Every year Apple should release a new Mac Pro, a single, top of the line model.
Apple should then continue to sell the previous years models as a lower priced option.
This would would make them much more enterprise friendly.
Enterprise customers love predictable upgrade cycles and standardized configurations.
They also love being able to buy the same machine they bought 2 years ago and having fewer varying models.
2015 Mac Pro $3000
2014 Mac Pro $2500
2013 Mac Pro $2000
I dont think their unit sales numbers nor the drop in technology costs YoY warrant that as a viable option.
The current direction of Apple is something that I have been wondering about for a while now, specifically the lack of focus on the high-end professional consumers served by the Mac Pro. The iOS platform has been incredibly lucerative and successful for Apple, however there is an obvious question about the future of the Mac hardware platform.
In the last five years, there's been a shift which has seen huge adoption of Mac laptops and desktops by a wide variety of people. It would seem this has just as much to do with the world's detachment from Microsoft as it does the adoption of Mac hardware. Now that the majority of a lot of people's "computing" is via web and mobile apps, the architecture of the interface matters a lot less. We've gone from a world where Windows was the dominant paradigm because it ran all the software, to a world where Windows is the exceptional paradigm. One goes with Windows if they have software that doesn't run as a web app. Otherwise, a Mac will do fine. The triumph of Apple engineering and the computer as a consumer device.
The concern is three-fold. The first is that iOS occupies a less defensible position in the market. As devices have shorter lifespans and apps are often replicated across platforms, the iOS is less sticky than the MacOS. Consumers seem to go from iPhones to Androids, now they're trying Blackberries again. As long as the device 1) makes phone calls, 2) receives/sends email, 3) receives/sends texts, 4) has a decent camera, and 5) has Facebook/Instagram/Twitter/Spotify/etc, I don't think people are really bothered about the device as much as might be considered. Prosumers, yes. The people who read this website, probably. But a lot of people really don't seem too bothered. And Apple didn't help with the cable shift of the last iPhone. All of a sudden, iPhones went from a gated community, to a fragmented community. When you have to carry a cable for an iPhone 5 like you do with a Galaxy, they all just become phones again.
The second concern is the proliferation of hardware. MacBook Pros with retina displays and without retina displays. iPads with retina displays and iPads without retina displays. This is a significant break from the Jobsian narrow ecosystem that made Apple so successful. There used to be one iPad. One MacBook. One product at each level. That is a luxury manufacturer logic. You do not get a choice, for part of what you are paying the premium for is curation of the device. Now with Tim, we see new product segments, and point-to-point competition. Business strategy that is technically correct but again breaks the ecosystem structure. This may seem minor, but along with the dividend payment, Apple is being run as a business rather than as a product religion. Last time this happened, it wasn't long before the company ended up replicating the dominant industry logic.
That comes to the third point, which is the very broken pro user market. Final Cut X was the first debacle. Apple took a long time to capture that attention, and actually get AVID on the run, and Adobe nearly out of the game. They were almost home-free on the pro video editing market. Almost... and then Final Cut X and utter disaster for the high-end market. Now, there's been a big shift back to AVID, and another shift to Adobe. From a business point of view, these are niche segments that don't pay for a lot of hardware in the end. From a spiritual point of view, those users are at the heart of the platform. The storytellers who have the need for the best hardware. Who's time is so valuable, they upgrade machines whenever there's a speed-bump that can save them an appreciable amount of time. Now the lack of focus on that segment from an Apple that is so cash rich can only mean that the segment is not considered important to the company. Huge mistake, and probably the biggest red flag here. Even if the pro video market was a complete loss leader for Apple, the pro video market is the defence industry of the computer industry... it's where all the black budgets and high-end stuff is created, tested, and purchased.
I think Tim is doing a fantastic job overall with Apple. However, the three points above are three very obvious non-Apple things that his Apple is doing, and should cause any investor cause for concern. Apple will always be around, and will always be producing hardware, but whether or not it can maintain its mystique and related margins has to do with the company's continued ability to capture the imaginations of a relatively small number of VERY influential content creators. If I were one of those creators today, I would be asking if Apple really cares. And that is a big big red flag.
Every year Apple should release a new Mac Pro, a single, top of the line model.
Apple should then continue to sell the previous years models as a lower priced option.
The chips don't come out that often, and they're already doing that.
Originally Posted by jnr1212
One MacBook.
There was never just one MacBook.
Now with Tim…
You mean with Steve. Everything out now was approved by him.
This may seem minor, but along with the dividend payment, Apple is being run as a business rather than as a product religion.
I rather think you can't say 'rather' yet. Steve did that, too.
Final Cut X was the first debacle.
And there goes your opinion.
…that segment from an Apple that is so cash rich…
Thing is, that segment isn't cash rich.
Even if the pro video market was a complete loss leader for Apple, the pro video market is the defence industry of the computer industry... it's where all the black budgets and high-end stuff is created, tested, and purchased.
Apple doesn't run "loss leaders". Apple doesn't run losses. Sounds like you want them to run more like a business than a product-oriented group…
You save the space that the Mac Pro would have taken up to begin with. There's more cabling with external storage + 2-3 peripherals but you can put them in a box if you like.
The size of the card isn't important. The fastest cards are not even 2x the performance of the iMac's GPU but cost $1000 or more each. Not to mention they are outside the safe power limits of the Mac Pro too.
The Radeon 7990 is 375W. I'm not sure how many professionals would put a 375W GPU into slots with a 300W power allocation but I doubt they'd be in business long.
The Mac Pro can obviously hold higher wattage cards than the iMac but they don't perform significantly better in practise. OS X doesn't support the highest-end GPUs on the software side either.
With the high-end GPUs out of the equation, you're just left with expansion cards, which work fine over Thunderbolt.
Marvin, my 2010 Mac Pro sits under my work area and takes up NO desk space. Furthermore, the space taken is vertical. Even if it were on the desk (few are) it would take up the space of an AJA ioXT HD interface and cables and perhaps a Thunderbolt hub minus the cables. The Mac Pro can even be placed in the next room, which it often is in professional settings.
As for the Radeon 7990's power requirements, the first PCI 3 GPU, we are talking about a new Mac Pro here. Is there any reason a larger power supply could no be fitted into a new Mac Pro design? There are many cards of lower price than the 7990 that are also quite large with great performance, particularly with CUDA. PCI 3 will provide great advantages for CUDA and OpenCL, card based technologies that greatly speed up video rendering and ray tracing, a staple of animation. Adobe is making amazing use of CUDA for video rendering. As for Apple software support, I imagine they have the engineering talent and budget (it would not require much, specs are published) to implement version 4 OpenGL. I can't imagine they would not further development and implementation of OpenCL, the standard they have been pushing.
And to working fine over Thunderbolt, as of now. the expansion card boxes are much more limited in wattage and card space than even the 2010 Mac Pro. How do you put a Radeon 7990 in them? And you will eventually hit the speed wall of even Thunderbolt 2 (not yet even available) compared to internal PCI 3 which is shipping. Plus another (big) box on your desk. And more cables. And more fan noise. And whoops, I accidentally pulled the plug. Not a pro environment in my opinion. iMacs for those who are satisfied, they have their place for many pro uses. But for most, they will not replace the Mac Pro workstation concept, in functionality, speed and elegance, where they are truly needed and appreciated.
Apple should offer a "MacPro" PLATFORM, upon which tech savvy power users can customize everything in their computer. Up to the tiniest details from a choice of modules.
This would cater that customer base best because they could really tailor the machine to their needs.
It is, in my opinion, pointless to offer "ready made" choices.
That's the other thing about Apple's direction with the computer line. It's often thought that Apple should offer what the potential buyer wants (e.g the customer is always right) but before the iMac, nobody asked for one. Some people will then mutter 'exactly, nobody asked for one' but look at the sales stats less than 2 years after it was introduced:
It made up over 50% of their sales. Tim Cook's role in overhauling Apple's supply chain had a very important implication for their computer designs too. One thing he said was:
"You kind of want to manage [Apple's inventory] like you're in the dairy business. If it gets past its freshness date, you have a problem."
When you offer customers a lot of choice then you can't always guarantee that you can turn over the inventory quickly. You might offer 10 GPU options and buy a bunch of stock and then 90% of buyers only stick with the entry GPU. This is exactly why the PC component industry struggles so much and why Intel/AMD/NVidia all want to move towards SoC. It's also why everyone seems to want a tablet now and their own operating system. It's a far more profitable business model to sell you an entire finished product than a part of one.
The ideal scenario (from the seller's perspective) where every manufacturer offers one device per category is not going to happen but out of all manufacturers, Apple will move the closest to this ideal because this is how they choose to do things.
They narrowed GPU options on the Mac Pro to one manufacturer and two models. Intel makes dozens of CPUs but you get to choose from 2-3 per main model. They removed the hardware RAID card BTO option too.
