You suggested that customers who buy Mac Pros and 17" laptops are "their most loyal and biggest spending high end power users". These are the same people who would say that if they don't get what they want they'll either stop buying until they get what they want or switch to a Windows PC. They spend big once and then don't upgrade and if there isn't a specific upgrade, the loyalty is worth nothing.
Buyers of 15" Macbook Pros are currently Apple's most loyal high-end power users and as such get the foremost attention from Apple.
If the prices drop in June and they don't bring back the 17", the people waiting will just have to deal with the 15". Apple has to make a higher screen resolution again and they might not be able to get the yields good enough to justify doing for such a low volume of customers.
It's still an important and high spending sector who use these for a living, just like Mac pro users. You may be happy with what's available NOW, I and many others are not.
Apple has already stopped selling the Mac Pro across Europe and this will be the case for the next few months. The 17" MBP has been out of sale for about a year now. If these events had a significant impact to Apple, you can bet they'd have done something about it.
It may be significant to some potential buyers but you use whatever tools are available.
but these things happen. Now that they have the manufacturing line setup for Retina displays, they shouldn't have the same problems they did before and they have no reason to change the design again.
Now that they have the manufacturing line setup for Retina displays, they shouldn't have the same problems they did before and they have no reason to change the design again.
Now that they have the manufacturing line setup for Retina displays, they shouldn't have the same problems they did before and they have no reason to change the design again.
Who is making them now?
I'd expect the same suppliers but this is obviously a fairly new process for all of them.
I'm curious how much of the market thinks like that, but I noticed the people upgrading 1,1s listed in there. Many may not be original owners, but regardless they really aren't a substantial market factor.
Quote:
You know what the problem is. When you hold onto hardware for 6 years, the resale value is gone so it's a much better option to upgrade components than to buy new.
I have told people before to treat computers as sunken costs. If they're too concerned with residual value, it's likely that they spent too much initially. Considering the number of 1,1s, some of those could also be secondary owners who don't really drive sales anyway.
Quote:
That doesn't explain the desire to upgrade.
I think many of those upgrading really don't know what they need. In the case of the 4,1s, they probably do so due to price. Intel launched the W3680 at $1000. The 5,1 was over $4000 by the time you added ram, tax, etc. Let's say they bought a 4,1 in 2009 and regularly upgraded every 2-3 years. Most of these guys were probably upgrading from 1,1s aside from new customers. It's unlikely that most of them would go for a 2010 model the following year as it was basically a revision B. They dropped in new cpus and gpus. Nothing else changed. Like the other vendors the sub $3000 segment retained the same cpus they used the prior year, so unless they both went for the upgrade and increased their budget, there wasn't anything of significance. Given the Sandy Bridge E delay and price drops from intel that didn't really filter down through most of the oems to the customer level, is it really that surprising that they made simple upgrades there as stopgap measures?
In terms of actual sales, it is a matter of how many people would buy an updated version this year or next year. Anyway I've stated before that the market for workstations will hold longer regardless of whether Apple makes one. It's not likely to drive their growth either way, but it is somewhat healthier than the low end of the PC market. Sometimes you get a bit hyperbolic with your assertions.
There's little chance that the people upgrading 1,1s are ever going to be purchasing 12 core models. If they started with the dualcore 2.66 models, they spent $2500 + whatever on memory at the time. The fbdimms were quite expensive at the time, but not all of them will be looking at much higher pricing tiers. If anything the past few years, I've considered the lower mac pros models to represent poor value unless you're locked in. I've never suggested Apple would be doomed without the mac pro. Really that kind of market should be more interesting for smaller companies.
Articles and speculation on quantum computing theory go back many years. Considering your wealth of links (not sure how you find so many) you probably already know this. I find them more interesting when they're close to being something that could be implemented in the near future. Right now gpu technologies interest me more than anything due to the way they can handle parallel processes.
Quote:
The CUDA revolution started with just a single research project:
I didn't know about that. I noticed it details some of the features in CUDA relative to OpenCL. I like some of the questions asked there, like the one below. What is interesting to me is that certain things that would have lagged on the 2009 era 8 cores can be driven by some of these gpu technologies when implemented. In a couple areas it's really interesting if you deal with a lot of still image and video data in raw formats. It allows you to apply quite a bit more adjustment without baking things like gamma and debayering. Of course it's software dependent, but I can see a lot of this stuff emerging.
Quote:
With OpenCL, you gain the advantage of cross-platform support, but lose automated tools, such as memory management, that are found with CUDA. It seems that as a scientist, you'd want to decrease your startup development costs, but at the same time, you'd want support for multiple platforms. What's the best way to reconcile this challenge?
Quote:
The smaller form factors will keep eroding away the large ones, especially the more that GPU computing takes hold:
I never disagreed with that, although I wonder how aggressively Apple will move to IGPs only. They often move 1-2 cycles earlier than technology is really ready figuring that whatever people need more will just upgrade the following cycle as it's not possible to expect everyone will upgrade annually. 3 years is probably a typical assessment. Many oems set up their available warranties to go as far as 3 years out including Apple.
