I agree that even if building and selling MP's is a net loss for Apple, it is important, at the very least, to maintain professional/workstation Operating System market share.
And, as much as I hesitate to EVER suggest that Apple would allow 3P hardware manufacturers again, I can see a possibility that Apple may partner with another producer of High End workstations (i.e. HP, etc.) to make the machines. They may even call it an apple and sell it on the Apple store. This would probably alleviate many of the operating losses caused by the MP.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
Quote:
Originally Posted by Jim W
A great deal of that computationally intensive processing takes place on the internal PCIe cards, such as NVidia Cuda graphics cards for Adobe, etc, etc.. Video I/O is on these cards. Where are there 16x Thunderbolt expansion options? The cruel fact is that most of that work will move to PC's. Heaven forbid! It will just be easier than Thunderbolt kludges connected to underpowered laptops and consumer level desktops. There is a halo effect from the high end creative work done on Mac Pros. I just hope Apple's new management realizes that. Not to mention that many people's livelihoods depend on them.
Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
A quick list of things indicating a troubled platform.
When updates do come they are uninspired.
The lack of TB support on the platform. Not that I find it compelling just that they used a laptop to introduce the technology.
Massive price bumps for no apparent reason.
The general lack of focus on new technology.
There is almost to public acknowledgement of the machines existence any more. Apple just doesn't promote the machine like it has in the past.
Now given this I suspect that Apple more than breaks even on hardware costs, that is the cost to assemble the machine. However I really doubt that they make enough money from the platform to justify the engineering expense to build the next iteration. In fact each new release looks like an attempt to limit engineering costs as much as possible.
Quote:
Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
There are a number of reasons why one shouldn't rely upon external expansion or maybe better said why such isn't desirable. I'm not even convinced that the current approach with PC based PCI-Express cards is all that great either. Unfortunately it is what we have as a standard right now. It wouldn't hurt Apple all that much to transition to a new technology here too. PCI-E 3 is soon to be here as is the transition to alternative storage tech, it is a really good time to focus on alternative board form factors. The right form factor could be applicable to the PRO, and the Mini. It could possibly be extended to the iMac and laptops.
Now I can see people already crying in their milk, why why why. Well it is simple, the industry needs a common format moving forward that supports PCI-E 3 and can be very low cost for the transition to solid state storage. In other words a primary motivator here is a card that effectively replaces the disk drive bay and the current PCI card slots.
Let's keep things under perspective. Despite all the legitimate complains from the professional users demanding the highest performance in a workstation, it is a very small market. In fact there wasn't enough money in this market to sustain SGI, SUN and all the other major workstation makers catering to this market. And the market using Apple workstations is probably even smaller since it seems its mostly graphics/video users, since lots of scientific/engineering software simply doesn't run in OSX.
Actually I think Apples market is probably bigger but the fact that it is bigger doesn't mean it is big enough. This is one of the reasons why I suggest that Apple needs to significantly expand the appeal of the Pro. It needs to be a platform that can be had in at least two different performance configurations. One should be built with a mainstream processor the other with a dual socket Xeon type mother board. They need to pull in sales from people that need the other features of the Pro besides hard thread count.
Quote:
In my area of work, all scientific simulation software that used to run exclusively on unix and required a very expensive branded workstation, was eventually ported to windows for desktop use, and linux to run in clusters for big jobs that a workstation simply cannot handle. OSX will never be supported, since it doesn't offer anything superior above linux for this purpose.
Actually I do see Apple getting some traction here. In many ways Mac OS/X is still more appealing than going the Windows route. Clusters though will likely stay with Linux.
Quote:
This situation is not new. so I doubt Apple is going to simply discontinue the MP now when they would have done so long ago when it was clear it had extremely limited commercial appeal. It seems they are content to have it available despite the low sales. In fact, HP Dell etc offerings are in the same price range, so I don't understand why people are complaining that they are a rip off.
The ripoff comes from the relatively high cost if you are looking for some of the expansion capability. For example I could effectively get buy with four cores and a solid OpenCL supporting video solution, something that could be had today in an iMac and probably in an Ivy Bridge Mini. However neither of those two platforms take care of other factors that are important, these are:
Ram expansion.
Storage expansion.
Maintainability.
I/O expansion.
You would have a very hard time convincing me that the Pro is any thing less than a ripoff if you are trying to answer these needs. For all of its power and usefulness the Mini falls flat on its face with all of the items above. It is especially shameful of Apple to realize that the MBPs are far more maintainable that the iMac or Mini.