Apple's direction is not to expand choice unless it's clear there will be a huge benefit but to streamline their manufacturing and reduce inventory. That setup is the most profitable.
With this in mind, people then point out that it shouldn't be all about counting the beans and that's not what it used to be about and if they don't get back there, the company with $147b in the bank will go bankrupt eventually. Well, that's what nearly drove them to bankruptcy before Tim came along and turned it around by streamlining their inventory.
It's much easier to tell what they will do going forward when you look at it from their point of view rather than from the perspective as a buyer. As a potential buyer, it's easy to rattle off fictional hardware that you'd buy - 'sure if they made a 19" laptop with dual GTX 680Ms, a 1TB SSD (heck make it dual 1TB in RAID 0) and a giant 300W power brick and it cost $6000, I'd buy it or a Mac Pro with 9 PCI slots and PCI SSD and quad CPUs with 256GB RAM'. Except you probably wouldn't buy it, you'd buy the lowest possible model with your minimum requirements (e.g it's a tower or it has a big screen) and try and upgrade it cheaper without paying Apple more money that you have to.
With the Mac Pro, people talk about money being no object and almost every time someone has a Mac Pro, it's an ancient one they are for some reason dying to upgrade now that Apple doesn't sell one or doesn't sell one they think is good value. If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
The people who work at Apple have been at this a very long time and they've seen technology and companies grow and die over 30+ years. That's a lot of experience. They don't always exercise the best judgement in everything they do but they are talented people and more often than not, they have driven technology in the right direction and that gives them a very long term vision that buyers don't need to have. Buyers don't think about what will happen in 10 years, buyers think about what the current costs and needs are.
Apple knows that in a few years, computers are going to be 10x faster than they are now no matter if it's distributed more to the GPU than the CPU and they are in the business of selling computers. So, they need to figure how to make it compelling enough to keep buying new computers while almost satisfying the current needs of the buyer. When they launched the iPad, they could have put in 1GB of RAM, all sorts of higher spec things but you wouldn't necessarily buy another one and that's no good. When buyers aren't compelled to keep buying, that's when sellers stop growing and this is where the PC industry is now.
Their strategy of streamlining has worked great so far, Tim won't undo supply chain efficiency unless there's a good reason to. I think it's clear that they will be getting rid of the old style laptops and a few years down the line might even consolidate the MBP and Air lines - remember 10x faster than they are now. They want to be soldering RAM onto the machines so that trend will probably continue where it makes sense - when DDR4 comes probably next year with double the density, they can solder 8/16/32GB onto iMacs. The Mac Pro might not impact the supply chain enough to bother locking it down but there's no harm in them doing that either.
Well, they are streamlining themselves right out of the professional marketplace. There is a huge movement to Adobe software in video editing. That is occurring now, not ten years from now. And I speak as the former exclusive Avid dealer for Hawaii. Final Cut X was a watershed and Adobe is taking advantage of it. Their products are cross platform and actually run better on PC's because of software optimaization and access to the best current hardware, particularly GPUs. This is also true in animation, particularly ray tracing., There are many of us, myself included, who would never consider moving to Windows, but large numbers are because of Apple's intransigence regarding Mac Pro workstation development. The high end hardware and software is available on Windows workstations like the HP's, and they are either moving or seriously considering it. We can theorize about the future of computing until we are blue in the face, but that does not change the fact that a new Mac Pro, not that much different than the old one in form but with up to date technology is needed now to stop the bleeding of professionals to Windows where they can configure to the current state of the art today and can count on future development of such systems. Why do you think Mac Pro diehards are so torn and emotional? We want to stay with the Mac, but can Apple give us what we need now and in the future?
Have to say it was a very good article by DED. I have been wondering what was coming ever since Tim C. said the Power Mac would get a refresh.
And I have been collecting a few older Power PC G5 dual units. They run fine, even the G4 dual tower. But the older OS systems will only run older browsers. So when the new units come out, the intel machines should finally come down in price.
Currently you can get a decent G5 dual for about $150..
The Mac Pro can even be placed in the next room, which it often is in professional settings.
With optical Thunderbolt cables, you can have peripherals across the street. A slight tripping hazard but only to non-professionals who don't watch where they're going.
As for the Radeon 7990's power requirements, the first PCI 3 GPU, we are talking about a new Mac Pro here. Is there any reason a larger power supply could no be fitted into a new Mac Pro design?
There are many cards of lower price than the 7990 that are also quite large with great performance, particularly with CUDA. PCI 3 will provide great advantages for CUDA and OpenCL, card based technologies that greatly speed up video rendering and ray tracing, a staple of animation. Adobe is making amazing use of CUDA for video rendering.
Why is it so important that they are large? The speed doesn't scale directly with the size of the cards. Apple could put a larger power supply in and support higher powered cards and run the fans a bit faster but they made a set of compromises when they designed it as they do with every machine. They want it to be quiet in operation so it can't get too hot. That's obviously why they went a bit nuts with the heatsinks. They have to consider what the potential gains would be by going that bit further.
When you look at one of the highest end GPUs - the latest Titan GPU - you can see that the OpenCL score for that in one benchmark is 1421:
For all of the extra effort, heat, power draw, size and cost, that's all you get. This year's iMac GPU will be 30% faster again so if you stocked up a Mac Pro when the Titan came out for say $3500 + display, you only end up 35-45% faster than this year's iMac and they have a sleek desktop whereas you have a power hungry tower. You're not going to buy another $1000 GPU after just 1 year.
If they did allow you to run 3 or more high-end GPUs, there's some advantage but again, they cost $1000 each and Apple wouldn't do this because so few people would take advantage of it. When you need lots of compute power, you can network multiple machines together.
And you will eventually hit the speed wall of even Thunderbolt 2 (not yet even available) compared to internal PCI 3 which is shipping.
Workstation chipsets don't have PCI 3 support - that doesn't come until Haswell chipsets. Some third party manufacturers have added it but just because a spec exists, doesn't mean that suddenly becomes the minimum requirement. These comparisons tend to go the route of comparing some consumer option to the highest possible option there is and the purpose is obviously to suggest the consumer option is therefore not up to the job. The false assumption being made is that there is a common task somewhere that's using the full extent of whatever spec exists e.g if it's possible to put 128GB of RAM in something, that must mean someone is using 128GB of RAM therefore the iMac is useless because it only support 32GB. That idea doesn't hold up in practise.
Plus another (big) box on your desk. And more cables. And more fan noise. And whoops, I accidentally pulled the plug. Not a pro environment in my opinion. iMacs for those who are satisfied, they have their place for many pro uses. But for most, they will not replace the Mac Pro workstation concept, in functionality, speed and elegance, where they are truly needed and appreciated.
The faster machines become, the less reliant people will be on specialised PCI cards for processing - it'll be done natively on the CPU/GPU. Apple was demoing native 4K on an iMac at NAB this year. IO standards will move to optical soon. The workstation concept won't disappear overnight but these advances will keep eroding the userbase and eventually the sellers will stop selling them.
Think 6 years down the line, say that SSDs go down in price 25% every year. Crucial has 1TB at $600 just now. That means that a 1TB drive will be $107, essentially today's HDD prices. CPU/GPU power will be about 8x faster with one likely weighted more than the other. Memory density will mean 32GB RAM is inexpensive. IO ports will exceed 50Gbps each.
You won't need Red Rocket cards or other processing cards. The internal GPUs in laptops will match current GTX Titans.
At that point, what will you need to be putting inside a tower?
a new Mac Pro, not that much different than the old one in form but with up to date technology is needed now to stop the bleeding of professionals to Windows
Windows PC sales are down vs last year. Apple suffered less than PC manufacturers. It's obvious people are moving to Premiere but not to Windows PCs in large numbers. I imagine some have moved to HP/Dell workstations but the phrase 'bleeding professionals to Windows' again suggests that professionals only use Mac Pros, which isn't the case. There are about 90 million PCs sold worldwide every quarter. Workstations are 1 million of those. If there's any bleeding, it's little more than from a paper cut.
That's the other thing about Apple's direction with the computer line. It's often thought that Apple should offer what the potential buyer wants (e.g the customer is always right) but before the iMac, nobody asked for one. Some people will then mutter 'exactly, nobody asked for one' but look at the sales stats less than 2 years after it was introduced:
It made up over 50% of their sales. Tim Cook's role in overhauling Apple's supply chain had a very important implication for their computer designs too. One thing he said was:
"You kind of want to manage [Apple's inventory] like you're in the dairy business. If it gets past its freshness date, you have a problem."
When you offer customers a lot of choice then you can't always guarantee that you can turn over the inventory quickly. You might offer 10 GPU options and buy a bunch of stock and then 90% of buyers only stick with the entry GPU. This is exactly why the PC component industry struggles so much and why Intel/AMD/NVidia all want to move towards SoC. It's also why everyone seems to want a tablet now and their own operating system. It's a far more profitable business model to sell you an entire finished product than a part of one.