Quote:
Servers will always exist but they don't use the tower form factor.
I'm not as hung up on form factors as you seem to think. I don't know why this comes up again and again. When I refer to a tower, it's what is there today. There are certain features I like such as independent keyboards as they allow more freedom in terms of placement. I have certain preferences in terms of displays and peripheral devices, but sometimes it's just a matter of flexibility.
I'd expect the same suppliers but this is obviously a fairly new process for all of them.
Missed this one. I suspect you'll eventually see some kind of reference implementation, which will make them cheaper. 1080 is not uncommon today. You even find it on 13" models. I personally placed far more value in the improvements in viewing angles and the reduced color temperature. The high res 15" models measured roughly 8000K whites for me. The rMBP is more like D65. The stock models were a bit low in resolution for my taste, but the high res ones were good. Of course we all have different priorities. I just hated the super cool whites and weird screen gamma.
I am not an Apple worshipping corporatist. In my world, the customer is always right. Apple sold hundreds of thousands of 17 inch MBPs the last year they were available so there is demand. Fail to satisfy a market segment long enough and it will ALWAYS find an alternative. Apple can easily afford to service these customers so why lose them? Good enough for most people is a dismal philosophy. Successful companies don't cede market segments. They fight for the whole pie. Their stock has taken a ferocious beating this year so no one is immune to failure.
I'm curious how much of the market thinks like that.
I have told people before to treat computers as sunken costs. If they're too concerned with residual value, it's likely that they spent too much initially. Considering the number of 1,1s, some of those could also be secondary owners who don't really drive sales anyway.
I don't think anyone would consider a computer to be an item to eventually throw away. Some of the MP1,1 owners tried to sell it and couldn't so that led them to upgrade the parts, which suggests they were the original owners. Here's another original owner wondering why his machine won't run Mountain Lion:
"I can't just throw my Mac Pro in garbage because I can't upgrade. Why do I have to upgrade? Well, Sibelius 7 works only in Mountain Lion, and soon, many other software will.
What do I have to change in my mac pro to make it compatible for upgrading to the latest version of mac OS?
I think Apple should be proud to help 7 year old expensive computers, and not just tell you "buy a new one, your computer is old". I guess that's how they can make more money"
This just seems to be the recurring theme of the Mac Pro owner.
Given the Sandy Bridge E delay and price drops from intel that didn't really filter down through most of the oems to the customer level, is it really that surprising that they made simple upgrades there as stopgap measures?
But they're upgrading to chips that are slower or the same as new machines that are available.
The Westmere 3.33GHz MP (2012) costs $2999. This is the same chips used in the upgrade thread:
So, after having a computer for 4 years, rather than sell it for $1000 and pay $2000 to upgrade, the choice is to pay $610 and get some thermal paste to do the upgrade or worst case, consider a $1500 upgrade from OWC. Anything to save paying Apple the full amount. I don't blame people for wanting to save $500-1400 but these people can't be described as big spenders or loyal to Apple.
The reason Apple charges so much is because people aren't buying from them often enough.
the market for workstations will hold longer regardless of whether Apple makes one. it is somewhat healthier than the low end of the PC market.
I don't believe this is the case. You've seen the sales figures for the workstation market. By 2015 they'll have the equivalent of a GTX 680 in a MBP and easily double the performance of the CPU, which is currently at 8-core 2009 MP level and they'll have DDR4 memory supporting up to 32GB. 512GB SSDs will be very affordable. More and more people will migrate down to nicer form factors and more affordable hardware.
I wonder how aggressively Apple will move to IGPs only.
I think they'll use Haswell's IGP in more machines. Even the entry iMac can get away with it, although it would have to be the GT3 version. I'd like to see them get rid of the 21.5" and bring out a 27" for $1299.
There are certain features I like such as independent keyboards as they allow more freedom in terms of placement. I have certain preferences in terms of displays and peripheral devices, but sometimes it's just a matter of flexibility.
You can get a laptop or iMac and have all of these things. You can even buy a Mac Pro enclosure, shut the MBP and sit it inside it and just pretend you have a 2009 8-core Mac Pro when people come over. They might wonder how you got SATA 6G and USB 3 but you can say you have PCI cards for that.
Apple can easily afford to service these customers so why lose them? Good enough for most people is a dismal philosophy. Successful companies don't cede market segments.
It's unlikely they'll lose all of the 17" customers. It's the same deal with the glossy/matte MBP. Steve Jobs mentioned that most people were buying glossy models so they dropped the matte versions. The Retina display makes it pretty much a non-issue now. 17" MBP owners have the option to buy a much larger external display - you even get USB-powered 20"+ displays.
Their stock has taken a ferocious beating this year so no one is immune to failure.
