However I stand by my assertion of Apple's customer base being driven by the unspoken desires of Apple's marketing. To the point that they see their product as the best in the world and then charging a price that reflects this mentality. Because in the end all the sheeple of the world see is that Apple logo on the back of your beat up MBP. Nothing else really matters to them (sheeple) beyond that.
How is this statement inapplicable to ANY company’s marketing department, except for the part where they’re not actually making the best products in the world?
I'm also not going to disregard one OS over another because it doesn't conform to the standards of independent tests conducted by others, or because I'm anti-Bill Gates/Steve Jobs/Linus Torvalds.
No, nor should you. If you hate something, get to know it better. Either you’ll keep hating it–but for correct reasons–or you’ll accept that it’s just not worth using, otherwise. No sense in blind hatred.
Intel chipsets and processors are expensive across the computing spectrum, not only for Apple.
Yes Intels top of the line chipsets are expensive but in the alternative OS world you have many machines built with far cheaper chip sets. Those chip sets can come from Intel or AMD. Frankly I haven't even looked at low end laptops in over a year now, but AMD's BRAZOS platform was very popular in the extreme low end machines awhile ago. So too cheaper Intel and AMD chipsets.
The point is a couple hundred dollars of an Apple laptop costs are directly related to the cost of Intels Chipset.
As for an alternative to Linux, that's a choice we all have to make. I've set up a media server running Linux Mint, it's great and I don't think I'll be changing that out any time soon. My Mac is for video and audio editing, and my PC is my work horse for daily messing about, database creation and other work. We could argue until we're both blue in the face about the pros and cons of each operating system, but in the end neither of us will convince the other that our way of thinking is superior.
The intent wasn't to convince you of anything! I bought a MBP and the Mac OS echo system because it is a far better platform, especially for a laptop. Frankly one could suggest that it is far better than any other platform right now for mobile computing. You basically get the power of UNIX in a very power efficient machine.
By the way I have Linux systems at home and a VM on the laptop. At work it is no choice other than Windows.
However I stand by my assertion of Apple's customer base being driven by the unspoken desires of Apple's marketing. To the point that they see their product as the best in the world and then charging a price that reflects this mentality. Because in the end all the sheeple of the world see is that Apple logo on the back of your beat up MBP. Nothing else really matters to them (sheeple) beyond that.
In many aspect the Mac is the best in the world. I have exposure to all of the major platforms and frankly Mac OS is the most trouble free of the lot. Does that mean it is perfect - no but it is certainly worth a few dollars more.
I have to wonder if a Mac user peed in your cereal this morning as you seem to be really upset with Mac users for some reason.
Don't get me wrong, I love my MBP, my iPhone and iPad, the Apple culture, and so on. But I'm also not going to disregard one OS over another because it doesn't conform to the standards of independent tests conducted by others, or because I'm anti-Bill Gates/Steve Jobs/Linus Torvalds.
Where are you pulling this from? Nobody here was talking about disregarding other OS's, I did say something about moving to a Mac as my primary machine from Linux. It was the right thing to do unless you are really into the constant tinkering and incompatible system updates that come out every six months.
I like to think I'm knowledgeable enough in technology as a whole to be completely at ease moving from one OS to another -seamlessly- and building a system from scratch to come to my own conclusions about how well it truly works. I'm far from some kind of genius, but I'm just as far from dumb. Different strokes for different folks, mate.
OK! Again I don't know what set you off here but what in the hell does building your own system have to do with the viability of bringing the Mac Book back or Apple slotting in a new machine into the line up.
How is this statement inapplicable to ANY company’s marketing department, except for the part where they’re not actually making the best products in the world?
Some people just don't grasp what marketing is. I see this all the time in the various Apple forums, in fact more so in Apple forums than others I view. The number one thing about marketing is that you can only sell or market if you will, those products that you have available at the moment.
We often see people bring up comments Steve Jobs made about the iPad and having to sharpen fingers. It is just some of Steve marketing genius at keeping people focused on what they had to sell at the moment. This doesn't have anything to do with the fact that iPhones existed at the time with much smaller screens or that the iPad Mini was already under development. It is simply selling the product that they where introducing at the time. Why people don't grasp this is beyond me.
Apple beyond a doubt is a company with a highly evolved marketing arm.
No, nor should you. If you hate something, get to know it better. Either you’ll keep hating it–but for correct reasons–or you’ll accept that it’s just not worth using, otherwise. No sense in blind hatred.
No sense at all in blind hatred. However I'm not sure how saying the Mac is the best platform out there for many (most) users is somehow twisted into a statement that implies the alternatives are crap.
Frankly the message that you are responding to made no sense to me at all. Like I said in the other reply it is like a Mac user peed in his cereal this morning.
Think about an iPad for a minute and imagine that SoC was built for performance not low thermal power. We would have a chip that could make for a very passable laptop.
So? The ARM based laptops we are talking about would be similar work horses. They would still run OS/X, still have large internal SSD storage, WiFi and what have you.
And they would lose all the x86 apps until ported to ARM and ARM chips "built for performance" are still only going to be bay trail range in power. If apple was satisfied with "passable" they'd have made netbooks which is what you are suggesting.
No thanks for now. Core i5 is far superior for laptop workloads. Maybe in a few years but by then intel would also be further along and our laptop workloads will be even more demanding.
And they would lose all the x86 apps until ported to ARM and ARM chips "built for performance" are still only going to be bay trail range in power. If apple was satisfied with "passable" they'd have made netbooks which is what you are suggesting.
A7 is a lot better than bay trail and if you up the clock by 1GHZ it would deliver nice single core performance. The idea here is that you are saving significant power over the i86 solutions. Add more cores and you wouldn't notice the difference between this A8ish ARM and an Intel i3 maybe even an i5.
No thanks for now. Core i5 is far superior for laptop workloads.