The ideal scenario (from the seller's perspective) where every manufacturer offers one device per category is not going to happen but out of all manufacturers, Apple will move the closest to this ideal because this is how they choose to do things.
They narrowed GPU options on the Mac Pro to one manufacturer and two models. Intel makes dozens of CPUs but you get to choose from 2-3 per main model. They removed the hardware RAID card BTO option too.
Apple's direction is not to expand choice unless it's clear there will be a huge benefit but to streamline their manufacturing and reduce inventory. That setup is the most profitable.
With this in mind, people then point out that it shouldn't be all about counting the beans and that's not what it used to be about and if they don't get back there, the company with $147b in the bank will go bankrupt eventually. Well, that's what nearly drove them to bankruptcy before Tim came along and turned it around by streamlining their inventory.
It's much easier to tell what they will do going forward when you look at it from their point of view rather than from the perspective as a buyer. As a potential buyer, it's easy to rattle off fictional hardware that you'd buy - 'sure if they made a 19" laptop with dual GTX 680Ms, a 1TB SSD (heck make it dual 1TB in RAID 0) and a giant 300W power brick and it cost $6000, I'd buy it or a Mac Pro with 9 PCI slots and PCI SSD and quad CPUs with 256GB RAM'. Except you probably wouldn't buy it, you'd buy the lowest possible model with your minimum requirements (e.g it's a tower or it has a big screen) and try and upgrade it cheaper without paying Apple more money that you have to.
With the Mac Pro, people talk about money being no object and almost every time someone has a Mac Pro, it's an ancient one they are for some reason dying to upgrade now that Apple doesn't sell one or doesn't sell one they think is good value. If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
The people who work at Apple have been at this a very long time and they've seen technology and companies grow and die over 30+ years. That's a lot of experience. They don't always exercise the best judgement in everything they do but they are talented people and more often than not, they have driven technology in the right direction and that gives them a very long term vision that buyers don't need to have. Buyers don't think about what will happen in 10 years, buyers think about what the current costs and needs are.
Apple knows that in a few years, computers are going to be 10x faster than they are now no matter if it's distributed more to the GPU than the CPU and they are in the business of selling computers. So, they need to figure how to make it compelling enough to keep buying new computers while almost satisfying the current needs of the buyer. When they launched the iPad, they could have put in 1GB of RAM, all sorts of higher spec things but you wouldn't necessarily buy another one and that's no good. When buyers aren't compelled to keep buying, that's when sellers stop growing and this is where the PC industry is now.
Their strategy of streamlining has worked great so far, Tim won't undo supply chain efficiency unless there's a good reason to. I think it's clear that they will be getting rid of the old style laptops and a few years down the line might even consolidate the MBP and Air lines - remember 10x faster than they are now. They want to be soldering RAM onto the machines so that trend will probably continue where it makes sense - when DDR4 comes probably next year with double the density, they can solder 8/16/32GB onto iMacs. The Mac Pro might not impact the supply chain enough to bother locking it down but there's no harm in them doing that either.
They won't touch DDR4 for the iMac by soldering it on-board. The designs for that would be a waste of design and manufacturing resources to modify the pre-existing motherboard design. it's far easier just to update the IC and interfaces for DDR4 following Intel's specs.
The modular idea doesn't work for a number of reasons (power/data connections) but the door is a big problem. They'd have to have variable sized doors. It also leaves uneven seams everywhere, Ive would not approve of that.
That's a mockup of the older model too. The newer one was this one:
No optical makes more sense. That was also before finding out about the Sandia CPU cooler. Right now Apple uses massive heatsinks, which are quite expensive. The following block of metal doesn't really do anything actively, fans just blow on it to keep it cold and the copper draws the heat away from the processor - that little white square is where the processor sits:
The Sandia cooler is a part Apple can machine out of aluminium very cheaply and is a fraction of the size:
That's what led to this:
I left some PCI slots on that one because it's not clear how they can make Thunderbolt work without an IGP, the slots are just half-length. They'd allow full-length outside but again, that would be the exception.
There's a difference with your disney example and also the car examples that crop up. What people often forget is that the chips inside the Mac Pro that do all the work (even the ones on the GPU cards) are actually incredibly small. They are not much larger than the chips you get in a laptop or iMac. The main difference is they get hotter so a significant portion of the Mac Pro design is for cooling. Using advanced cooling methods means those chips can fit into a smaller space.
You can see that today - the super slim iMac performs at the level of a 2009 Mac Pro, significantly outperforms its GPU - and yet the iMac is whisper quiet at full load. Intel and NVidia just improved performance per watt.
The bulk of the Mac Pro doesn't represent high performance but cooling inefficiency just like the mainframes that preceded it. The size of the storage represents the lack of areal density.
Some people see the Mac Pro as highlighting the best of computing but it really showcases the worst of it. The iMac and iPad represent the best of it because they hide the inefficiency to the point that you can't tell there is any. The iPad especially because it's passively cooled and everything is designed around real-time interaction.
Right now, some people are still in the phase of 'you can't do xyz with anything less than a Mac pro' and the scenarios get ever more elaborate but the fact is that people are doing the highest-end workflows without them on the workstation side and hundreds of server blades doing the processing:
When I saw those films (Flight and Looper), I didn't know the visual effects were designed with iMacs. While some people will continue to suggest that iMacs are not for professional/high-end/intensive/tight-deadline/color-accurate work, people are using them for it. A Mac Pro will do the same job but it wouldn't necessarily improve the workflow. When it comes to local real-time workflows, the difference is not that great.
That's designed for Haswell though. The Mac Pro will be Ivy Bridge. It would be nice if it was Haswell though.
500-600 W PSU is DOA. Unless you are classifying an 850W PS at 92 Platinum rating it's knee capping the entire system for expansion.
The redesign isn't a matter of linear scaling. Your removal of all that top level destroys any significant improvement of cooling rates and allowing for larger diameter convective heat transfer at lower rpms.
My entire digital arts, career for the past twenty three years has been possible because of my top of the line Mac Desktop. Started with the Mac2 then a Mac2FXm, Quadra, two Power Macs, G3,G4,G5, and four different Mac Pros. I have created music albums, film soundtracks, complex animations, websites and just about every single kind of digital art you can imagine on these macs. I usually get a new Mac desktop every two to three years. The rig I use now has 64GB ram and three cinema displays. All the professionals I know use these high horsepower machines. I can't wait to see what Apple has in store for the digital art professionals that need these powerful desktops.
[...] If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
Me? Honestly? "What am I missing? What the hell do these people want? What is missing from my product offering that's keeping them from buying? Features? Service? Price?" That's what I'D ask, but I don't have Tim's experience.
I understand your points. I just don't know how a guy like me makes Apple aware of the trigger issues for "stale buyers."
I haven't purchased an iPhone since the 4. By the time I was out of the contract, other players were offering bigger screens. I want that, I just don't want one from the other players. So I buy nothing. There's no way for Apple to know that I'll buy a new iPhone the second they make one with a 5" or bigger screen.
I haven't purchased a MacBook Pro since 2010. By the time I paid this one off they had stopped making a 17" screen. I like that form factor enough that, for me, it outweighs the many benefits of the newer models. There's no way for Apple to know there's a buyer for a 17" MacBook Pro standing ready to drop *ANOTHER* $4000.
I didn't even submit a capital request for a new tower at work this year even though we really need one because there wasn't enough advantage over what we have now to justify the cost. To Apple, that just looks like "the Pro isn't selling very well," as opposed to "we really need to improve this thing to make it attractive to buyers."
I grok what you're saying about supply and inventory et al. They make their best guesses at what people will want and hope for the best. That doesn't make the guy who hasn't bought anything for six years an invalid data point though. If he's like me, he's WILLING to buy, but the particular carrots Apple chose to dangle aren't the ones I want. Sometimes that'll be for good reason (Apple's figures show it won't sell) but sometimes Apple just picks characteristics that appeal to one group of buyers but not another. When something sells 20 million units, it's hard to know if adding that one extra feature would have driven it to 25 million, or if the extra 5 million would be worth the administrative and capital overhead.
Marvin, all I can say is that you must not make your living doing professional video, animation or even high end photography. 35%, 50%,90% gains are very substantial when you are doing projects that bill in the thousands and even tens of thousands of dollars. Feature films in the millions. Plenty to pay for the difference between a Mac Pro with the best GPU's that allow such speed gains. It is even important to people who bill $150 an hour. What if a factory manager was told he could get a 90% gain in productivity and profit for a minimal difference in investment? He would jump on it, and so do pros currently using workstations. Believe it or not many people load their machines with 64-128 GB of RAM. Check out Mac Performance Guide sometime. And he is only optimizing for photography. Photoshop files that run into the gigabytes with layers from large sensor still cameras. You want as much of that in RAM as possible. Huge time differences in processing. The differences in video and animation are just as stark, if not more so. Besides the fact that new hardware can be easily added with PCI as needed, where the cutting edge of graphics card performance currently is being developed, not years in the future, and after several generations of iMacs with soldered GPUs and RAM. A new Mac Pro would undoubtedly have Thunderbolt as well, but not as the only option for expansion.