That isn't because of how they're doing though. That's because of how people think they are doing. They are still making record sales every quarter and are still the most profitable company in the world. You're right that no one is immune to failure but the number 1 company in the world is the last company to be concerned about.
I still don't want integrated graphics in the iMac. I want the best possible stuff Apple can put in for the price they choose to have not "what is good enough for the masses."
To use a sports analogy, I want me a Joe Montana in my Mac and not a Steve Young even though a Steve Young is good enough.
I disagree. Top of the line halo products have a benefit that goes beyond that market segment of audio visual professionals, something Steve Jobs understood. He always wanted to be at the top, without excuses or rationalizations,or telling customers to buy something they really weren't excited about and just make due.
I disagree. Top of the line halo products have a benefit that goes beyond that market segment of audio visual professionals, something Steve Jobs understood.
You know what else he understood:
Even way back in 2007 they said that 2/3 of their lineup sales were notebooks and at that point, the PC industry was at 40% notebooks.
In terms of 'halo products', that gets mentioned a few times and the analogy is to sports cars but it doesn't hold up. People don't aspire to own a 17" laptop or a Mac Pro. They are cumbersome and unattractive form factors. The 17" MBP didn't perform any better than the 15" so its appeal was purely down to screen size. The entry Mac Pros don't perform much better than the iMac and by the time you hit the ones that do at $3000-4000, they are too expensive.
Apple's halo products are the ones that people admire the most and those are the iPhone and iPad.
It's unlikely they'll lose all of the 17" customers. It's the same deal with the glossy/matte MBP. Steve Jobs mentioned that most people were buying glossy models so they dropped the matte versions. The Retina display makes it pretty much a non-issue now. 17" MBP owners have the option to buy a much larger external display - you even get USB-powered 20"+ displays.
That isn't because of how they're doing though. That's because of how people think they are doing. They are still making record sales every quarter and are still the most profitable company in the world. You're right that no one is immune to failure but the number 1 company in the world is the last company to be concerned about.
Jobs said a lot of things. He was a salesman, so it was to be expected.
Quote:
I don't believe this is the case. You've seen the sales figures for the workstation market. By 2015 they'll have the equivalent of a GTX 680 in a MBP and easily double the performance of the CPU, which is currently at 8-core 2009 MP level and they'll have DDR4 memory supporting up to 32GB. 512GB SSDs will be very affordable. More and more people will migrate down to nicer form factors and more affordable hardware.
Well technically you can get 32GB into a notebook today. Thinkpads seem to be popular for that. The main reason seems to be for multiple VMs. Your numbers are quite aggressive though if you're comparing them to current specs. Haswell is supposed to be around a 15% cpu boost. Will broadwell make up the rest? Also 2014 would be a secondary revision of Maxwell. Have they projected something similar in tdp to the 650m today will match a GTX 680 within 2 years? I've read the links on that one, and that is a pretty bold prediction.
Quote:
You can get a laptop or iMac and have all of these things. You can even buy a Mac Pro enclosure, shut the MBP and sit it inside it and just pretend you have a 2009 8-core Mac Pro when people come over. They might wonder how you got SATA 6G and USB 3 but you can say you have PCI cards for that.
That comment of mine wasn't specific to the mac pro.
Quote:
Originally Posted by Marvin
In terms of 'halo products', that gets mentioned a few times and the analogy is to sports cars but it doesn't hold up. People don't aspire to own a 17" laptop or a Mac Pro. They are cumbersome and unattractive form factors. The 17" MBP didn't perform any better than the 15" so its appeal was purely down to screen size. The entry Mac Pros don't perform much better than the iMac and by the time you hit the ones that do at $3000-4000, they are too expensive.
Apple's halo products are the ones that people admire the most and those are the iPhone and iPad.
Would you really buy a porsche if the dealer wouldn't sell it to you with pirelli tires? I too grow tired of endless car analogies. The last funny one was an exhaust pipe joke from Futurama.
Jobs said a lot of things. He was a salesman, so it was to be expected.
I wouldn't say that's an apt description. There was an element of that but a salesman is solely there to sell you something with a value lower than you are expected to pay. Steve Jobs rationalised the product choices and there's a very important design process behind that.
Your numbers are quite aggressive though if you're comparing them to current specs. Haswell is supposed to be around a 15% cpu boost. Will broadwell make up the rest? Also 2014 would be a secondary revision of Maxwell. Have they projected something similar in tdp to the 650m today will match a GTX 680 within 2 years? I've read the links on that one, and that is a pretty bold prediction.
I don't expect Intel to stick to 15% year on year with their CPUs but I guess that will depend on how much they focus on the IGP. Say that they go 1.15 (2013), 1.30 (2014), 1.30 (2015), then it's about double by 2015. NVidia claims 75% increase this year but I'd say it's closer to 30% for the refresh. To match the 680, it would need something between a 3x-7x speedup, depending on what's being measured. They plan to more than double Kepler with Maxwell so if they hit 50% at the Maxwell refresh, that at least gets to the lower end of the requirement to match a 680 in 2015 and they definitely will by 2016 with Volta.