It certainly is if you compare a 1.3 GHz chip to a 3GHz chip. The real question is this how much overhead does Apple have clock wise in their A7. Will it run at 2, 2.5 or even 3 GHz? I don't know but the point is the chip is relatively slow at the 1.3 GHz clock range and it probably runs at that clock rate for thermal reasons and nothing else.
Maybe in a few years but by then intel would also be further along and our laptop workloads will be even more demanding.
The thing here is that Apple can tailor the chip to their specific needs. This is especially important in the realm of GPU performance and optimizations for OpenCL. Not that Intel is still behind the eight ball there, but it took them a long time to take GPUs seriously.
Would it take Apple years to catch up to Intel? Hard to say, but we have seen some amazing strides of late. If they double core performance yet again will the gap with Intel be significant? I don't see anything stopping Apple from producing such a laptop, one that they can sell at a significantly reduced price relative to the AIRs. Maybe it ends up as an iOS device, which frankly I don't want to see, or something in between. I just see Apple with an opportunity here to shake up the PC industry even more.
"But the Bay Trail's Geekbench score of 2,935 surpassed those of the LG G2 (2,154) and the Apple iPhone 5 as loaded with iOS 7 (1,296) by a significant margin."
"To put it nicely, the Bay Trail reference model obliterated the Atom Z2760–based systems that have come before it. Its PCMark 7 score of 2,560 was markedly higher than those of the closest contenders, the Asus VivoTab Smart ME400C (which scored 1,438) and theLenovo ThinkPad Tablet 2 (1,410). Whereas those systems required huge chunks of time to complete our HandBrake video conversion test (Lenovo's tablet was the fastest, needing 5 minutes 16 seconds), Bay Trail wrapped up the task in just 2:17. Its rendering score in CineBench R11.5 was an eye-popping improvement as well: 1.47, almost two and a half times the next-highest result (the 0.63 of the HP ElitePad 900).?
And Bay Trail is infinitely faster than the A7 at running x86 apps.
Quote:
The idea here is that you are saving significant power over the i86 solutions. Add more cores and you wouldn't notice the difference between this A8ish ARM and an Intel i3 maybe even an i5.
If you up the clock of the ARM by 1GHz you're going to significantly increase power use.
Add more cores and you're going to significantly increase power use.
The distance between Bay Trail/A7 performance and i5 performance is large.
The MBA geekbench score is 4471 for the 1.5 Ghz dual core i5.
By the time you've bridged that performance gap you've lost any power advantage vs the lowest power Haswells.
And you've lost x86 app compatibility. Any any Rosetta-like hack will cost you a lot in terms of performance as well.
Quote:
It certainly is if you compare a 1.3 GHz chip to a 3GHz chip.
It certainly is if you compare a 1.4 Ghz A7 (iPad Air) vs a 1.3 Ghz dual core i5 (MBA).
The real question is this how much overhead does Apple have clock wise in their A7. Will it run at 2, 2.5 or even 3 GHz? I don't know but the point is the chip is relatively slow at the 1.3 GHz clock range and it probably runs at that clock rate for thermal reasons and nothing else.
Thermal and power. Not thermal and nothing else. And as you say, there's no certainty that the A7 will clock all that fast.
Quote:
The thing here is that Apple can tailor the chip to their specific needs. This is especially important in the realm of GPU performance and optimizations for OpenCL. Not that Intel is still behind the eight ball there, but it took them a long time to take GPUs seriously.
The thing here is that the iPad Air is amazingly fast for iPad workflows while the i5 MBA is just okay for laptop workflows. Given that the A7 is half the speed of the i5 in the MBA it would make for a 3 year regression in MBA performance back to slightly above the 2010 Core 2 Duo MBA.
And tailor all you want…it still won't run x86 apps and would still be slow for running things like iMovie, Xcode, Eclipse, Pixelmator, Aperture, etc which is pretty much all I turn my Mac on for these days.
My kids use a Core 2 Duo mac mini (about par with iPad Air geek bench scores) and OMG, even with a SSD the thing is slow for more than just browsing. Even in comparison to my 3 year old Core i5 MBP which is also nothing much to speak of these days.
Quote:
Would it take Apple years to catch up to Intel? Hard to say, but we have seen some amazing strides of late. If they double core performance yet again will the gap with Intel be significant? I don't see anything stopping Apple from producing such a laptop, one that they can sell at a significantly reduced price relative to the AIRs. Maybe it ends up as an iOS device, which frankly I don't want to see, or something in between. I just see Apple with an opportunity here to shake up the PC industry even more.
Given that the Bay Trail SOCs are more or less price competitive with ARM SOCs I see no real price advantage in losing x86 compatibility to go A7. And the MBA running either an A7 or Bay Trail ends up in netbook range of performance.
That isn't going to shake up the industry at all.
What stops Apple from producing such a laptop today is that it would suck as a laptop in comparison to the MBA, have crappy ASP and cheapens the brand. Maybe in a few years the performance penalty won't be as bad as moving from 2013 to 2010. Today it is.
"But the Bay Trail's Geekbench score of 2,935 surpassed those of the LG G2 (2,154) and the Apple iPhone 5 as loaded with iOS 7 (1,296) by a significant margin."
"To put it nicely, the Bay Trail reference model obliterated the Atom Z2760–based systems that have come before it. Its PCMark 7 score of 2,560 was markedly higher than those of the closest contenders, the Asus VivoTab Smart ME400C (which scored 1,438) and theLenovo ThinkPad Tablet 2 (1,410). Whereas those systems required huge chunks of time to complete our HandBrake video conversion test (Lenovo's tablet was the fastest, needing 5 minutes 16 seconds), Bay Trail wrapped up the task in just 2:17. Its rendering score in CineBench R11.5 was an eye-popping improvement as well: 1.47, almost two and a half times the next-highest result (the 0.63 of the HP ElitePad 900).?