Pros are making their decisions based on what comes from Apple by the end of this year. Hopefully Apple comes through. They don't have six years to wait for blue sky technology. I think most would be happy to move four years beyond 2010. That bleeding paper cut going to Windows is the pro market, and it is a lot more important than you seem to give it credit.
I think he has a point here. There is a tremendous opportunity here. The consumer market is saturated. Why not try to boost profits with high-end enterprise oriented hardware.
I just don't know how a guy like me makes Apple aware of the trigger issues for "stale buyers."
I didn't even submit a capital request for a new tower at work this year even though we really need one because there wasn't enough advantage over what we have now to justify the cost. To Apple, that just looks like "the Pro isn't selling very well," as opposed to "we really need to improve this thing to make it attractive to buyers."
There are a lot of factors that Apple has to weigh up. They have all the data for buying trends. A few people have pointed out that Apple would sell more if they just sold the right options but again, you have to think about the decisions from their point of view.
If you sold a tower for $2500 and you had an iMac at $2000 and a buyer comes up to you and says 'I'll buy from you if you make a tower for $1500'. Right from the start, you know the maximum sale price is below what you offer. The threat being made by the consumer is they won't buy at all unless that option exists. However you know that you will drop the average selling price and therefore the profits and to more than just that one customer. So what you'd have to weigh up is will the volume of customers buying the cheaper option increase enough to offset the lower profit. If not, your only option is to say to the customer, 'sorry, we don't swing that way'.
The same applies for the expensive models e.g 'I'll buy if you make a 17" or if you make a great value Mac Pro'. Apple has to decide to source materials, develop the manufacturing line, design marketing material, train staff to support it and so on. The 15" form factor is the most popular size:
The ratios won't be quite the same for Apple as they don't have the in-between sizes and the ratios will have changed since then because of netbooks failing but it's about 5:1 in favour of the 15" vs 17". Assuming that this isn't based solely on price, Apple appeals to a wider audience with a 15" so a 15" at the 17" price point has a wider appeal and the same or higher profit margin. They may well lose out on the sale to 17" buyers but they make the sale to 15" buyers and some of the previous 17" buyers move to 15". The decision as a seller is to get the best return.
On the Mac Pro side, the audience has pretty much flat-lined at 1 million units per quarter and this has been the case for years. It clearly hasn't followed the growth trends that the rest of the PC industry has seen as many buyers have migrated down and now the majority of the industry is on laptops (70% laptops, 28% desktops, 2% workstations). This small audience is willing to spend a lot initially, which generates a lot of profit. However, this audience doesn't spend frequently because work machines should be stable and you can't just be changing operating system versions every year. This inevitably slows down the buying. On top of that, the owners can upgrade almost everything inside it and they do (off-the-shelf CPUs, GPUs etc).
Apple knows this, they have the internet too and they can see people doing upgrades and they will have data from some partner suppliers, possibly NVidia. From what they see happening, they make decisions about what they should sell to get the best return. It doesn't really matter if someone is making a scientific discovery on a Mac or not if Apple doesn't know about it and the public doesn't know about it. All Apple knows is that person has paid the same as anyone else.
It's more likely that people who spend more on computers do so because they need it and that leads to the conclusion that this is an important group of buyers. Pixar, ILM etc are important buyers too and they have all HP and Dell. They are important in their own right though, their importance to Apple is how much they'd be willing to spend on Apple products. The celebrity factor doesn't work when it comes to workstations. Chris Rock owning a Mac is more influential than somebody making a scientific discovery on a tower hidden under a desk:
"Chris Rock told People in April that he was an avid Mac user. "If they had water, I would drink the Mac water,""
Some people who own Mac Pros think they are the most important people just like people who have the nicest cars or the biggest house or the CEO position at a company. It's not the labels or the products that make the people important but what they do. A Mac Pro or a 17" laptop won't be the most suitable machine for a very important and highly paid job. In no way does that make the job or the person doing it less important or less professional than people/jobs for whom a 17" MBP or MP are more suitable.
Apple's decisions are always presented as being 'suitable or unsuitable' for certain types of work. In every case these days, it's 'more suitable or less suitable'. An iMac is less suitable for animation work than a Mac Pro but an iMac with cloud rendering can be more suitable than a Mac Pro because it can end up cheaper as people have demonstrated and you don't have to source your own display. A Macbook Pro is less suitable for audio work than a Mac Pro for performance but more suitable if you need to take the machine with you. Buyers make compromises, sellers make compromises and sometimes sellers' compromises won't match the buyers' but they both make the decisions purposefully for the best return and you can't assume as a buyer that your compromises give Apple a better return, even beyond financial.
35%, 50%,90% gains are very substantial when you are doing projects that bill in the thousands and even tens of thousands of dollars. Feature films in the millions.
On the render farm side a 90% increase would be beneficial but individuals can't do what you are describing. The videos I posted earlier are people who do this. The visual effects guy was sitting with Robert Zemeckis at Skywalker Ranch and he wanted a change done to a 900 frame sequence for the movie Flight ASAP. A single workstation can't do that regardless if it is a top of the line Mac Pro because even a Mac Pro would take at least 30 minutes per frame so about 19 days. They used cloud rendering and scaled up to 100 nodes to get it done in time using a laptop (they use Macbook Pros). They did the changes on the 'workstation' (MBP/iMac) in Maya and Nuke and shoved the processing to the render nodes. The other freelancer guy needed to do a 100 machine hour render in 2 days; Mac Pro or not, you can't do that with one machine by definition because you don't have enough hours in the day.
Some computer has to sit in the cloud for these jobs but it's not towers any more, it's server blades, which Apple doesn't make. Someone could of course build their own Mac Pro farm but it's too expensive because it's >$6000 per 12-core unit. It's actually more cost-effective just now to use Mac Minis. The visual effects guy mentioned that it would cost them around $2m to scale up to the capacity they'd need. Even the huge animation companies have talked about outsourcing their render farms because it's not cost-effective to maintain them.
There's no question the Mac Pro is faster and that's better, it's just not the difference between getting a job done or not and I don't understand the opposition to doing these things on say a Macbook Pro. Isn't it good that we now have inexpensive machines that sit in our laps that can churn through high-end jobs? I think it's great you can now buy a laptop and get a similar experience to a Mac Pro.
What if a factory manager was told he could get a 90% gain in productivity
90% performance gain doesn't mean a 90% gain in productivity unless the machine is only doing raw processing, which means it's not being used as a workstation but a render node. That's not to say the productivity gains you do get aren't valuable but the gains get less the more that machines improve.
Pros are making their decisions based on what comes from Apple by the end of this year. Hopefully Apple comes through.
People said that last year though and the year before. We know that no one can force them to do what they don't want to do. They have to be convinced that it's worthwhile. They need market projections, design requirements etc. Just saying 'I know a few important people who will buy one' can't possibly convince a company to roll out a new production line.
They've made the decision to make a new Pro so there's not much uncertainty over them making one but we'll know better what their future plans are when we see what they've done with it. The 17" MBP decision will be clearer once the prices shift on the Retina models. It's to be expected now though that these machines aren't growth models and even yearly upgrade cycles aren't really essential. Intel takes some responsibility for this - Apple can only offer what they sell.
Just this month Red Shark released its Lightworks movie editing program for Linux. It is a program that seems to be on par with anything from Avid or Adobe. It's even set up to create 3D movies, AND IT'S FREE!
You remind me of all those Photoshop killers (GIMP and Co.). A free program will just not have the final polish that professionals need. It will have some nice touches but will lack something that is badly needed. Just look at those free MS Office competitors. What percentage of people uses them?
One more things:) A Free program will never cut it as there is always chance that it will be abandoned by the developers. What if no update comes in years?
Pros are making their decisions based on what comes from Apple by the end of this year. Hopefully Apple comes through.
Quote:
Originally Posted by Marvin
People said that last year though and the year before.
...and they did. Our primary Avid suite, once the foundation for Apple technology in our plant, is now running on HP machines. Engineering refused to pay premium coin for woefully outdated hardware so they went with Windows instead. On the audio side, the latest version of Pro Tools offers serious workflow improvements but won't run on our 2009 Mac. In the absence of a "new" Mac Pro the fear is that we'll wind up with Windows machines, too.
Inertia is hard to overcome in a production setting. It took Apple completely ignoring the Pro tower segment to uproot it. I imagine it will be similarly difficult to persuade the decision makers to even look at going back to Apple now that we've already gone through a migration. Why would they? There's a lot of effort involved in switching everything over, and I can't imagine there being much will to do it all again.