It is a bit optimistic but it's reasonable to suggest that laptops will be in that class of performance by that time.
I don't expect Intel to stick to 15% year on year with their CPUs but I guess that will depend on how much they focus on the IGP. Say that they go 1.15 (2013), 1.30 (2014), 1.30 (2015), then it's about double by 2015. NVidia claims 75% increase this year but I'd say it's closer to 30% for the refresh. To match the 680, it would need something between a 3x-7x speedup, depending on what's being measured. They plan to more than double Kepler with Maxwell so if they hit 50% at the Maxwell refresh, that at least gets to the lower end of the requirement to match a 680 in 2015 and they definitely will by 2016 with Volta.
It is a bit optimistic but it's reasonable to suggest that laptops will be in that class of performance by that time.
In terms of mainstream components, their focus appears to be on IGPs and power management. I can't find any truly interesting tech articles on the subject. The numbers claimed are once again pretty far out there, but none of the tech sites are very clear on sources. Personally I think by the time mobile graphics make it as far as you're suggesting, Apple will be in the process of pulling discrete graphics from more notebooks. As I've said they often do things just before a big change. Most people don't upgrade annually, so many will wait for the following year if the initial version is disappointing. The 27" imac is the only one where I'm totally sure discrete graphics will be around a while longer. It's supposed popularity suggests that it's being used by those with higher performance requirements. Other than that I don't see a terribly bright future for consumer desktops. I mentioned that I think things like ergonomics tend to be better, but when was the last time you saw anyone in their 20s that owned computer furniture? They have their notebook set down wherever and just look their phone much of the time (speaking partly from experience, I'm not that much older).
I am not an Apple worshipping corporatist. In my world, the customer is always right. Apple sold hundreds of thousands of 17 inch MBPs the last year they were available so there is demand. Fail to satisfy a market segment long enough and it will ALWAYS find an alternative. Apple can easily afford to service these customers so why lose them? Good enough for most people is a dismal philosophy. Successful companies don't cede market segments. They fight for the whole pie. Their stock has taken a ferocious beating this year so no one is immune to failure.
I still don't want integrated graphics in the iMac.
I thought you wanted a Mini?
I want the best possible stuff Apple can put in for the price they choose to have not "what is good enough for the masses."
At some point having an external GPU will be a bit of technology saved for extreme cases. There is nothing extreme about the markets the iMac services. Now if you re talking about the subject of this thread, the Mac Book Pro, the discussion gets even more interesting.
Apple still needs to make strides in power savings and at the same time increase performance significantly. Again we come back to what is more important , ultimate performance or simply an incremental improvement over todays hardware? If the next rev of the MBP came with a Intel chip that bested todays combo of an integrated chip and an external chip would you really object to such an APU based machine? Especially in the intro model.
To use a sports analogy, I want me a Joe Montana in my Mac and not a Steve Young even though a Steve Young is good enough.
Here is where we could easily be satisfied with a two model strategy from Apple. One motherboard with Integrated only and one with an external GPU. In the end I'm willing to bet the vast majority of sales will go to the integrated MBP, especially if that allows them to widen the price differential between the entry level machine and the top of the line machine.
This doesn't even get into the discussion of heterogeneous computing as hardware slowly moves to support that type of functionality widely. The reality is in the not to near future a discrete GPU might actually end up being a performance negative for many users.
What makes you think that they sold that many 17" laptops? I ask because they are seldom seen out in the wild.
It was reported as an estimate in 2011 by one of the main Apple analysts, that around 300,000 were sold that year. I find the the mentality of those who defend companies who don't satisfy customers bizarre, to say the least. I would not be a total surprise that at some point Apple will release a monster notebook because the pent up demand will warrant it.
It was reported as an estimate in 2011 by one of the main Apple analysts, that around 300,000 were sold that year. I find the the mentality of those who defend companies who don't satisfy customers bizarre, to say the least. I would not be a total surprise that at some point Apple will release a monster notebook because the pent up demand will warrant it.
I really doubt those numbers. For one thing you would walk into the local Apple store and have a hard time even finding the 17" MBP. That is a sign right there of limited demand.
As to defending Apple that isn't my intention, the fact is you cant satisfy all the people all the time. To believe otherwise is just foolish.
As to a monster notebook that might be possible. It should be noted that Apple hasn't really commented much on the 17" MBP going missing, that could mean they are working on something new to replace the old model. That would still be a surprise though as I just don't believe the demand was as strong as you have indicated. I've seen a bit of a trend away form really large notebook PC's at work as people simply don't want to lug around all that mass.
Having said that, I want Apple to use the best possible stuff available in general without anything sub-par.