Thermal and power. Not thermal and nothing else. And as you say, there's no certainty that the A7 will clock all that fast.
The thing here is that the iPad Air is amazingly fast for iPad workflows while the i5 MBA is just okay for laptop workflows. Given that the A7 is half the speed of the i5 in the MBA it would make for a 3 year regression in MBA performance back to slightly above the 2010 Core 2 Duo MBA.
And tailor all you want…it still won't run x86 apps and would still be slow for running things like iMovie, Xcode, Eclipse, Pixelmator, Aperture, etc which is pretty much all I turn my Mac on for these days.
My kids use a Core 2 Duo mac mini (about par with iPad Air geek bench scores) and OMG, even with a SSD the thing is slow for more than just browsing. Even in comparison to my 3 year old Core i5 MBP which is also nothing much to speak of these days.
Given that the Bay Trail SOCs are more or less price competitive with ARM SOCs I see no real price advantage in losing x86 compatibility to go A7. And the MBA running either an A7 or Bay Trail ends up in netbook range of performance.
That isn't going to shake up the industry at all.
What stops Apple from producing such a laptop today is that it would suck as a laptop in comparison to the MBA, have crappy ASP and cheapens the brand. Maybe in a few years the performance penalty won't be as bad as moving from 2013 to 2010. Today it is.
You're delusional if you think Intel's Bay Trail is on par with the industry's system-wide adoption of ARM.
A lot better? It looks like Bay Trail is on par with the A7 if not better in terms of raw compute power.
The fact remains that the A7 is cooler and running on an older generation process. That pretty much by definition makes it a lot better. Besides in mobile it isn't just about raw CPU compute power! it is what the whole SoC offers that is important.
"<strong style="border:0px;color:rgb(34,149,171);font-style:normal;margin:0px;padding:0px;vertical-align:baseline;">Update</strong> : Intel responded with <a href="" style="color:rgb(34,149,171);margin:0px;padding:0px;vertical-align:baseline;">a Bay Trail run under IE11</a> , which comes in at 329.6 ms."
"But the Bay Trail's Geekbench score of 2,935 surpassed those of the <a href="http://www.pcmag.com/article2/0,2817,2422814,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">LG G2</a> (2,154) and the <a href="http://www.pcmag.com/article2/0,2817,2409999,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">Apple iPhone 5</a> as loaded with <a href="http://www.pcmag.com/article2/0,2817,2424169,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">iOS 7</a> (1,296) by a significant margin."
"To put it nicely, the Bay Trail reference model obliterated the Atom Z2760–based systems that have come before it. Its PCMark 7 score of 2,560 was markedly higher than those of the closest contenders, the <a href="http://www.pcmag.com/article2/0,2817,2415674,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">Asus VivoTab Smart ME400C</a> (which scored 1,438) and the<a href="http://www.pcmag.com/article2/0,2817,2415773,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">Lenovo ThinkPad Tablet 2</a> (1,410). Whereas those systems required huge chunks of time to complete our HandBrake video conversion test (Lenovo's tablet was the fastest, needing 5 minutes 16 seconds), Bay Trail wrapped up the task in just 2:17. Its rendering score in CineBench R11.5 was an eye-popping improvement as well: 1.47, almost two and a half times the next-highest result (the 0.63 of the <a href="http://www.pcmag.com/article2/0,2817,2417732,00.asp" style="color:rgb(0,123,161);margin:0px;padding:0px;vertical-align:baseline;" target="_blank">HP ElitePad 900</a> ).?
iPad Air scores 2643 on Geekbench vs 2935 for Bay Trail.
I suppose if geek bench was important to me I'd be a little concerned. On the other hand this just highlights that Bay Trail isn't really all that incredible. Atom has been around for awhile now and basically Intel latest effort barely pulls ahead of the first ARM 64 bit machine to be delivered. Do you not see a problem here, even with Intels advanced process technology they can't deliver an overwhelming win on a poor synthetic benchmark.
And Bay Trail is infinitely faster than the A7 at running x86 apps.
Who really gives two hoots about X86 apps. That is why I have a laptop.
If you up the clock of the ARM by 1GHz you're going to significantly increase power use.
Certainly but will you be in the same territory as the "I" series chips from Intel. If they hit ten watts it is still far better than the Intel offerings.
Add more cores and you're going to significantly increase power use.
Cores themselves don't add significant power. If I remember correctly Global Foundries indicated some time ago that the basic ARM core was pulling 500 milliwatts. Often most of the power goes to supporting hardware especially the caches.
The distance between Bay Trail/A7 performance and i5 performance is large.
The MBA geekbench score is 4471 for the 1.5 Ghz dual core i5.
By the time you've bridged that performance gap you've lost any power advantage vs the lowest power Haswells.
We really don't know that. Engineers are always coming up with new architectures. Sometimes the architectures are really radical like the "Mill" processor, sometimes they are simply improvements to an existing design.
And you've lost x86 app compatibility. Any any Rosetta-like hack will cost you a lot in terms of performance as well.
Again the desire to be compatible with a dying platform is not a big concern for me. Apple would have a massive software library right out of the gate if they simply provide a compatibility mode to run iOS apps as well as native ARM Mac OS apps. Frankly Apple already has most of the infrastructure in place on Mac OS to do this. All they need to do is make it transparent to the user.
It certainly is if you compare a 1.4 Ghz A7 (iPad Air) vs a 1.3 Ghz dual core i5 (MBA).
1465 single core vs 2366 for the i5
2643 multi core vs 4471 for the i5
That really isn't that bad and is the whole point of my proposal. That gap can be easily narrowed by improvement to the core and adding more cores.
Thermal and power. Not thermal and nothing else. And as you say, there's no certainty that the A7 will clock all that fast.
