Lol. No. If it were that easy we'd see more ARM servers. It isn't easy and ARM inroads into the server market are far below expectations as micro servers didn't explode like folks were claiming it would.
Yes but Intel really hasn't made any progress with Micro Servers either. It is a problem of marketing not understanding customer needs more than technology failing. Servers will get smaller for the same reason everything else electronic has. The question is how do you package up a Micro Server to make it useful to those that would buy it? It is interesting that the Mini has had some success here but that is due more to a static size (fixed platform) than anything.
Intel sold 100 million CPUs in Q4 2014. 2015 has been better than 2014 for them from what I remember. They are doing fine and still ahead in process.
I'm not too sure about that, Intel had a few layoffs due to poor sales. Even if I'm wrong about that Inels sales figures don't matter even if they are better than last years. It isn't Intels sales figures that are significant it it rather Apples sales figures of ARM based chips that is. If Apple gets to the point where they shift 250+ million a year (very possible) then that justifies a huge amount of R&D. I haven't actually looked at the numbers but Apple can't be far from that 250M number now. In any event with all of that R&D available to them producing high performance desktop chips isn't an issue.
Beyond all of that Intel hasn't been ahead on process for at least the last year. A8 had a much high component density that the chips Intel was shipping at the time, even if the chip was built on a larger process. Beyond that Intels chips are still thermally challenged relative to the Arm solutions Apple ships.
There is no way Apple moves from Intel for the MacBooks in the next few years.
Well maybe they won't call them Mac Books. I wouldn't want to see the market confusion that would cause, that doesn't however mean that I wouldn't want to see an ARM based Apple laptop with Mac OS running on it.
Your comments regarding micro servers are thoughtful and cogent.
The volumes of CPUs Apple sells would easily allow the design team the kind of funding to design high performance chips. Right now Apple is focusing on energy consumption and designing as performance into the SOC based on a given TDP.
If Apple decides to, they could design a high performance model of the chip tethered to a cord. Designing such a chip would have technical challenges, but Apple's design team is up to the task. Their mobile design team has been superior to Intel's and are now firmly in the lead. When Intel has to compare their Core series with Apple's ARM SOCs, that's quite the development. Especially with the SOC costing less than a tenth of the equivalent Core CPU.
It is inevitable that Apple will move their entire line over to ARM CPUs. They are no longer dependent on Intel's schedule and they can reap the benefits of their own chips which will not be made available to everyone else via Intel. Losing Apple will be a massive blow to Intel. But Otellini started that process. Apple has learned that they are able to produce high performance processors and able to outcompete Intel in the mobile realm. They will move into high performance computing apart from mobile. That's Intel's real problem. Because the volumes of chips Apple moves will allow the financial investments in extending mobile CPU development over to high performance development also.
Correction to the above post. Apple does not sell processors. I meant to say the number of processors Apple produces. They do sell the processors as part of the mobile devices.
Your comments regarding micro servers are thoughtful and cogent.
The volumes of CPUs Apple sells would easily allow the design team the kind of funding to design high performance chips. Right now Apple is focusing on energy consumption and designing as performance into the SOC based on a given TDP.
If Apple decides to, they could design a high performance model of the chip tethered to a cord. Designing such a chip would have technical challenges, but Apple's design team is up to the task. Their mobile design team has been superior to Intel's and are now firmly in the lead. When Intel has to compare their Core series with Apple's ARM SOCs, that's quite the development. Especially with the SOC costing less than a tenth of the equivalent Core CPU.
It is inevitable that Apple will move their entire line over to ARM CPUs. They are no longer dependent on Intel's schedule and they can reap the benefits of their own chips which will not be made available to everyone else via Intel. Losing Apple will be a massive blow to Intel. But Otellini started that process. Apple has learned that they are able to produce high performance processors and able to outcompete Intel in the mobile realm. They will move into high performance computing apart from mobile. That's Intel's real problem. Because the volumes of chips Apple moves will allow the financial investments in extending mobile CPU development over to high performance development also.
It's interesting that Apple is still quite tied to MS wrt x86 and mainstream applications, so that will definitely keep some product lines with Intel processors. But look for Apple to displace x86 products over time with ARM products including entire product lines.
Intel's dilemma is that they earn high margins from their near monopoly in desktop processors, but they are beginning to lose the lower end of the laptop/notebook market to ARM, and are barely relevant in tablets or smartphones, also primarily ARM. Apple saves big with the A Series, and is able to customize its processors which is something that Intel has not, and likely won't be interested in doing. What I would fear, if I was Intel, is that MS sees a growth market for Windows on ARM, and even though it has failed with the Surface RT, might find its way back again, with Apple as a silent, though major, supporter.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
If Apple is making over 85% of the profits in handsets, and Samsung over 10%, is Android market share all that important? It's not like Apple is hurting for developers or apps.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
That is called collusion. Neither company is going to attempt that.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
If MS wanted to hurt the Android platform, they needn't have made MS Office available to it. However, MS also doesn't want Google Office apps to gain any more traction, so maybe it was strategic from that point. Plus, MS gets some serious money each time an Android phone ships.