Then they would never have a range of products to draw in a breath of customers. To survive as a company Apple needs to provide real choice from which all of the possible customers can choose from.
Comments
Relative to Apple's shipping volumes, the number of users you know would have to be at least in the tens of thousands to be significant though.
You suggested that customers who buy Mac Pros and 17" laptops are "their most loyal and biggest spending high end power users". These are the same people who would say that if they don't get what they want they'll either stop buying until they get what they want or switch to a Windows PC. They spend big once and then don't upgrade and if there isn't a specific upgrade, the loyalty is worth nothing.
Buyers of 15" Macbook Pros are currently Apple's most loyal high-end power users and as such get the foremost attention from Apple.
If the prices drop in June and they don't bring back the 17", the people waiting will just have to deal with the 15". Apple has to make a higher screen resolution again and they might not be able to get the yields good enough to justify doing for such a low volume of customers.
Apple has already stopped selling the Mac Pro across Europe and this will be the case for the next few months. The 17" MBP has been out of sale for about a year now. If these events had a significant impact to Apple, you can bet they'd have done something about it.
It may be significant to some potential buyers but you use whatever tools are available.
They need to fix some of the quality problems:
http://www.digitaltrends.com/computing/macbook-pro-retina-lg-class-action-suit-started/
but these things happen. Now that they have the manufacturing line setup for Retina displays, they shouldn't have the same problems they did before and they have no reason to change the design again.
Who is making them now?
I'd expect the same suppliers but this is obviously a fairly new process for all of them.
Quote:
Originally Posted by Marvin
It always stays at a couple no matter how many examples come along. It must be at least a handful by now.
http://www.tomshardware.co.uk/forum/360597-28-upgrade-macpro
http://forums.macrumors.com/showthread.php?t=1505370
http://forums.macrumors.com/showthread.php?t=781908
http://macperformanceguide.com/Mac-MacPro-upgrade.html
http://www.macobserver.com/tmo/article/step_by_step_mac_pro_processor_upgrade
http://www.ifixit.com/Answers/View/118042/Mac+Pro+1.1+Change+out+CPU
http://store.apple.com/uk/question/answers/mac/can-i-upgrade-the-cpu-on-mac-pro/Q2DCC7FHTU2UUT9X2
http://forum.netkas.org/index.php?topic=3829.0
http://www.staze.org/mac-pro-11-cpu-upgrade/
http://forums.appleinsider.com/t/144215/upgrading-a-2006-macpro-2-66-ghz-machine
https://discussions.apple.com/thread/3969568?start=0&tstart=0
https://discussions.apple.com/thread/4894228?start=0&tstart=0
https://discussions.apple.com/thread/4074230?start=0&tstart=0
https://discussions.apple.com/thread/4685505?start=0&tstart=0
http://forums.macrumors.com/showthread.php?t=1375414
http://forums.macrumors.com/showthread.php?t=1540397
http://forums.macrumors.com/showthread.php?t=1533098
http://forums.macrumors.com/showthread.php?t=1419182
http://www.dpreview.com/forums/post/41988597
http://www.waitingoutside.com/2010/02/upgrading-a-2006-apple-mac-pro-11-with-2-xeon-x5355-processors/
http://www.gearslutz.com/board/music-computers/668475-mac-pro-1-1-cpu-upgrade-advice.html
I'm curious how much of the market thinks like that, but I noticed the people upgrading 1,1s listed in there. Many may not be original owners, but regardless they really aren't a substantial market factor.
Quote:
You know what the problem is. When you hold onto hardware for 6 years, the resale value is gone so it's a much better option to upgrade components than to buy new.
I have told people before to treat computers as sunken costs. If they're too concerned with residual value, it's likely that they spent too much initially. Considering the number of 1,1s, some of those could also be secondary owners who don't really drive sales anyway.
Quote:
That doesn't explain the desire to upgrade.
I think many of those upgrading really don't know what they need. In the case of the 4,1s, they probably do so due to price. Intel launched the W3680 at $1000. The 5,1 was over $4000 by the time you added ram, tax, etc. Let's say they bought a 4,1 in 2009 and regularly upgraded every 2-3 years. Most of these guys were probably upgrading from 1,1s aside from new customers. It's unlikely that most of them would go for a 2010 model the following year as it was basically a revision B. They dropped in new cpus and gpus. Nothing else changed. Like the other vendors the sub $3000 segment retained the same cpus they used the prior year, so unless they both went for the upgrade and increased their budget, there wasn't anything of significance. Given the Sandy Bridge E delay and price drops from intel that didn't really filter down through most of the oems to the customer level, is it really that surprising that they made simple upgrades there as stopgap measures?
In terms of actual sales, it is a matter of how many people would buy an updated version this year or next year. Anyway I've stated before that the market for workstations will hold longer regardless of whether Apple makes one. It's not likely to drive their growth either way, but it is somewhat healthier than the low end of the PC market. Sometimes you get a bit hyperbolic with your assertions.