The arguments supporting the Mac Pro because of the computational performance are very valid but that isn't the only reason to desire a Mac Pro. I suspect many mis this point when dismissing the Pro. For example the Pro is currently the only piece of Apple hardware that could even be considered as a storage server.
What I posted was relevant and representative of my personal experience. Why don't you move on to a different forum.
His request was rational in the context of this thread. To many have already posted going on and on about how well their G4 performs. The reality is if your hardware is fast enough for you or has the other capabilities you need then this forum isn't about your needs.
Quote:
Also, everyone, don't forget that just because Apple discontinues the MP, doesn't mean they are out of the workstation market for good. The MP's design, while elegant and extremely well engineered, is very dated.
Well this I agree with. The whole platform is in need of an overhaul.
Quote:
I agree that even if building and selling MP's is a net loss for Apple, it is important, at the very least, to maintain professional/workstation Operating System market share.
Trying to sell hardware at a lost has killed many a company. In Apples position though they would open themselves up to lawsuits.
Quote:
And, as much as I hesitate to EVER suggest that Apple would allow 3P hardware manufacturers again, I can see a possibility that Apple may partner with another producer of High End workstations (i.e. HP, etc.) to make the machines. They may even call it an apple and sell it on the Apple store. This would probably alleviate many of the operating losses caused by the MP.
What advantage would there be for Apple. All of this hardware is made in China, often in the same plants.
Quote:
Doing something like this for the iMac or the Mac Mini would be unthinkable, given the tight engineering specs and proprietary designs used in both. BUt the Mac Pro is pretty much just a shiny PC built with a really well designed case. There's no particular reason they couldn't license the case to HP, and HP couldn't obtain the same components through its sources and build an identical computer that would come preinstalled with OSX.
again there is no rational reason to do this. More importantly Apple needs to innovate on the desktop, you can't do that by relabeling somebody else's hardware.
Quote:
Again, I hesitate to suggest that Apple would ever again experiment with 3P hardware licenses. BUt if they were going to, the Mac Pro seems a logical place to do it.
The problem with the Mac Pro is that Apple is building a computer for the past not one that is suitable for the future. There are many technologies coming on line that will permit a radically different Mac Pro that allows the machine to focus on high performance. Apple just needs to focus on the hardware a bit to get there.
You have made some interesting points, especially in respect of SGI and Sun. A colleague and I however, ported a LINUX only scientific package to OS X with very little effort. The advantage? To be fair, this applies in my case if not in general but it relieved me of the need to run that awful OS, especially the great pretender Ubuntu. Honestly, to me, this was a great relief and gave me a complete processing chain under the one OS (the one to rule them all lol)! Others would disagree.
It's nice to see 8 instances of the application running in 8 cores, each one fully taxed.
All the best.
I agree that porting to OSX is not a big deal (depending on the libraries involved and licensing of any proprietary stuff used in the application ). However the market is already so small that it's not worth to offer full commercial support. Perhaps the market can grow if these companies were to target the academic market, selling a OSX version for university students in math/science/engineering ) like Mathematica and MatLab do) and this would increase demand for workstation quality hardware for the mac platform.
Now I can see people already crying in their milk, why why why. Well it is simple, the industry needs a common format moving forward that supports PCI-E 3 and can be very low cost for the transition to solid state storage. In other words a primary motivator here is a card that effectively replaces the disk drive bay and the current PCI card slots.
Considering the push for lower wattage consumption across basically all components and the push from mobile devices upward and servers down, it's pretty reasonable to expect some updates on standards for these things. What I hope doesn't catch on or continue is Apple's trend toward junk technology where things that used to be drop-in type replacements all become one week turnaround repair items. Overall longevity on things like batteries has improved, but component reliability as a whole is all over the place. Premature display backlight failure isn't as common these days as it was a few years ago, but hard drive reliability is in many cases down (both ssds and seagate which amuses me because it has "gate" in the name). Pushing everything into breakout boxes seems like a very interim design in an industry that has been focused on consolidation.
The overall dissection of the mac pro design by a few others has shown a lack of understanding over what would influence the price and audience. If we're talking about people using it for work, it comes down to TCO and what you get for it. Some of these guys have $10,000 or more in accompanying hardware addons, software, and peripheral devices needed to make it a fully usable machine. As for thunderbolt, if you need something today, it will probably cost more via a thunderbolt item for lesser performance, and Intel really needs to help push it. They keep hitting snags on cost or development and end up backing away each time.