What do you bet that Chris Rock was first impressed by a Mac on a film or video shoot? And why was it there? Ultimately, because almost everything is edited on Avids using Mac Pros. This is where Hollywood saw their first Macs. The editors liked them, then the directors liked them, then, when PowerBooks came out, they showed up on the set being used for script, continuity, wardrobe, production management, and eventually the stars thought they were cool, so they bought them, and voila, everybody who could see the stars thought they were cool, which is EVERYBODY, started buying them. And it all started with the ultra cool Mac Pro in the edit room (or it's earlier equivalents). Then they started to appear in the feature films themselves. Why? Because everyone on set was using them, and they were near at hand. Apple eventually got the idea, and subsidized product placement was born for computers. Even now, the ratio of Macs on screen is much higher than in the general computer population. And it all started with a high end Mac in the edit suite. Where they still are in Hollywood. There is a huge marketing advantage, a halo if you will, that having a machine like the Mac Pro confers that you just don't get with an iMac. They have a presence. Secretaries and middle management don't use them. Not to mention that it is because form follows function. The editors and engineers know that. It is an intangible of coolness that a bunch of small objects spread out on a desk surrounding an iMac or MacBook Pro just ain't got. Leaving aside the fact that they do the job better. The Corvette effect. Marketing commercials can't buy and hardware specs don't account for.
Comments
I dont think their unit sales numbers nor the drop in technology costs YoY warrant that as a viable option.
The current direction of Apple is something that I have been wondering about for a while now, specifically the lack of focus on the high-end professional consumers served by the Mac Pro. The iOS platform has been incredibly lucerative and successful for Apple, however there is an obvious question about the future of the Mac hardware platform.
In the last five years, there's been a shift which has seen huge adoption of Mac laptops and desktops by a wide variety of people. It would seem this has just as much to do with the world's detachment from Microsoft as it does the adoption of Mac hardware. Now that the majority of a lot of people's "computing" is via web and mobile apps, the architecture of the interface matters a lot less. We've gone from a world where Windows was the dominant paradigm because it ran all the software, to a world where Windows is the exceptional paradigm. One goes with Windows if they have software that doesn't run as a web app. Otherwise, a Mac will do fine. The triumph of Apple engineering and the computer as a consumer device.
The concern is three-fold. The first is that iOS occupies a less defensible position in the market. As devices have shorter lifespans and apps are often replicated across platforms, the iOS is less sticky than the MacOS. Consumers seem to go from iPhones to Androids, now they're trying Blackberries again. As long as the device 1) makes phone calls, 2) receives/sends email, 3) receives/sends texts, 4) has a decent camera, and 5) has Facebook/Instagram/Twitter/Spotify/etc, I don't think people are really bothered about the device as much as might be considered. Prosumers, yes. The people who read this website, probably. But a lot of people really don't seem too bothered. And Apple didn't help with the cable shift of the last iPhone. All of a sudden, iPhones went from a gated community, to a fragmented community. When you have to carry a cable for an iPhone 5 like you do with a Galaxy, they all just become phones again.
The second concern is the proliferation of hardware. MacBook Pros with retina displays and without retina displays. iPads with retina displays and iPads without retina displays. This is a significant break from the Jobsian narrow ecosystem that made Apple so successful. There used to be one iPad. One MacBook. One product at each level. That is a luxury manufacturer logic. You do not get a choice, for part of what you are paying the premium for is curation of the device. Now with Tim, we see new product segments, and point-to-point competition. Business strategy that is technically correct but again breaks the ecosystem structure. This may seem minor, but along with the dividend payment, Apple is being run as a business rather than as a product religion. Last time this happened, it wasn't long before the company ended up replicating the dominant industry logic.
That comes to the third point, which is the very broken pro user market. Final Cut X was the first debacle. Apple took a long time to capture that attention, and actually get AVID on the run, and Adobe nearly out of the game. They were almost home-free on the pro video editing market. Almost... and then Final Cut X and utter disaster for the high-end market. Now, there's been a big shift back to AVID, and another shift to Adobe. From a business point of view, these are niche segments that don't pay for a lot of hardware in the end. From a spiritual point of view, those users are at the heart of the platform. The storytellers who have the need for the best hardware. Who's time is so valuable, they upgrade machines whenever there's a speed-bump that can save them an appreciable amount of time. Now the lack of focus on that segment from an Apple that is so cash rich can only mean that the segment is not considered important to the company. Huge mistake, and probably the biggest red flag here. Even if the pro video market was a complete loss leader for Apple, the pro video market is the defence industry of the computer industry... it's where all the black budgets and high-end stuff is created, tested, and purchased.
I think Tim is doing a fantastic job overall with Apple. However, the three points above are three very obvious non-Apple things that his Apple is doing, and should cause any investor cause for concern. Apple will always be around, and will always be producing hardware, but whether or not it can maintain its mystique and related margins has to do with the company's continued ability to capture the imaginations of a relatively small number of VERY influential content creators. If I were one of those creators today, I would be asking if Apple really cares. And that is a big big red flag.
Originally Posted by Johnny Mozzarella
Every year Apple should release a new Mac Pro, a single, top of the line model.
Apple should then continue to sell the previous years models as a lower priced option.
The chips don't come out that often, and they're already doing that.
Originally Posted by jnr1212
One MacBook.
There was never just one MacBook.
Now with Tim…
You mean with Steve. Everything out now was approved by him.
This may seem minor, but along with the dividend payment, Apple is being run as a business rather than as a product religion.
I rather think you can't say 'rather' yet. Steve did that, too.
Final Cut X was the first debacle.
And there goes your opinion.
…that segment from an Apple that is so cash rich…
Thing is, that segment isn't cash rich.
Even if the pro video market was a complete loss leader for Apple, the pro video market is the defence industry of the computer industry... it's where all the black budgets and high-end stuff is created, tested, and purchased.
Apple doesn't run "loss leaders". Apple doesn't run losses. Sounds like you want them to run more like a business than a product-oriented group…
Thanks, Marvin, for an educational post, as always.
Quote:
Originally Posted by Marvin
You save the space that the Mac Pro would have taken up to begin with. There's more cabling with external storage + 2-3 peripherals but you can put them in a box if you like.
The size of the card isn't important. The fastest cards are not even 2x the performance of the iMac's GPU but cost $1000 or more each. Not to mention they are outside the safe power limits of the Mac Pro too.
The Radeon 7990 is 375W. I'm not sure how many professionals would put a 375W GPU into slots with a 300W power allocation but I doubt they'd be in business long.
The Mac Pro can obviously hold higher wattage cards than the iMac but they don't perform significantly better in practise. OS X doesn't support the highest-end GPUs on the software side either.
With the high-end GPUs out of the equation, you're just left with expansion cards, which work fine over Thunderbolt.
Marvin, my 2010 Mac Pro sits under my work area and takes up NO desk space. Furthermore, the space taken is vertical. Even if it were on the desk (few are) it would take up the space of an AJA ioXT HD interface and cables and perhaps a Thunderbolt hub minus the cables. The Mac Pro can even be placed in the next room, which it often is in professional settings.
As for the Radeon 7990's power requirements, the first PCI 3 GPU, we are talking about a new Mac Pro here. Is there any reason a larger power supply could no be fitted into a new Mac Pro design? There are many cards of lower price than the 7990 that are also quite large with great performance, particularly with CUDA. PCI 3 will provide great advantages for CUDA and OpenCL, card based technologies that greatly speed up video rendering and ray tracing, a staple of animation. Adobe is making amazing use of CUDA for video rendering. As for Apple software support, I imagine they have the engineering talent and budget (it would not require much, specs are published) to implement version 4 OpenGL. I can't imagine they would not further development and implementation of OpenCL, the standard they have been pushing.
And to working fine over Thunderbolt, as of now. the expansion card boxes are much more limited in wattage and card space than even the 2010 Mac Pro. How do you put a Radeon 7990 in them? And you will eventually hit the speed wall of even Thunderbolt 2 (not yet even available) compared to internal PCI 3 which is shipping. Plus another (big) box on your desk. And more cables. And more fan noise. And whoops, I accidentally pulled the plug. Not a pro environment in my opinion. iMacs for those who are satisfied, they have their place for many pro uses. But for most, they will not replace the Mac Pro workstation concept, in functionality, speed and elegance, where they are truly needed and appreciated.
That's the other thing about Apple's direction with the computer line. It's often thought that Apple should offer what the potential buyer wants (e.g the customer is always right) but before the iMac, nobody asked for one. Some people will then mutter 'exactly, nobody asked for one' but look at the sales stats less than 2 years after it was introduced:
http://www.apple.com/pr/library/2000/01/19Apple-Reports-First-Quarter-Profit-of-183-Million.html
It made up over 50% of their sales. Tim Cook's role in overhauling Apple's supply chain had a very important implication for their computer designs too. One thing he said was:
"You kind of want to manage [Apple's inventory] like you're in the dairy business. If it gets past its freshness date, you have a problem."