There's little chance that the people upgrading 1,1s are ever going to be purchasing 12 core models. If they started with the dualcore 2.66 models, they spent $2500 + whatever on memory at the time. The fbdimms were quite expensive at the time, but not all of them will be looking at much higher pricing tiers. If anything the past few years, I've considered the lower mac pros models to represent poor value unless you're locked in. I've never suggested Apple would be doomed without the mac pro. Really that kind of market should be more interesting for smaller companies.
Quote:
It depends on what technology comes along:
http://phys.org/news/2013-01-qubit-bodes-future-quantum.html
Articles and speculation on quantum computing theory go back many years. Considering your wealth of links (not sure how you find so many) you probably already know this. I find them more interesting when they're close to being something that could be implemented in the near future. Right now gpu technologies interest me more than anything due to the way they can handle parallel processes.
Quote:
The CUDA revolution started with just a single research project:
http://www.tomshardware.com/reviews/ian-buck-nvidia,2393.html
I didn't know about that. I noticed it details some of the features in CUDA relative to OpenCL. I like some of the questions asked there, like the one below. What is interesting to me is that certain things that would have lagged on the 2009 era 8 cores can be driven by some of these gpu technologies when implemented. In a couple areas it's really interesting if you deal with a lot of still image and video data in raw formats. It allows you to apply quite a bit more adjustment without baking things like gamma and debayering. Of course it's software dependent, but I can see a lot of this stuff emerging.
Quote:
With OpenCL, you gain the advantage of cross-platform support, but lose automated tools, such as memory management, that are found with CUDA. It seems that as a scientist, you'd want to decrease your startup development costs, but at the same time, you'd want support for multiple platforms. What's the best way to reconcile this challenge?
Quote:
The smaller form factors will keep eroding away the large ones, especially the more that GPU computing takes hold:
http://www.xbitlabs.com/news/graphics/display/20130412175120_Nvidia_Next_Generation_Maxwell_Architecture_Will_Break_New_Grounds.html
I never disagreed with that, although I wonder how aggressively Apple will move to IGPs only. They often move 1-2 cycles earlier than technology is really ready figuring that whatever people need more will just upgrade the following cycle as it's not possible to expect everyone will upgrade annually. 3 years is probably a typical assessment. Many oems set up their available warranties to go as far as 3 years out including Apple.
Quote:
Servers will always exist but they don't use the tower form factor.
I'm not as hung up on form factors as you seem to think. I don't know why this comes up again and again. When I refer to a tower, it's what is there today. There are certain features I like such as independent keyboards as they allow more freedom in terms of placement. I have certain preferences in terms of displays and peripheral devices, but sometimes it's just a matter of flexibility.
Quote:
Originally Posted by Marvin
I'd expect the same suppliers but this is obviously a fairly new process for all of them.
Missed this one. I suspect you'll eventually see some kind of reference implementation, which will make them cheaper. 1080 is not uncommon today. You even find it on 13" models. I personally placed far more value in the improvements in viewing angles and the reduced color temperature. The high res 15" models measured roughly 8000K whites for me. The rMBP is more like D65. The stock models were a bit low in resolution for my taste, but the high res ones were good. Of course we all have different priorities. I just hated the super cool whites and weird screen gamma.
I am not an Apple worshipping corporatist. In my world, the customer is always right. Apple sold hundreds of thousands of 17 inch MBPs the last year they were available so there is demand. Fail to satisfy a market segment long enough and it will ALWAYS find an alternative. Apple can easily afford to service these customers so why lose them? Good enough for most people is a dismal philosophy. Successful companies don't cede market segments. They fight for the whole pie. Their stock has taken a ferocious beating this year so no one is immune to failure.
I don't think anyone would consider a computer to be an item to eventually throw away. Some of the MP1,1 owners tried to sell it and couldn't so that led them to upgrade the parts, which suggests they were the original owners. Here's another original owner wondering why his machine won't run Mountain Lion:
https://discussions.apple.com/thread/4879441?start=0&tstart=0
"I can't just throw my Mac Pro in garbage because I can't upgrade. Why do I have to upgrade? Well, Sibelius 7 works only in Mountain Lion, and soon, many other software will.
What do I have to change in my mac pro to make it compatible for upgrading to the latest version of mac OS?
I think Apple should be proud to help 7 year old expensive computers, and not just tell you "buy a new one, your computer is old". I guess that's how they can make more money"
This just seems to be the recurring theme of the Mac Pro owner.
But they're upgrading to chips that are slower or the same as new machines that are available.