The thing here is that the iPad Air is amazingly fast for iPad workflows while the i5 MBA is just okay for laptop workflows. Given that the A7 is half the speed of the i5 in the MBA it would make for a 3 year regression in MBA performance back to slightly above the 2010 Core 2 Duo MBA.
In return we get a laptop that is $300 cheaper if not more. That is if the current A7 chip is used. I don't expect the current A7 to be used as is.
And tailor all you want…it still won't run x86 apps and would still be slow for running things like iMovie, Xcode, Eclipse, Pixelmator, Aperture, etc which is pretty much all I turn my Mac on for these days.
So? Really .that is you, many users have other usage patterns where the machine would be fine. Even then the specifics of how each of those apps would perform on an ARM platform is unknown.
My kids use a Core 2 Duo mac mini (about par with iPad Air geek bench scores) and OMG, even with a SSD the thing is slow for more than just browsing. Even in comparison to my 3 year old Core i5 MBP which is also nothing much to speak of these days.
But it is good enough for your kids right?
Given that the Bay Trail SOCs are more or less price competitive with ARM SOCs I see no real price advantage in losing x86 compatibility to go A7. And the MBA running either an A7 or Bay Trail ends up in netbook range of performance.
I see no good reason to be x86 compatible anymore. That was a big issue for me when I purchased my 2008 MBP all those years ago, it isn't anymore though. In fact i86 is a big drag on innovation if you ask me.
That isn't going to shake up the industry at all.
What stops Apple from producing such a laptop today is that it would suck as a laptop in comparison to the MBA, have crappy ASP and cheapens the brand. Maybe in a few years the performance penalty won't be as bad as moving from 2013 to 2010. Today it is.
You are assuming the machine would come with iPad AIR performance levels and you also seem to discount the importance of the GPU and other hardware in the SoC. As for the nonsense about cheapening the brand, they sell the Apple TV for less than $99, they still sell the iPad 2 and frankly this year they learned that the can't command the high prices on the laptops that they would like to have.
A company Apples size won't remain in business without reasonably priced hardware relative to the rest of the market. They learned that lesson this year when they had to discount and promote the laptops that where introduced with very high prices. An ARM based machine gives them the option to create an entry level machine (I don't want to call it a laptop) that is low enough in price that it is considered a different product from the AIR and MBP lines. The only thing I can see as being really important here is that it supports a version of Mac OS as its primary OS, to allow user access instead of a buttoned down iOS machine.
You're delusional if you think Intel's Bay Trail is on par with the industry's system-wide adoption of ARM.
Did you notice that most of that post was about benchmarks that don't really matter? Apple has been making very significant progress with their DIY SoC. I have complete confidence that they can keep this up for at least a couple of years more. Even if thesis laptop/undefined device, doesn't exist there are solid rumors about a larger iPad floating around. Such a machine would require a more powerful version of the A7 so if these near term rumors are correct I would expect a beefed up A7 would be available early next year.
Of course what gets beefed up in the chip might not be what some want to see improved. The likely hood is that the chip would be accelerated GPU wise to drive far more pixels. Of course to address the bandwidth issues new caches or maybe a wider memory interface would likely happen. Apple has a history of being able to tweak their "A" series chips to get the performance they need for a specific machine, so I could see all sorts of tweaks happening to the current A7 architecture. We aren't even talking about what might be in an A8 chip here.
The fact remains that the A7 is cooler and running on an older generation process. That pretty much by definition makes it a lot better. Besides in mobile it isn't just about raw CPU compute power! it is what the whole SoC offers that is important.
The Bay Trail tested by Anand showed 2.5W at the SoC level under load which is on par with estimates of the A7 TDP. That Intel is a half node ahead isn't an issue for Intel going forward given they'll be at 14nm before anyone else.
Quote:
I suppose if geek bench was important to me I'd be a little concerned. On the other hand this just highlights that Bay Trail isn't really all that incredible. Atom has been around for awhile now and basically Intel latest effort barely pulls ahead of the first ARM 64 bit machine to be delivered. Do you not see a problem here, even with Intels advanced process technology they can't deliver an overwhelming win on a poor synthetic benchmark.
ARM has been around longer than Atom. So what? The fact is that Intel has vastly improved the power consumption part of the performance per watt equation across it's entire line up as fast as ARM (including the A7) has increased the performance part of the performance per watt equation AND can still move up the performance envelope to the Core i7 performance levels.
Nope, I don't see a problem at all given 11.5W TDP Haswells with 4.5W SDPs in the works. The primary advantage that ARM still enjoys is die size.
Quote:
Who really gives two hoots about X86 apps. That is why I have a laptop.
Because you're advocating replacing the Core i5 with an uber A7 in a new MacBook model.
"Think about an iPad for a minute and imagine that SoC was built for performance not low thermal power. We would have a chip that could make for a very passable laptop.
So? The ARM based laptops we are talking about would be similar work horses. They would still run OS/X, still have large internal SSD storage, WiFi and what have you."
The answer is that no, they wouldn't be similar work horses for laptop workflows NOR would they run x86 apps.
Quote:
Certainly but will you be in the same territory as the "I" series chips from Intel. If they hit ten watts it is still far better than the Intel offerings.
If you hit 10W you are not "far better" than an 11.5W TDP Core i5. While the Y series Haswells are a bit MIA at least they exist.
Quote:
Cores themselves don't add significant power. If I remember correctly Global Foundries indicated some time ago that the basic ARM core was pulling 500 milliwatts. Often most of the power goes to supporting hardware especially the caches.
LOL. Some time ago is the operative word. ARM cores are not so basic anymore.
Quote:
Again the desire to be compatible with a dying platform is not a big concern for me. Apple would have a massive software library right out of the gate if they simply provide a compatibility mode to run iOS apps as well as native ARM Mac OS apps.