When you offer customers a lot of choice then you can't always guarantee that you can turn over the inventory quickly. You might offer 10 GPU options and buy a bunch of stock and then 90% of buyers only stick with the entry GPU. This is exactly why the PC component industry struggles so much and why Intel/AMD/NVidia all want to move towards SoC. It's also why everyone seems to want a tablet now and their own operating system. It's a far more profitable business model to sell you an entire finished product than a part of one.
The ideal scenario (from the seller's perspective) where every manufacturer offers one device per category is not going to happen but out of all manufacturers, Apple will move the closest to this ideal because this is how they choose to do things.
They narrowed GPU options on the Mac Pro to one manufacturer and two models. Intel makes dozens of CPUs but you get to choose from 2-3 per main model. They removed the hardware RAID card BTO option too.
Apple's direction is not to expand choice unless it's clear there will be a huge benefit but to streamline their manufacturing and reduce inventory. That setup is the most profitable.
With this in mind, people then point out that it shouldn't be all about counting the beans and that's not what it used to be about and if they don't get back there, the company with $147b in the bank will go bankrupt eventually. Well, that's what nearly drove them to bankruptcy before Tim came along and turned it around by streamlining their inventory.
It's much easier to tell what they will do going forward when you look at it from their point of view rather than from the perspective as a buyer. As a potential buyer, it's easy to rattle off fictional hardware that you'd buy - 'sure if they made a 19" laptop with dual GTX 680Ms, a 1TB SSD (heck make it dual 1TB in RAID 0) and a giant 300W power brick and it cost $6000, I'd buy it or a Mac Pro with 9 PCI slots and PCI SSD and quad CPUs with 256GB RAM'. Except you probably wouldn't buy it, you'd buy the lowest possible model with your minimum requirements (e.g it's a tower or it has a big screen) and try and upgrade it cheaper without paying Apple more money that you have to.
With the Mac Pro, people talk about money being no object and almost every time someone has a Mac Pro, it's an ancient one they are for some reason dying to upgrade now that Apple doesn't sell one or doesn't sell one they think is good value. If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
The people who work at Apple have been at this a very long time and they've seen technology and companies grow and die over 30+ years. That's a lot of experience. They don't always exercise the best judgement in everything they do but they are talented people and more often than not, they have driven technology in the right direction and that gives them a very long term vision that buyers don't need to have. Buyers don't think about what will happen in 10 years, buyers think about what the current costs and needs are.
Apple knows that in a few years, computers are going to be 10x faster than they are now no matter if it's distributed more to the GPU than the CPU and they are in the business of selling computers. So, they need to figure how to make it compelling enough to keep buying new computers while almost satisfying the current needs of the buyer. When they launched the iPad, they could have put in 1GB of RAM, all sorts of higher spec things but you wouldn't necessarily buy another one and that's no good. When buyers aren't compelled to keep buying, that's when sellers stop growing and this is where the PC industry is now.
Their strategy of streamlining has worked great so far, Tim won't undo supply chain efficiency unless there's a good reason to. I think it's clear that they will be getting rid of the old style laptops and a few years down the line might even consolidate the MBP and Air lines - remember 10x faster than they are now. They want to be soldering RAM onto the machines so that trend will probably continue where it makes sense - when DDR4 comes probably next year with double the density, they can solder 8/16/32GB onto iMacs. The Mac Pro might not impact the supply chain enough to bother locking it down but there's no harm in them doing that either.
Well, they are streamlining themselves right out of the professional marketplace. There is a huge movement to Adobe software in video editing. That is occurring now, not ten years from now. And I speak as the former exclusive Avid dealer for Hawaii. Final Cut X was a watershed and Adobe is taking advantage of it. Their products are cross platform and actually run better on PC's because of software optimaization and access to the best current hardware, particularly GPUs. This is also true in animation, particularly ray tracing., There are many of us, myself included, who would never consider moving to Windows, but large numbers are because of Apple's intransigence regarding Mac Pro workstation development. The high end hardware and software is available on Windows workstations like the HP's, and they are either moving or seriously considering it. We can theorize about the future of computing until we are blue in the face, but that does not change the fact that a new Mac Pro, not that much different than the old one in form but with up to date technology is needed now to stop the bleeding of professionals to Windows where they can configure to the current state of the art today and can count on future development of such systems. Why do you think Mac Pro diehards are so torn and emotional? We want to stay with the Mac, but can Apple give us what we need now and in the future?
And I have been collecting a few older Power PC G5 dual units. They run fine, even the G4 dual tower. But the older OS systems will only run older browsers. So when the new units come out, the intel machines should finally come down in price.
Currently you can get a decent G5 dual for about $150..
With optical Thunderbolt cables, you can have peripherals across the street. A slight tripping hazard but only to non-professionals who don't watch where they're going.
Why is it so important that they are large? The speed doesn't scale directly with the size of the cards. Apple could put a larger power supply in and support higher powered cards and run the fans a bit faster but they made a set of compromises when they designed it as they do with every machine. They want it to be quiet in operation so it can't get too hot. That's obviously why they went a bit nuts with the heatsinks. They have to consider what the potential gains would be by going that bit further.
When you look at one of the highest end GPUs - the latest Titan GPU - you can see that the OpenCL score for that in one benchmark is 1421:
http://www.tomshardware.com/reviews/geforce-gtx-titan-opencl-cuda-workstation,3474-19.html
This GPU costs $1000 and the power draw is around 250W. The iMac with the 680MX scores 814 here:
http://barefeats.com/imac12g4.html
So the Titan is 75% faster for computation. There are a few other tests that have similar results, the Passmark one shows 90% faster:
http://www.videocardbenchmark.net/high_end_gpus.html
For all of the extra effort, heat, power draw, size and cost, that's all you get. This year's iMac GPU will be 30% faster again so if you stocked up a Mac Pro when the Titan came out for say $3500 + display, you only end up 35-45% faster than this year's iMac and they have a sleek desktop whereas you have a power hungry tower. You're not going to buy another $1000 GPU after just 1 year.
If they did allow you to run 3 or more high-end GPUs, there's some advantage but again, they cost $1000 each and Apple wouldn't do this because so few people would take advantage of it. When you need lots of compute power, you can network multiple machines together.
The same way with a Mac Pro - you don't. If you had a large enough power supply you could do either one though.
Workstation chipsets don't have PCI 3 support - that doesn't come until Haswell chipsets. Some third party manufacturers have added it but just because a spec exists, doesn't mean that suddenly becomes the minimum requirement. These comparisons tend to go the route of comparing some consumer option to the highest possible option there is and the purpose is obviously to suggest the consumer option is therefore not up to the job. The false assumption being made is that there is a common task somewhere that's using the full extent of whatever spec exists e.g if it's possible to put 128GB of RAM in something, that must mean someone is using 128GB of RAM therefore the iMac is useless because it only support 32GB. That idea doesn't hold up in practise.
The faster machines become, the less reliant people will be on specialised PCI cards for processing - it'll be done natively on the CPU/GPU. Apple was demoing native 4K on an iMac at NAB this year. IO standards will move to optical soon. The workstation concept won't disappear overnight but these advances will keep eroding the userbase and eventually the sellers will stop selling them.
Think 6 years down the line, say that SSDs go down in price 25% every year. Crucial has 1TB at $600 just now. That means that a 1TB drive will be $107, essentially today's HDD prices. CPU/GPU power will be about 8x faster with one likely weighted more than the other. Memory density will mean 32GB RAM is inexpensive. IO ports will exceed 50Gbps each.
You won't need Red Rocket cards or other processing cards. The internal GPUs in laptops will match current GTX Titans.
At that point, what will you need to be putting inside a tower?
Windows PC sales are down vs last year. Apple suffered less than PC manufacturers. It's obvious people are moving to Premiere but not to Windows PCs in large numbers. I imagine some have moved to HP/Dell workstations but the phrase 'bleeding professionals to Windows' again suggests that professionals only use Mac Pros, which isn't the case. There are about 90 million PCs sold worldwide every quarter. Workstations are 1 million of those. If there's any bleeding, it's little more than from a paper cut.
Quote:
Originally Posted by Marvin
That's the other thing about Apple's direction with the computer line. It's often thought that Apple should offer what the potential buyer wants (e.g the customer is always right) but before the iMac, nobody asked for one. Some people will then mutter 'exactly, nobody asked for one' but look at the sales stats less than 2 years after it was introduced:
http://www.apple.com/pr/library/2000/01/19Apple-Reports-First-Quarter-Profit-of-183-Million.html
It made up over 50% of their sales. Tim Cook's role in overhauling Apple's supply chain had a very important implication for their computer designs too. One thing he said was:
"You kind of want to manage [Apple's inventory] like you're in the dairy business. If it gets past its freshness date, you have a problem."