The Westmere 3.33GHz MP (2012) costs $2999. This is the same chips used in the upgrade thread:
http://forums.appleinsider.com/t/156913/my-new-mac-pro-a-2009-with-a-intel-xeon-westmere-3-33-6-cores
So, after having a computer for 4 years, rather than sell it for $1000 and pay $2000 to upgrade, the choice is to pay $610 and get some thermal paste to do the upgrade or worst case, consider a $1500 upgrade from OWC. Anything to save paying Apple the full amount. I don't blame people for wanting to save $500-1400 but these people can't be described as big spenders or loyal to Apple.
The reason Apple charges so much is because people aren't buying from them often enough.
I don't believe this is the case. You've seen the sales figures for the workstation market. By 2015 they'll have the equivalent of a GTX 680 in a MBP and easily double the performance of the CPU, which is currently at 8-core 2009 MP level and they'll have DDR4 memory supporting up to 32GB. 512GB SSDs will be very affordable. More and more people will migrate down to nicer form factors and more affordable hardware.
I think they'll use Haswell's IGP in more machines. Even the entry iMac can get away with it, although it would have to be the GT3 version. I'd like to see them get rid of the 21.5" and bring out a 27" for $1299.
You can get a laptop or iMac and have all of these things. You can even buy a Mac Pro enclosure, shut the MBP and sit it inside it and just pretend you have a 2009 8-core Mac Pro when people come over. They might wonder how you got SATA 6G and USB 3 but you can say you have PCI cards for that.
It's unlikely they'll lose all of the 17" customers. It's the same deal with the glossy/matte MBP. Steve Jobs mentioned that most people were buying glossy models so they dropped the matte versions. The Retina display makes it pretty much a non-issue now. 17" MBP owners have the option to buy a much larger external display - you even get USB-powered 20"+ displays.
That isn't because of how they're doing though. That's because of how people think they are doing. They are still making record sales every quarter and are still the most profitable company in the world. You're right that no one is immune to failure but the number 1 company in the world is the last company to be concerned about.
I still don't want integrated graphics in the iMac. I want the best possible stuff Apple can put in for the price they choose to have not "what is good enough for the masses."
To use a sports analogy, I want me a Joe Montana in my Mac and not a Steve Young even though a Steve Young is good enough.
I disagree. Top of the line halo products have a benefit that goes beyond that market segment of audio visual professionals, something Steve Jobs understood. He always wanted to be at the top, without excuses or rationalizations,or telling customers to buy something they really weren't excited about and just make due.
You know what else he understood:
Even way back in 2007 they said that 2/3 of their lineup sales were notebooks and at that point, the PC industry was at 40% notebooks.
In terms of 'halo products', that gets mentioned a few times and the analogy is to sports cars but it doesn't hold up. People don't aspire to own a 17" laptop or a Mac Pro. They are cumbersome and unattractive form factors. The 17" MBP didn't perform any better than the 15" so its appeal was purely down to screen size. The entry Mac Pros don't perform much better than the iMac and by the time you hit the ones that do at $3000-4000, they are too expensive.
Apple's halo products are the ones that people admire the most and those are the iPhone and iPad.
Quote:
Originally Posted by Marvin
It's unlikely they'll lose all of the 17" customers. It's the same deal with the glossy/matte MBP. Steve Jobs mentioned that most people were buying glossy models so they dropped the matte versions. The Retina display makes it pretty much a non-issue now. 17" MBP owners have the option to buy a much larger external display - you even get USB-powered 20"+ displays.
That isn't because of how they're doing though. That's because of how people think they are doing. They are still making record sales every quarter and are still the most profitable company in the world. You're right that no one is immune to failure but the number 1 company in the world is the last company to be concerned about.
Jobs said a lot of things. He was a salesman, so it was to be expected.
Quote:
I don't believe this is the case. You've seen the sales figures for the workstation market. By 2015 they'll have the equivalent of a GTX 680 in a MBP and easily double the performance of the CPU, which is currently at 8-core 2009 MP level and they'll have DDR4 memory supporting up to 32GB. 512GB SSDs will be very affordable. More and more people will migrate down to nicer form factors and more affordable hardware.
Well technically you can get 32GB into a notebook today. Thinkpads seem to be popular for that. The main reason seems to be for multiple VMs. Your numbers are quite aggressive though if you're comparing them to current specs. Haswell is supposed to be around a 15% cpu boost. Will broadwell make up the rest? Also 2014 would be a secondary revision of Maxwell. Have they projected something similar in tdp to the 650m today will match a GTX 680 within 2 years? I've read the links on that one, and that is a pretty bold prediction.
Quote:
You can get a laptop or iMac and have all of these things. You can even buy a Mac Pro enclosure, shut the MBP and sit it inside it and just pretend you have a 2009 8-core Mac Pro when people come over. They might wonder how you got SATA 6G and USB 3 but you can say you have PCI cards for that.
That comment of mine wasn't specific to the mac pro.