LOL. Macs are a dying platform? So why build an ARM based OSX laptop at all? And iOS apps would run about as well on a MBA as OSX apps would run on iPad. Poorly since they are designed for different user interaction.
Quote:
Frankly Apple already has most of the infrastructure in place on Mac OS to do this. All they need to do is make it transparent to the user.
That really isn't that bad and is the whole point of my proposal. That gap can be easily narrowed by improvement to the core and adding more cores.
Gee, all Apple needs to do is improve their leading edge ARM processor and add more cores.
Quote:
In return we get a laptop that is $300 cheaper if not more. That is if the current A7 chip is used. I don't expect the current A7 to be used as is.
If Apple wanted to make $699 laptops it can. And Bay Trail is going to be price competitive with ARM SoCs according to Intel and geared toward $599 tablets. It would still suck but at least it maintains app compatibility with the rest of the Mac line.
Quote:
So? Really .that is you, many users have other usage patterns where the machine would be fine. Even then the specifics of how each of those apps would perform on an ARM platform is unknown.
While every pickup truck owner does different things with their trucks the base requirements remain the same. Which is why they have beefier engines than subcompacts.
Every time a real world example, benchmark, use case is provided all you do is get upset and wave your arms around even harder. No, it's not just "me". It's everyone that has stuff that even the iPad Air would be asininely slow in comparison to the MBA.
Quote:
But it is good enough for your kids right?
Only for tasks the iPad can do. Which except for MS Office is all it does.
Quote:
I see no good reason to be x86 compatible anymore. That was a big issue for me when I purchased my 2008 MBP all those years ago, it isn't anymore though. In fact i86 is a big drag on innovation if you ask me.
So you see no good reason that an "entry level" Macbook should be compatible with the rest of the Mac lineup in terms of applications.
What? You're going to replace all the MBPs and iMacs and MPs with more magical A7 cores?
Quote:
You are assuming the machine would come with iPad AIR performance levels and you also seem to discount the importance of the GPU and other hardware in the SoC. As for the nonsense about cheapening the brand, they sell the Apple TV for less than $99, they still sell the iPad 2 and frankly this year they learned that the can't command the high prices on the laptops that they would like to have.
I can show you an Intel x86 SoC running in a reference platform with similar performance to the A7 SoC. You cannot show me a running ARM based SoC with similar performance to the Core i5 in a reference platform. The GPU is a wash. The Bay Trail FFRD is about 30% slower than the Tegra 4 GPU and slightly faster than the Adreno 225 GPU. Cloverview was paired with the PowerVR SGX545. Merrifield will be paired with a PowerVR 6 Rogue GPU. It's a SoC and Intel will use PowerVR when it thinks it needs to and has done so in the past.
Cheapening the brand isn't price. It's providing a slow laptop to save $300 when the iPad Air can already do everything an ARM based laptop can do by adding a keyboard and the MBA is about as low as you want to go for laptop workflows. Otherwise we'd see Core i3 Macbooks.
Quote:
A company Apples size won't remain in business without reasonably priced hardware relative to the rest of the market.
Oh bullshit. Apple has based it's entire business in not providing "reasonably priced hardware relative to the rest of the market" but instead "exceptional hardware at a higher price relative to the rest of the market".
Really...listen to yourself with the Apple is Doomed TM bullshit. OMG they're going to go out of business without an xMac. Oh way, ARM laptop.
Quote:
An ARM based machine gives them the option to create an entry level machine (I don't want to call it a laptop) that is low enough in price that it is considered a different product from the AIR and MBP lines. The only thing I can see as being really important here is that it supports a version of Mac OS as its primary OS, to allow user access instead of a buttoned down iOS machine.
AKA Netbook. No shit you don't want to call it a laptop. If you're going to wish for unicorns you're better off wishing for an xMac. At least it wouldn't suck.
Mac price points have been marching north for over a decade and Apple has not changed despite your assertion otherwise. In fact, if they don't simply cancel the Mac Mini, I expect the Mac Mini price to really jump up. There's no need for a $599 Mac Mini anymore in Apple's thinking for "switchers". The iPad/iPhone halo took care of exposing folks to OSX. They don't really make "entry level" anything.
If they keep the mini at all I expect to see just a $999 somewhat slower quad core i7 base model without Iris Pro, with 1TB Fusion standard and a $1199 slightly faster quad core i7 server model with Iris Pro and dual 1TB drives. BTO would just be for RAM and a couple SSD options.
If you need a truck Apple will sell you one with a beefy V8 and not a V6.
Comments
Don’t get me wrong, I love my MBP, my iPhone and iPad, the Apple culture, and so on. But…
This is not a guideline of what to do.
How is this statement inapplicable to ANY company’s marketing department, except for the part where they’re not actually making the best products in the world?
No, nor should you. If you hate something, get to know it better. Either you’ll keep hating it–but for correct reasons–or you’ll accept that it’s just not worth using, otherwise. No sense in blind hatred.
The point is a couple hundred dollars of an Apple laptop costs are directly related to the cost of Intels Chipset. The intent wasn't to convince you of anything! I bought a MBP and the Mac OS echo system because it is a far better platform, especially for a laptop. Frankly one could suggest that it is far better than any other platform right now for mobile computing. You basically get the power of UNIX in a very power efficient machine.
By the way I have Linux systems at home and a VM on the laptop. At work it is no choice other than Windows. In many aspect the Mac is the best in the world. I have exposure to all of the major platforms and frankly Mac OS is the most trouble free of the lot. Does that mean it is perfect - no but it is certainly worth a few dollars more.
I have to wonder if a Mac user peed in your cereal this morning as you seem to be really upset with Mac users for some reason. Where are you pulling this from? Nobody here was talking about disregarding other OS's, I did say something about moving to a Mac as my primary machine from Linux. It was the right thing to do unless you are really into the constant tinkering and incompatible system updates that come out every six months.