When you offer customers a lot of choice then you can't always guarantee that you can turn over the inventory quickly. You might offer 10 GPU options and buy a bunch of stock and then 90% of buyers only stick with the entry GPU. This is exactly why the PC component industry struggles so much and why Intel/AMD/NVidia all want to move towards SoC. It's also why everyone seems to want a tablet now and their own operating system. It's a far more profitable business model to sell you an entire finished product than a part of one.
The ideal scenario (from the seller's perspective) where every manufacturer offers one device per category is not going to happen but out of all manufacturers, Apple will move the closest to this ideal because this is how they choose to do things.
They narrowed GPU options on the Mac Pro to one manufacturer and two models. Intel makes dozens of CPUs but you get to choose from 2-3 per main model. They removed the hardware RAID card BTO option too.
Apple's direction is not to expand choice unless it's clear there will be a huge benefit but to streamline their manufacturing and reduce inventory. That setup is the most profitable.
With this in mind, people then point out that it shouldn't be all about counting the beans and that's not what it used to be about and if they don't get back there, the company with $147b in the bank will go bankrupt eventually. Well, that's what nearly drove them to bankruptcy before Tim came along and turned it around by streamlining their inventory.
It's much easier to tell what they will do going forward when you look at it from their point of view rather than from the perspective as a buyer. As a potential buyer, it's easy to rattle off fictional hardware that you'd buy - 'sure if they made a 19" laptop with dual GTX 680Ms, a 1TB SSD (heck make it dual 1TB in RAID 0) and a giant 300W power brick and it cost $6000, I'd buy it or a Mac Pro with 9 PCI slots and PCI SSD and quad CPUs with 256GB RAM'. Except you probably wouldn't buy it, you'd buy the lowest possible model with your minimum requirements (e.g it's a tower or it has a big screen) and try and upgrade it cheaper without paying Apple more money that you have to.
With the Mac Pro, people talk about money being no object and almost every time someone has a Mac Pro, it's an ancient one they are for some reason dying to upgrade now that Apple doesn't sell one or doesn't sell one they think is good value. If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
The people who work at Apple have been at this a very long time and they've seen technology and companies grow and die over 30+ years. That's a lot of experience. They don't always exercise the best judgement in everything they do but they are talented people and more often than not, they have driven technology in the right direction and that gives them a very long term vision that buyers don't need to have. Buyers don't think about what will happen in 10 years, buyers think about what the current costs and needs are.
Apple knows that in a few years, computers are going to be 10x faster than they are now no matter if it's distributed more to the GPU than the CPU and they are in the business of selling computers. So, they need to figure how to make it compelling enough to keep buying new computers while almost satisfying the current needs of the buyer. When they launched the iPad, they could have put in 1GB of RAM, all sorts of higher spec things but you wouldn't necessarily buy another one and that's no good. When buyers aren't compelled to keep buying, that's when sellers stop growing and this is where the PC industry is now.
Their strategy of streamlining has worked great so far, Tim won't undo supply chain efficiency unless there's a good reason to. I think it's clear that they will be getting rid of the old style laptops and a few years down the line might even consolidate the MBP and Air lines - remember 10x faster than they are now. They want to be soldering RAM onto the machines so that trend will probably continue where it makes sense - when DDR4 comes probably next year with double the density, they can solder 8/16/32GB onto iMacs. The Mac Pro might not impact the supply chain enough to bother locking it down but there's no harm in them doing that either.
They won't touch DDR4 for the iMac by soldering it on-board. The designs for that would be a waste of design and manufacturing resources to modify the pre-existing motherboard design. it's far easier just to update the IC and interfaces for DDR4 following Intel's specs.
Quote:
Originally Posted by Marvin
The modular idea doesn't work for a number of reasons (power/data connections) but the door is a big problem. They'd have to have variable sized doors. It also leaves uneven seams everywhere, Ive would not approve of that.
That's a mockup of the older model too. The newer one was this one:
No optical makes more sense. That was also before finding out about the Sandia CPU cooler. Right now Apple uses massive heatsinks, which are quite expensive. The following block of metal doesn't really do anything actively, fans just blow on it to keep it cold and the copper draws the heat away from the processor - that little white square is where the processor sits:
The Sandia cooler is a part Apple can machine out of aluminium very cheaply and is a fraction of the size:
That's what led to this:
I left some PCI slots on that one because it's not clear how they can make Thunderbolt work without an IGP, the slots are just half-length. They'd allow full-length outside but again, that would be the exception.
There's a difference with your disney example and also the car examples that crop up. What people often forget is that the chips inside the Mac Pro that do all the work (even the ones on the GPU cards) are actually incredibly small. They are not much larger than the chips you get in a laptop or iMac. The main difference is they get hotter so a significant portion of the Mac Pro design is for cooling. Using advanced cooling methods means those chips can fit into a smaller space.
You can see that today - the super slim iMac performs at the level of a 2009 Mac Pro, significantly outperforms its GPU - and yet the iMac is whisper quiet at full load. Intel and NVidia just improved performance per watt.
The bulk of the Mac Pro doesn't represent high performance but cooling inefficiency just like the mainframes that preceded it. The size of the storage represents the lack of areal density.
Some people see the Mac Pro as highlighting the best of computing but it really showcases the worst of it. The iMac and iPad represent the best of it because they hide the inefficiency to the point that you can't tell there is any. The iPad especially because it's passively cooled and everything is designed around real-time interaction.
Right now, some people are still in the phase of 'you can't do xyz with anything less than a Mac pro' and the scenarios get ever more elaborate but the fact is that people are doing the highest-end workflows without them on the workstation side and hundreds of server blades doing the processing:
When I saw those films (Flight and Looper), I didn't know the visual effects were designed with iMacs. While some people will continue to suggest that iMacs are not for professional/high-end/intensive/tight-deadline/color-accurate work, people are using them for it. A Mac Pro will do the same job but it wouldn't necessarily improve the workflow. When it comes to local real-time workflows, the difference is not that great.
That's designed for Haswell though. The Mac Pro will be Ivy Bridge. It would be nice if it was Haswell though.
500-600 W PSU is DOA. Unless you are classifying an 850W PS at 92 Platinum rating it's knee capping the entire system for expansion.
The redesign isn't a matter of linear scaling. Your removal of all that top level destroys any significant improvement of cooling rates and allowing for larger diameter convective heat transfer at lower rpms.
They won't do that.
I can't wait to see what Apple has in store for the digital art professionals that need these powerful desktops.
Quote:
Originally Posted by Marvin
[...] If you were on the other side listening to that and seeing someone not buying from you for as many as 6 years or more, what decision would you make?
Me? Honestly? "What am I missing? What the hell do these people want? What is missing from my product offering that's keeping them from buying? Features? Service? Price?" That's what I'D ask, but I don't have Tim's experience.
I understand your points. I just don't know how a guy like me makes Apple aware of the trigger issues for "stale buyers."
I haven't purchased an iPhone since the 4. By the time I was out of the contract, other players were offering bigger screens. I want that, I just don't want one from the other players. So I buy nothing. There's no way for Apple to know that I'll buy a new iPhone the second they make one with a 5" or bigger screen.
I haven't purchased a MacBook Pro since 2010. By the time I paid this one off they had stopped making a 17" screen. I like that form factor enough that, for me, it outweighs the many benefits of the newer models. There's no way for Apple to know there's a buyer for a 17" MacBook Pro standing ready to drop *ANOTHER* $4000.
I didn't even submit a capital request for a new tower at work this year even though we really need one because there wasn't enough advantage over what we have now to justify the cost. To Apple, that just looks like "the Pro isn't selling very well," as opposed to "we really need to improve this thing to make it attractive to buyers."
I grok what you're saying about supply and inventory et al. They make their best guesses at what people will want and hope for the best. That doesn't make the guy who hasn't bought anything for six years an invalid data point though. If he's like me, he's WILLING to buy, but the particular carrots Apple chose to dangle aren't the ones I want. Sometimes that'll be for good reason (Apple's figures show it won't sell) but sometimes Apple just picks characteristics that appeal to one group of buyers but not another. When something sells 20 million units, it's hard to know if adding that one extra feature would have driven it to 25 million, or if the extra 5 million would be worth the administrative and capital overhead.