Quote:
Originally Posted by Marvin
In terms of 'halo products', that gets mentioned a few times and the analogy is to sports cars but it doesn't hold up. People don't aspire to own a 17" laptop or a Mac Pro. They are cumbersome and unattractive form factors. The 17" MBP didn't perform any better than the 15" so its appeal was purely down to screen size. The entry Mac Pros don't perform much better than the iMac and by the time you hit the ones that do at $3000-4000, they are too expensive.
Apple's halo products are the ones that people admire the most and those are the iPhone and iPad.
Would you really buy a porsche if the dealer wouldn't sell it to you with pirelli tires? I too grow tired of endless car analogies. The last funny one was an exhaust pipe joke from Futurama.
I wouldn't say that's an apt description. There was an element of that but a salesman is solely there to sell you something with a value lower than you are expected to pay. Steve Jobs rationalised the product choices and there's a very important design process behind that.
I don't expect Intel to stick to 15% year on year with their CPUs but I guess that will depend on how much they focus on the IGP. Say that they go 1.15 (2013), 1.30 (2014), 1.30 (2015), then it's about double by 2015. NVidia claims 75% increase this year but I'd say it's closer to 30% for the refresh. To match the 680, it would need something between a 3x-7x speedup, depending on what's being measured. They plan to more than double Kepler with Maxwell so if they hit 50% at the Maxwell refresh, that at least gets to the lower end of the requirement to match a 680 in 2015 and they definitely will by 2016 with Volta.
It is a bit optimistic but it's reasonable to suggest that laptops will be in that class of performance by that time.
Quote:
Originally Posted by Marvin
I don't expect Intel to stick to 15% year on year with their CPUs but I guess that will depend on how much they focus on the IGP. Say that they go 1.15 (2013), 1.30 (2014), 1.30 (2015), then it's about double by 2015. NVidia claims 75% increase this year but I'd say it's closer to 30% for the refresh. To match the 680, it would need something between a 3x-7x speedup, depending on what's being measured. They plan to more than double Kepler with Maxwell so if they hit 50% at the Maxwell refresh, that at least gets to the lower end of the requirement to match a 680 in 2015 and they definitely will by 2016 with Volta.
It is a bit optimistic but it's reasonable to suggest that laptops will be in that class of performance by that time.
In terms of mainstream components, their focus appears to be on IGPs and power management. I can't find any truly interesting tech articles on the subject. The numbers claimed are once again pretty far out there, but none of the tech sites are very clear on sources. Personally I think by the time mobile graphics make it as far as you're suggesting, Apple will be in the process of pulling discrete graphics from more notebooks. As I've said they often do things just before a big change. Most people don't upgrade annually, so many will wait for the following year if the initial version is disappointing. The 27" imac is the only one where I'm totally sure discrete graphics will be around a while longer. It's supposed popularity suggests that it's being used by those with higher performance requirements. Other than that I don't see a terribly bright future for consumer desktops. I mentioned that I think things like ergonomics tend to be better, but when was the last time you saw anyone in their 20s that owned computer furniture? They have their notebook set down wherever and just look their phone much of the time (speaking partly from experience, I'm not that much older).
Apple still needs to make strides in power savings and at the same time increase performance significantly. Again we come back to what is more important , ultimate performance or simply an incremental improvement over todays hardware? If the next rev of the MBP came with a Intel chip that bested todays combo of an integrated chip and an external chip would you really object to such an APU based machine? Especially in the intro model.
Here is where we could easily be satisfied with a two model strategy from Apple. One motherboard with Integrated only and one with an external GPU. In the end I'm willing to bet the vast majority of sales will go to the integrated MBP, especially if that allows them to widen the price differential between the entry level machine and the top of the line machine.
This doesn't even get into the discussion of heterogeneous computing as hardware slowly moves to support that type of functionality widely. The reality is in the not to near future a discrete GPU might actually end up being a performance negative for many users.
Having said that, I want Apple to use the best possible stuff available in general without anything sub-par.
Quote:
Originally Posted by wizard69
What makes you think that they sold that many 17" laptops? I ask because they are seldom seen out in the wild.
It was reported as an estimate in 2011 by one of the main Apple analysts, that around 300,000 were sold that year. I find the the mentality of those who defend companies who don't satisfy customers bizarre, to say the least. I would not be a total surprise that at some point Apple will release a monster notebook because the pent up demand will warrant it.
I really doubt those numbers. For one thing you would walk into the local Apple store and have a hard time even finding the 17" MBP. That is a sign right there of limited demand.
As to defending Apple that isn't my intention, the fact is you cant satisfy all the people all the time. To believe otherwise is just foolish.
As to a monster notebook that might be possible. It should be noted that Apple hasn't really commented much on the 17" MBP going missing, that could mean they are working on something new to replace the old model. That would still be a surprise though as I just don't believe the demand was as strong as you have indicated. I've seen a bit of a trend away form really large notebook PC's at work as people simply don't want to lug around all that mass.
Then they would never have a range of products to draw in a breath of customers. To survive as a company Apple needs to provide real choice from which all of the possible customers can choose from.