OK! Again I don't know what set you off here but what in the hell does building your own system have to do with the viability of bringing the Mac Book back or Apple slotting in a new machine into the line up.
We often see people bring up comments Steve Jobs made about the iPad and having to sharpen fingers. It is just some of Steve marketing genius at keeping people focused on what they had to sell at the moment. This doesn't have anything to do with the fact that iPhones existed at the time with much smaller screens or that the iPad Mini was already under development. It is simply selling the product that they where introducing at the time. Why people don't grasp this is beyond me.
Apple beyond a doubt is a company with a highly evolved marketing arm.
No sense at all in blind hatred. However I'm not sure how saying the Mac is the best platform out there for many (most) users is somehow twisted into a statement that implies the alternatives are crap.
Frankly the message that you are responding to made no sense to me at all. Like I said in the other reply it is like a Mac user peed in his cereal this morning.
And they would lose all the x86 apps until ported to ARM and ARM chips "built for performance" are still only going to be bay trail range in power. If apple was satisfied with "passable" they'd have made netbooks which is what you are suggesting.
No thanks for now. Core i5 is far superior for laptop workloads. Maybe in a few years but by then intel would also be further along and our laptop workloads will be even more demanding.
The thing here is that Apple can tailor the chip to their specific needs. This is especially important in the realm of GPU performance and optimizations for OpenCL. Not that Intel is still behind the eight ball there, but it took them a long time to take GPUs seriously.
Would it take Apple years to catch up to Intel? Hard to say, but we have seen some amazing strides of late. If they double core performance yet again will the gap with Intel be significant? I don't see anything stopping Apple from producing such a laptop, one that they can sell at a significantly reduced price relative to the AIRs. Maybe it ends up as an iOS device, which frankly I don't want to see, or something in between. I just see Apple with an opportunity here to shake up the PC industry even more.
A7 is a lot better than bay trail and if you up the clock by 1GHZ it would deliver nice single core performance.
A lot better? It looks like Bay Trail is on par with the A7 if not better in terms of raw compute power.
http://www.anandtech.com/show/7335/the-iphone-5s-review/5
"But the Bay Trail's Geekbench score of 2,935 surpassed those of the LG G2 (2,154) and the Apple iPhone 5 as loaded with iOS 7 (1,296) by a significant margin."
"To put it nicely, the Bay Trail reference model obliterated the Atom Z2760–based systems that have come before it. Its PCMark 7 score of 2,560 was markedly higher than those of the closest contenders, the Asus VivoTab Smart ME400C (which scored 1,438) and theLenovo ThinkPad Tablet 2 (1,410). Whereas those systems required huge chunks of time to complete our HandBrake video conversion test (Lenovo's tablet was the fastest, needing 5 minutes 16 seconds), Bay Trail wrapped up the task in just 2:17. Its rendering score in CineBench R11.5 was an eye-popping improvement as well: 1.47, almost two and a half times the next-highest result (the 0.63 of the HP ElitePad 900).?
http://www.pcmag.com/article2/0,2817,2424262,00.asp
iPad Air scores 2643 on Geekbench vs 2935 for Bay Trail.
http://www.primatelabs.com/blog/2013/10/ipad-air-benchmarks/
And Bay Trail is infinitely faster than the A7 at running x86 apps.
If you up the clock of the ARM by 1GHz you're going to significantly increase power use.
Add more cores and you're going to significantly increase power use.
The distance between Bay Trail/A7 performance and i5 performance is large.
The MBA geekbench score is 4471 for the 1.5 Ghz dual core i5.
By the time you've bridged that performance gap you've lost any power advantage vs the lowest power Haswells.
And you've lost x86 app compatibility. Any any Rosetta-like hack will cost you a lot in terms of performance as well.
It certainly is if you compare a 1.4 Ghz A7 (iPad Air) vs a 1.3 Ghz dual core i5 (MBA).
1465 single core vs 2366 for the i5
2643 multi core vs 4471 for the i5
http://feedproxy.google.com/~r/primatelabsblog/~3/pzWGX26-fmg/retina-macbook-pro-benchmarks
http://feedproxy.google.com/~r/primatelabsblog/~3/fNJU6q41xM0/ipad-air-benchmarks
Thermal and power. Not thermal and nothing else. And as you say, there's no certainty that the A7 will clock all that fast.
The thing here is that the iPad Air is amazingly fast for iPad workflows while the i5 MBA is just okay for laptop workflows. Given that the A7 is half the speed of the i5 in the MBA it would make for a 3 year regression in MBA performance back to slightly above the 2010 Core 2 Duo MBA.
And tailor all you want…it still won't run x86 apps and would still be slow for running things like iMovie, Xcode, Eclipse, Pixelmator, Aperture, etc which is pretty much all I turn my Mac on for these days.
My kids use a Core 2 Duo mac mini (about par with iPad Air geek bench scores) and OMG, even with a SSD the thing is slow for more than just browsing. Even in comparison to my 3 year old Core i5 MBP which is also nothing much to speak of these days.
Given that the Bay Trail SOCs are more or less price competitive with ARM SOCs I see no real price advantage in losing x86 compatibility to go A7. And the MBA running either an A7 or Bay Trail ends up in netbook range of performance.
That isn't going to shake up the industry at all.
What stops Apple from producing such a laptop today is that it would suck as a laptop in comparison to the MBA, have crappy ASP and cheapens the brand. Maybe in a few years the performance penalty won't be as bad as moving from 2013 to 2010. Today it is.
A lot better? It looks like Bay Trail is on par with the A7 if not better in terms of raw compute power.
http://www.anandtech.com/show/7335/the-iphone-5s-review/5
"But the Bay Trail's Geekbench score of 2,935 surpassed those of the LG G2 (2,154) and the Apple iPhone 5 as loaded with iOS 7 (1,296) by a significant margin."
