the moment.
Surveillance industry? HiSilicon has been a leading player in camera SoCs for many years. At one point, it was claimed that as much as 70% of camera SoCs were HiSilicon. Yes. All over the US too.
That's right, basically everywhere you go, there are cameras watching you. London, New York, Paris, Milan, Barcelona and many smaller towns and cities. Government installed. Then there are private homes, business etc. Transportation. This isn't limited to China but you label it surveillance.
Fine, but surveillance is basically everywhere.
IoT. Is that in Apple's 'entire' product range? Another nope. Yet again, HiSilicon is there and offering solutions to industry partners with built in Open Harmony support.
So now you can look up Ascend Max, Mini, Tiny, Nano plus CANN/Mindspore etc and see how true the OP's statement was.
https://aibusiness.com/document.asp?doc_id=769053
That was the point. Not the craziness you are attempting to inject here.
Do you know what year all this became reality?
https://www.huawei.com/en/technology-insights/publications/huawei-tech/86/driving-ai-to-new-horizons