In addition, MS doesn't want to work with Google, especially because of their "search wars," but MS can't get in bed with Apple because they are still being looked at hard for the shit they pulled in the '80s.
From what I've heard MS is putting their weight behind low cost phones. I don't think many people think of a Windows phone and think of quality or ease of use... most certainly people don't think of Windows for safety.
Yes but Intel really hasn't made any progress with Micro Servers either.
Because there's no market to make progress in.
Quote:
It is a problem of marketing not understanding customer needs more than technology failing. Servers will get smaller for the same reason everything else electronic has. The question is how do you package up a Micro Server to make it useful to those that would buy it? It is interesting that the Mini has had some success here but that is due more to a static size (fixed platform) than anything.
No, because it's a solution in search of a problem. Everyone looked at it and shrugged. Virtualization solves the same problem more efficiently and provides more flexibility in terms of changing needs. It is easy to turn a big server in to 1000 little ones. It is harder to turn 1000 little ones into 1 big one and it turned out the little ones weren't 20X as effective as claimed by HP (half rack = 10 racks of 1U servers).
Hence the fact that instead of being 10% of the market it is 1% of the market. Everyone wanted to see 10x power, heat and cost savings much less 20x so the problem with marketing wasn't they couldn't sell people on the idea but that micro servers didn't live up to the hype or marketing.
Quote:
I'm not too sure about that, Intel had a few layoffs due to poor sales. Even if I'm wrong about that Inels sales figures don't matter even if they are better than last years. It isn't Intels sales figures that are significant it it rather Apples sales figures of ARM based chips that is. If Apple gets to the point where they shift 250+ million a year (very possible) then that justifies a huge amount of R&D. I haven't actually looked at the numbers but Apple can't be far from that 250M number now. In any event with all of that R&D available to them producing high performance desktop chips isn't an issue.
They screwed up their last round of layoffs no doubt and they didn't see the growth they expected in 2015. That said, the difference between Intel and Apple is chip making is a core focus for Intel and just a strategic benefit for Apple.
And creating a high performance server chip has been a significant;y more challenging problem than lowering TDP down to 4.5W which isn't much higher than the estimated 3W TDP for the A9 and performance is higher for real world use. And efficiency isn't linear when you move up in TDP to get more performance.
Look at intels performance per watt at server level TDP vs ARM:
Even if the A9 was 2x better than the X-Gene it would still be lower than the Atom.
Quote:
Beyond all of that Intel hasn't been ahead on process for at least the last year. A8 had a much high component density that the chips Intel was shipping at the time, even if the chip was built on a larger process. Beyond that Intels chips are still thermally challenged relative to the Arm solutions Apple ships.
4.5W vs 3.0W for the A9. The A9X is probably around 4-4.5W TDP.
Quote:
Well maybe they won't call them Mac Books. I wouldn't want to see the market confusion that would cause, that doesn't however mean that I wouldn't want to see an ARM based Apple laptop with Mac OS running on it.
What would be better is a 12" Macbook with a touchscreen. I dunno how many times I've tried to pinch and zoom or swipe on my MBP without thinking.
Comments
Beyond all of that Intel hasn't been ahead on process for at least the last year. A8 had a much high component density that the chips Intel was shipping at the time, even if the chip was built on a larger process. Beyond that Intels chips are still thermally challenged relative to the Arm solutions Apple ships.
Well maybe they won't call them Mac Books. I wouldn't want to see the market confusion that would cause, that doesn't however mean that I wouldn't want to see an ARM based Apple laptop with Mac OS running on it.
Your comments regarding micro servers are thoughtful and cogent.
The volumes of CPUs Apple sells would easily allow the design team the kind of funding to design high performance chips. Right now Apple is focusing on energy consumption and designing as performance into the SOC based on a given TDP.
If Apple decides to, they could design a high performance model of the chip tethered to a cord. Designing such a chip would have technical challenges, but Apple's design team is up to the task. Their mobile design team has been superior to Intel's and are now firmly in the lead. When Intel has to compare their Core series with Apple's ARM SOCs, that's quite the development. Especially with the SOC costing less than a tenth of the equivalent Core CPU.