"To put it nicely, the Bay Trail reference model obliterated the Atom Z2760–based systems that have come before it. Its PCMark 7 score of 2,560 was markedly higher than those of the closest contenders, the Asus VivoTab Smart ME400C (which scored 1,438) and theLenovo ThinkPad Tablet 2 (1,410). Whereas those systems required huge chunks of time to complete our HandBrake video conversion test (Lenovo's tablet was the fastest, needing 5 minutes 16 seconds), Bay Trail wrapped up the task in just 2:17. Its rendering score in CineBench R11.5 was an eye-popping improvement as well: 1.47, almost two and a half times the next-highest result (the 0.63 of the HP ElitePad 900).?
http://www.pcmag.com/article2/0,2817,2424262,00.asp
iPad Air scores 2643 on Geekbench vs 2935 for Bay Trail.
http://www.primatelabs.com/blog/2013/10/ipad-air-benchmarks/
And Bay Trail is infinitely faster than the A7 at running x86 apps.
If you up the clock of the ARM by 1GHz you're going to significantly increase power use.
Add more cores and you're going to significantly increase power use.
The distance between Bay Trail/A7 performance and i5 performance is large.
The MBA geekbench score is 4471 for the 1.5 Ghz dual core i5.
By the time you've bridged that performance gap you've lost any power advantage vs the lowest power Haswells.
And you've lost x86 app compatibility. Any any Rosetta-like hack will cost you a lot in terms of performance as well.
It certainly is if you compare a 1.4 Ghz A7 (iPad Air) vs a 1.3 Ghz dual core i5 (MBA).
1465 single core vs 2366 for the i5
2643 multi core vs 4471 for the i5
http://feedproxy.google.com/~r/primatelabsblog/~3/pzWGX26-fmg/retina-macbook-pro-benchmarks
http://feedproxy.google.com/~r/primatelabsblog/~3/fNJU6q41xM0/ipad-air-benchmarks
Thermal and power. Not thermal and nothing else. And as you say, there's no certainty that the A7 will clock all that fast.
The thing here is that the iPad Air is amazingly fast for iPad workflows while the i5 MBA is just okay for laptop workflows. Given that the A7 is half the speed of the i5 in the MBA it would make for a 3 year regression in MBA performance back to slightly above the 2010 Core 2 Duo MBA.
And tailor all you want…it still won't run x86 apps and would still be slow for running things like iMovie, Xcode, Eclipse, Pixelmator, Aperture, etc which is pretty much all I turn my Mac on for these days.
My kids use a Core 2 Duo mac mini (about par with iPad Air geek bench scores) and OMG, even with a SSD the thing is slow for more than just browsing. Even in comparison to my 3 year old Core i5 MBP which is also nothing much to speak of these days.
Given that the Bay Trail SOCs are more or less price competitive with ARM SOCs I see no real price advantage in losing x86 compatibility to go A7. And the MBA running either an A7 or Bay Trail ends up in netbook range of performance.
That isn't going to shake up the industry at all.
What stops Apple from producing such a laptop today is that it would suck as a laptop in comparison to the MBA, have crappy ASP and cheapens the brand. Maybe in a few years the performance penalty won't be as bad as moving from 2013 to 2010. Today it is.
You're delusional if you think Intel's Bay Trail is on par with the industry's system-wide adoption of ARM.
The fact remains that the A7 is cooler and running on an older generation process. That pretty much by definition makes it a lot better. Besides in mobile it isn't just about raw CPU compute power! it is what the whole SoC offers that is important. I suppose if geek bench was important to me I'd be a little concerned. On the other hand this just highlights that Bay Trail isn't really all that incredible. Atom has been around for awhile now and basically Intel latest effort barely pulls ahead of the first ARM 64 bit machine to be delivered. Do you not see a problem here, even with Intels advanced process technology they can't deliver an overwhelming win on a poor synthetic benchmark. Who really gives two hoots about X86 apps. That is why I have a laptop. Certainly but will you be in the same territory as the "I" series chips from Intel. If they hit ten watts it is still far better than the Intel offerings. Cores themselves don't add significant power. If I remember correctly Global Foundries indicated some time ago that the basic ARM core was pulling 500 milliwatts. Often most of the power goes to supporting hardware especially the caches. We really don't know that. Engineers are always coming up with new architectures. Sometimes the architectures are really radical like the "Mill" processor, sometimes they are simply improvements to an existing design. Again the desire to be compatible with a dying platform is not a big concern for me. Apple would have a massive software library right out of the gate if they simply provide a compatibility mode to run iOS apps as well as native ARM Mac OS apps. Frankly Apple already has most of the infrastructure in place on Mac OS to do this. All they need to do is make it transparent to the user. That really isn't that bad and is the whole point of my proposal. That gap can be easily narrowed by improvement to the core and adding more cores. In return we get a laptop that is $300 cheaper if not more. That is if the current A7 chip is used. I don't expect the current A7 to be used as is. So? Really .that is you, many users have other usage patterns where the machine would be fine. Even then the specifics of how each of those apps would perform on an ARM platform is unknown. But it is good enough for your kids right? I see no good reason to be x86 compatible anymore. That was a big issue for me when I purchased my 2008 MBP all those years ago, it isn't anymore though. In fact i86 is a big drag on innovation if you ask me.
You are assuming the machine would come with iPad AIR performance levels and you also seem to discount the importance of the GPU and other hardware in the SoC. As for the nonsense about cheapening the brand, they sell the Apple TV for less than $99, they still sell the iPad 2 and frankly this year they learned that the can't command the high prices on the laptops that they would like to have.
A company Apples size won't remain in business without reasonably priced hardware relative to the rest of the market. They learned that lesson this year when they had to discount and promote the laptops that where introduced with very high prices. An ARM based machine gives them the option to create an entry level machine (I don't want to call it a laptop) that is low enough in price that it is considered a different product from the AIR and MBP lines. The only thing I can see as being really important here is that it supports a version of Mac OS as its primary OS, to allow user access instead of a buttoned down iOS machine.