Marvin, all I can say is that you must not make your living doing professional video, animation or even high end photography. 35%, 50%,90% gains are very substantial when you are doing projects that bill in the thousands and even tens of thousands of dollars. Feature films in the millions. Plenty to pay for the difference between a Mac Pro with the best GPU's that allow such speed gains. It is even important to people who bill $150 an hour. What if a factory manager was told he could get a 90% gain in productivity and profit for a minimal difference in investment? He would jump on it, and so do pros currently using workstations. Believe it or not many people load their machines with 64-128 GB of RAM. Check out Mac Performance Guide sometime. And he is only optimizing for photography. Photoshop files that run into the gigabytes with layers from large sensor still cameras. You want as much of that in RAM as possible. Huge time differences in processing. The differences in video and animation are just as stark, if not more so. Besides the fact that new hardware can be easily added with PCI as needed, where the cutting edge of graphics card performance currently is being developed, not years in the future, and after several generations of iMacs with soldered GPUs and RAM. A new Mac Pro would undoubtedly have Thunderbolt as well, but not as the only option for expansion.
Pros are making their decisions based on what comes from Apple by the end of this year. Hopefully Apple comes through. They don't have six years to wait for blue sky technology. I think most would be happy to move four years beyond 2010. That bleeding paper cut going to Windows is the pro market, and it is a lot more important than you seem to give it credit.
I think he has a point here. There is a tremendous opportunity here. The consumer market is saturated. Why not try to boost profits with high-end enterprise oriented hardware.
Originally Posted by AdonisSMU
The consumer market is saturated.
At 15% marketshare?
There are a lot of factors that Apple has to weigh up. They have all the data for buying trends. A few people have pointed out that Apple would sell more if they just sold the right options but again, you have to think about the decisions from their point of view.
If you sold a tower for $2500 and you had an iMac at $2000 and a buyer comes up to you and says 'I'll buy from you if you make a tower for $1500'. Right from the start, you know the maximum sale price is below what you offer. The threat being made by the consumer is they won't buy at all unless that option exists. However you know that you will drop the average selling price and therefore the profits and to more than just that one customer. So what you'd have to weigh up is will the volume of customers buying the cheaper option increase enough to offset the lower profit. If not, your only option is to say to the customer, 'sorry, we don't swing that way'.
The same applies for the expensive models e.g 'I'll buy if you make a 17" or if you make a great value Mac Pro'. Apple has to decide to source materials, develop the manufacturing line, design marketing material, train staff to support it and so on. The 15" form factor is the most popular size:
http://blog.laptopmag.com/data-shows-clueless-consumers-buying-15-inch-notebooks-ruining-market
The ratios won't be quite the same for Apple as they don't have the in-between sizes and the ratios will have changed since then because of netbooks failing but it's about 5:1 in favour of the 15" vs 17". Assuming that this isn't based solely on price, Apple appeals to a wider audience with a 15" so a 15" at the 17" price point has a wider appeal and the same or higher profit margin. They may well lose out on the sale to 17" buyers but they make the sale to 15" buyers and some of the previous 17" buyers move to 15". The decision as a seller is to get the best return.
On the Mac Pro side, the audience has pretty much flat-lined at 1 million units per quarter and this has been the case for years. It clearly hasn't followed the growth trends that the rest of the PC industry has seen as many buyers have migrated down and now the majority of the industry is on laptops (70% laptops, 28% desktops, 2% workstations). This small audience is willing to spend a lot initially, which generates a lot of profit. However, this audience doesn't spend frequently because work machines should be stable and you can't just be changing operating system versions every year. This inevitably slows down the buying. On top of that, the owners can upgrade almost everything inside it and they do (off-the-shelf CPUs, GPUs etc).
Apple knows this, they have the internet too and they can see people doing upgrades and they will have data from some partner suppliers, possibly NVidia. From what they see happening, they make decisions about what they should sell to get the best return. It doesn't really matter if someone is making a scientific discovery on a Mac or not if Apple doesn't know about it and the public doesn't know about it. All Apple knows is that person has paid the same as anyone else.
It's more likely that people who spend more on computers do so because they need it and that leads to the conclusion that this is an important group of buyers. Pixar, ILM etc are important buyers too and they have all HP and Dell. They are important in their own right though, their importance to Apple is how much they'd be willing to spend on Apple products. The celebrity factor doesn't work when it comes to workstations. Chris Rock owning a Mac is more influential than somebody making a scientific discovery on a tower hidden under a desk:
http://www.forbes.com/2010/07/01/apple-microsoft-seinfeld-technology-celebrities.html
"Chris Rock told People in April that he was an avid Mac user. "If they had water, I would drink the Mac water,""
Some people who own Mac Pros think they are the most important people just like people who have the nicest cars or the biggest house or the CEO position at a company. It's not the labels or the products that make the people important but what they do. A Mac Pro or a 17" laptop won't be the most suitable machine for a very important and highly paid job. In no way does that make the job or the person doing it less important or less professional than people/jobs for whom a 17" MBP or MP are more suitable.
Apple's decisions are always presented as being 'suitable or unsuitable' for certain types of work. In every case these days, it's 'more suitable or less suitable'. An iMac is less suitable for animation work than a Mac Pro but an iMac with cloud rendering can be more suitable than a Mac Pro because it can end up cheaper as people have demonstrated and you don't have to source your own display. A Macbook Pro is less suitable for audio work than a Mac Pro for performance but more suitable if you need to take the machine with you. Buyers make compromises, sellers make compromises and sometimes sellers' compromises won't match the buyers' but they both make the decisions purposefully for the best return and you can't assume as a buyer that your compromises give Apple a better return, even beyond financial.
On the render farm side a 90% increase would be beneficial but individuals can't do what you are describing. The videos I posted earlier are people who do this. The visual effects guy was sitting with Robert Zemeckis at Skywalker Ranch and he wanted a change done to a 900 frame sequence for the movie Flight ASAP. A single workstation can't do that regardless if it is a top of the line Mac Pro because even a Mac Pro would take at least 30 minutes per frame so about 19 days. They used cloud rendering and scaled up to 100 nodes to get it done in time using a laptop (they use Macbook Pros). They did the changes on the 'workstation' (MBP/iMac) in Maya and Nuke and shoved the processing to the render nodes. The other freelancer guy needed to do a 100 machine hour render in 2 days; Mac Pro or not, you can't do that with one machine by definition because you don't have enough hours in the day.
Some computer has to sit in the cloud for these jobs but it's not towers any more, it's server blades, which Apple doesn't make. Someone could of course build their own Mac Pro farm but it's too expensive because it's >$6000 per 12-core unit. It's actually more cost-effective just now to use Mac Minis. The visual effects guy mentioned that it would cost them around $2m to scale up to the capacity they'd need. Even the huge animation companies have talked about outsourcing their render farms because it's not cost-effective to maintain them.
There's no question the Mac Pro is faster and that's better, it's just not the difference between getting a job done or not and I don't understand the opposition to doing these things on say a Macbook Pro. Isn't it good that we now have inexpensive machines that sit in our laps that can churn through high-end jobs? I think it's great you can now buy a laptop and get a similar experience to a Mac Pro.
90% performance gain doesn't mean a 90% gain in productivity unless the machine is only doing raw processing, which means it's not being used as a workstation but a render node. That's not to say the productivity gains you do get aren't valuable but the gains get less the more that machines improve.
People said that last year though and the year before. We know that no one can force them to do what they don't want to do. They have to be convinced that it's worthwhile. They need market projections, design requirements etc. Just saying 'I know a few important people who will buy one' can't possibly convince a company to roll out a new production line.
They've made the decision to make a new Pro so there's not much uncertainty over them making one but we'll know better what their future plans are when we see what they've done with it. The 17" MBP decision will be clearer once the prices shift on the Retina models. It's to be expected now though that these machines aren't growth models and even yearly upgrade cycles aren't really essential. Intel takes some responsibility for this - Apple can only offer what they sell.
Quote:
Originally Posted by Smallwheels
Just this month Red Shark released its Lightworks movie editing program for Linux. It is a program that seems to be on par with anything from Avid or Adobe. It's even set up to create 3D movies, AND IT'S FREE!
You remind me of all those Photoshop killers (GIMP and Co.). A free program will just not have the final polish that professionals need. It will have some nice touches but will lack something that is badly needed. Just look at those free MS Office competitors. What percentage of people uses them?
One more things:) A Free program will never cut it as there is always chance that it will be abandoned by the developers. What if no update comes in years?
Quote:
Originally Posted by Jim W
Pros are making their decisions based on what comes from Apple by the end of this year. Hopefully Apple comes through.
Quote:
Originally Posted by Marvin
People said that last year though and the year before.
...and they did. Our primary Avid suite, once the foundation for Apple technology in our plant, is now running on HP machines. Engineering refused to pay premium coin for woefully outdated hardware so they went with Windows instead. On the audio side, the latest version of Pro Tools offers serious workflow improvements but won't run on our 2009 Mac. In the absence of a "new" Mac Pro the fear is that we'll wind up with Windows machines, too.
Inertia is hard to overcome in a production setting. It took Apple completely ignoring the Pro tower segment to uproot it. I imagine it will be similarly difficult to persuade the decision makers to even look at going back to Apple now that we've already gone through a migration. Why would they? There's a lot of effort involved in switching everything over, and I can't imagine there being much will to do it all again.