It is inevitable that Apple will move their entire line over to ARM CPUs. They are no longer dependent on Intel's schedule and they can reap the benefits of their own chips which will not be made available to everyone else via Intel. Losing Apple will be a massive blow to Intel. But Otellini started that process. Apple has learned that they are able to produce high performance processors and able to outcompete Intel in the mobile realm. They will move into high performance computing apart from mobile. That's Intel's real problem. Because the volumes of chips Apple moves will allow the financial investments in extending mobile CPU development over to high performance development also.
@Wizard69.
Your comments regarding micro servers are thoughtful and cogent.
The volumes of CPUs Apple sells would easily allow the design team the kind of funding to design high performance chips. Right now Apple is focusing on energy consumption and designing as performance into the SOC based on a given TDP.
If Apple decides to, they could design a high performance model of the chip tethered to a cord. Designing such a chip would have technical challenges, but Apple's design team is up to the task. Their mobile design team has been superior to Intel's and are now firmly in the lead. When Intel has to compare their Core series with Apple's ARM SOCs, that's quite the development. Especially with the SOC costing less than a tenth of the equivalent Core CPU.
It is inevitable that Apple will move their entire line over to ARM CPUs. They are no longer dependent on Intel's schedule and they can reap the benefits of their own chips which will not be made available to everyone else via Intel. Losing Apple will be a massive blow to Intel. But Otellini started that process. Apple has learned that they are able to produce high performance processors and able to outcompete Intel in the mobile realm. They will move into high performance computing apart from mobile. That's Intel's real problem. Because the volumes of chips Apple moves will allow the financial investments in extending mobile CPU development over to high performance development also.
It's interesting that Apple is still quite tied to MS wrt x86 and mainstream applications, so that will definitely keep some product lines with Intel processors. But look for Apple to displace x86 products over time with ARM products including entire product lines.
Intel's dilemma is that they earn high margins from their near monopoly in desktop processors, but they are beginning to lose the lower end of the laptop/notebook market to ARM, and are barely relevant in tablets or smartphones, also primarily ARM. Apple saves big with the A Series, and is able to customize its processors which is something that Intel has not, and likely won't be interested in doing. What I would fear, if I was Intel, is that MS sees a growth market for Windows on ARM, and even though it has failed with the Surface RT, might find its way back again, with Apple as a silent, though major, supporter.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
If Apple is making over 85% of the profits in handsets, and Samsung over 10%, is Android market share all that important? It's not like Apple is hurting for developers or apps.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
That is called collusion. Neither company is going to attempt that.
This may happen since Google Android controls over 81% of the smartphone market and Apple SoC may help Microsoft in so fashion. Not such how, but I can see an allies to block Android from gain even more market share!
If MS wanted to hurt the Android platform, they needn't have made MS Office available to it. However, MS also doesn't want Google Office apps to gain any more traction, so maybe it was strategic from that point. Plus, MS gets some serious money each time an Android phone ships.
In addition, MS doesn't want to work with Google, especially because of their "search wars," but MS can't get in bed with Apple because they are still being looked at hard for the shit they pulled in the '80s.
From what I've heard MS is putting their weight behind low cost phones. I don't think many people think of a Windows phone and think of quality or ease of use... most certainly people don't think of Windows for safety.
Yes but Intel really hasn't made any progress with Micro Servers either.
Because there's no market to make progress in.
No, because it's a solution in search of a problem. Everyone looked at it and shrugged. Virtualization solves the same problem more efficiently and provides more flexibility in terms of changing needs. It is easy to turn a big server in to 1000 little ones. It is harder to turn 1000 little ones into 1 big one and it turned out the little ones weren't 20X as effective as claimed by HP (half rack = 10 racks of 1U servers).
Hence the fact that instead of being 10% of the market it is 1% of the market. Everyone wanted to see 10x power, heat and cost savings much less 20x so the problem with marketing wasn't they couldn't sell people on the idea but that micro servers didn't live up to the hype or marketing.
They screwed up their last round of layoffs no doubt and they didn't see the growth they expected in 2015. That said, the difference between Intel and Apple is chip making is a core focus for Intel and just a strategic benefit for Apple.
And creating a high performance server chip has been a significant;y more challenging problem than lowering TDP down to 4.5W which isn't much higher than the estimated 3W TDP for the A9 and performance is higher for real world use. And efficiency isn't linear when you move up in TDP to get more performance.
Look at intels performance per watt at server level TDP vs ARM:
http://www.anandtech.com/show/8357/exploring-the-low-end-and-micro-server-platforms/17
Even if the A9 was 2x better than the X-Gene it would still be lower than the Atom.
4.5W vs 3.0W for the A9. The A9X is probably around 4-4.5W TDP.
What would be better is a 12" Macbook with a touchscreen. I dunno how many times I've tried to pinch and zoom or swipe on my MBP without thinking.