Did you notice that most of that post was about benchmarks that don't really matter? Apple has been making very significant progress with their DIY SoC. I have complete confidence that they can keep this up for at least a couple of years more. Even if thesis laptop/undefined device, doesn't exist there are solid rumors about a larger iPad floating around. Such a machine would require a more powerful version of the A7 so if these near term rumors are correct I would expect a beefed up A7 would be available early next year.
Of course what gets beefed up in the chip might not be what some want to see improved. The likely hood is that the chip would be accelerated GPU wise to drive far more pixels. Of course to address the bandwidth issues new caches or maybe a wider memory interface would likely happen. Apple has a history of being able to tweak their "A" series chips to get the performance they need for a specific machine, so I could see all sorts of tweaks happening to the current A7 architecture. We aren't even talking about what might be in an A8 chip here.
You're delusional if you think Intel's Bay Trail is on par with the industry's system-wide adoption of ARM.
You're delusional if you think your post responded to mine in any relatable fashion. Who was talking about adoption rates? Nobody.
The fact remains that the A7 is cooler and running on an older generation process. That pretty much by definition makes it a lot better. Besides in mobile it isn't just about raw CPU compute power! it is what the whole SoC offers that is important.
The Bay Trail tested by Anand showed 2.5W at the SoC level under load which is on par with estimates of the A7 TDP. That Intel is a half node ahead isn't an issue for Intel going forward given they'll be at 14nm before anyone else.
ARM has been around longer than Atom. So what? The fact is that Intel has vastly improved the power consumption part of the performance per watt equation across it's entire line up as fast as ARM (including the A7) has increased the performance part of the performance per watt equation AND can still move up the performance envelope to the Core i7 performance levels.
Nope, I don't see a problem at all given 11.5W TDP Haswells with 4.5W SDPs in the works. The primary advantage that ARM still enjoys is die size.
Because you're advocating replacing the Core i5 with an uber A7 in a new MacBook model.
"Think about an iPad for a minute and imagine that SoC was built for performance not low thermal power. We would have a chip that could make for a very passable laptop.
So? The ARM based laptops we are talking about would be similar work horses. They would still run OS/X, still have large internal SSD storage, WiFi and what have you."
The answer is that no, they wouldn't be similar work horses for laptop workflows NOR would they run x86 apps.
If you hit 10W you are not "far better" than an 11.5W TDP Core i5. While the Y series Haswells are a bit MIA at least they exist.
LOL. Some time ago is the operative word. ARM cores are not so basic anymore.
LOL. Macs are a dying platform? So why build an ARM based OSX laptop at all? And iOS apps would run about as well on a MBA as OSX apps would run on iPad. Poorly since they are designed for different user interaction.
That really isn't that bad and is the whole point of my proposal. That gap can be easily narrowed by improvement to the core and adding more cores.
Gee, all Apple needs to do is improve their leading edge ARM processor and add more cores.
If Apple wanted to make $699 laptops it can. And Bay Trail is going to be price competitive with ARM SoCs according to Intel and geared toward $599 tablets. It would still suck but at least it maintains app compatibility with the rest of the Mac line.
While every pickup truck owner does different things with their trucks the base requirements remain the same. Which is why they have beefier engines than subcompacts.
Every time a real world example, benchmark, use case is provided all you do is get upset and wave your arms around even harder. No, it's not just "me". It's everyone that has stuff that even the iPad Air would be asininely slow in comparison to the MBA.
Only for tasks the iPad can do. Which except for MS Office is all it does.
So you see no good reason that an "entry level" Macbook should be compatible with the rest of the Mac lineup in terms of applications.
What? You're going to replace all the MBPs and iMacs and MPs with more magical A7 cores?
I can show you an Intel x86 SoC running in a reference platform with similar performance to the A7 SoC. You cannot show me a running ARM based SoC with similar performance to the Core i5 in a reference platform. The GPU is a wash. The Bay Trail FFRD is about 30% slower than the Tegra 4 GPU and slightly faster than the Adreno 225 GPU. Cloverview was paired with the PowerVR SGX545. Merrifield will be paired with a PowerVR 6 Rogue GPU. It's a SoC and Intel will use PowerVR when it thinks it needs to and has done so in the past.
Cheapening the brand isn't price. It's providing a slow laptop to save $300 when the iPad Air can already do everything an ARM based laptop can do by adding a keyboard and the MBA is about as low as you want to go for laptop workflows. Otherwise we'd see Core i3 Macbooks.
Oh bullshit. Apple has based it's entire business in not providing "reasonably priced hardware relative to the rest of the market" but instead "exceptional hardware at a higher price relative to the rest of the market".
Really...listen to yourself with the Apple is Doomed TM bullshit. OMG they're going to go out of business without an xMac. Oh way, ARM laptop.
AKA Netbook. No shit you don't want to call it a laptop. If you're going to wish for unicorns you're better off wishing for an xMac. At least it wouldn't suck.
Mac price points have been marching north for over a decade and Apple has not changed despite your assertion otherwise. In fact, if they don't simply cancel the Mac Mini, I expect the Mac Mini price to really jump up. There's no need for a $599 Mac Mini anymore in Apple's thinking for "switchers". The iPad/iPhone halo took care of exposing folks to OSX. They don't really make "entry level" anything.
If they keep the mini at all I expect to see just a $999 somewhat slower quad core i7 base model without Iris Pro, with 1TB Fusion standard and a $1199 slightly faster quad core i7 server model with Iris Pro and dual 1TB drives. BTO would just be for RAM and a couple SSD options.
If you need a truck Apple will sell you one with a beefy V8 and not a V6.