I'm using a current 12 core Mac Pro and I have render times of up to 20 minutes a frame, sometimes longer. So I don't want to wait a few years for an iMac or Mac Mini to have the power of what is needed NOW.
...
. That will not be Apple if they kill the Mac Pro. As I will switch within 3 months of that announcement.
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
Considering the push for lower wattage consumption across basically all components and the push from mobile devices upward and servers down, it's pretty reasonable to expect some updates on standards for these things. What I hope doesn't catch on or continue is Apple's trend toward junk technology where things that used to be drop-in type replacements all become one week turnaround repair items. Overall longevity on things like batteries has improved, but component reliability as a whole is all over the place. Premature display backlight failure isn't as common these days as it was a few years ago, but hard drive reliability is in many cases down (both ssds and seagate which amuses me because it has "gate" in the name). Pushing everything into breakout boxes seems like a very interim design in an industry that has been focused on consolidation.
I can't say I blame you as your concern is justified. The problem is as things get smaller and more highly integrated the ability to effectively repair a device goes down. If the GPU in a Sandy Bridge chip goes north you loose the CPU.
I foresee a day when things like iPad and iPhone are one solid block of material with everything inside a mystery.
Quote:
The overall dissection of the mac pro design by a few others has shown a lack of understanding over what would influence the price and audience. If we're talking about people using it for work, it comes down to TCO and what you get for it. Some of these guys have $10,000 or more in accompanying hardware addons,
yes I realize that. My point is there isn't enough of those sorts of people to keep the Mac Pro profitable. This is the number one reason why I believe Apple has to find a way to appeal to more users with the next Pro. The idea is to pick up additional sales from users that don't need the flat out performance but rather need a chassis to serve other needs.
Quote:
software, and peripheral devices needed to make it a fully usable machine. As for thunderbolt, if you need something today, it will probably cost more via a thunderbolt item for lesser performance, and Intel really needs to help push it. They keep hitting snags on cost or development and end up backing away each time.
TB will fail if Intel can't open it up to competing hardware. The only way to get costs down is via micro controllers with build in TB ports. Many peripherals these days run on single chip processors, you can't even do that today with TB. Intel could give the TB chips away and still fail to grab the traction they need in the market place. Right at the moment I don't see TB being cost effective anytime soon. Worst yet we have yet to see an example of any device with more than one TB port, even if TB could replace internal slots right now there are no examples of machines with enough TB ports to replace internal slots.
I foresee a day when things like iPad and iPhone are one solid block of material with everything inside a mystery.
yes I realize that. My point is there isn't enough of those sorts of people to keep the Mac Pro profitable. This is the number one reason why I believe Apple has to find a way to appeal to more users with the next Pro. The idea is to pick up additional sales from users that don't need the flat out performance but rather need a chassis to serve other needs.
TB will fail if Intel can't open it up to competing hardware. The only way to get costs down is via micro controllers with build in TB ports. Many peripherals these days run on single chip processors, you can't even do that today with TB. Intel could give the TB chips away and still fail to grab the traction they need in the market place. Right at the moment I don't see TB being cost effective anytime soon. Worst yet we have yet to see an example of any device with more than one TB port, even if TB could replace internal slots right now there are no examples of machines with enough TB ports to replace internal slots.
27" imac has two thunderbolt ports. I don't know if they share bandwidth. The next generation chips are supposed to have more channels and lowered cost, but unlike usb3 thunderbolt isn't going to be part of the chipset.
I think Apple is reaching a point with this stuff where even if they started work on a successor to that line today, it would be too late. They missed a few features that really hurt their chances of going after windows users early on with the mac pro. I felt at the time to really keep that line going they needed to leverage into windows territory by converting people from Dell and HP workstations to Apple even if they were still running Windows. They had a really strong price point but just lacked a few features. If they got more workstation users using macs regardless of what OS was run, the machine would potentially be in a better spot today.
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
Yes, that is absolutely true! But in 18 months, I better not still be using this 12 core machine, I better be using a 16 or 24 or ??? what ever the next faster revision is.
That's why PRO is attached to the name Mac... Professionals require updates..
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
But is he going to be rendering the same size files in 18 months? Hell, if I was still working with MiniDV video, I could still be using my old G5 tower.
But is he going to be rendering the same size files in 18 months? Hell, if I was still working with MiniDV video, I could still be using my old G5 tower.
I don't see why not. The last stop for video quality is Super Hi-Vision, so once we start shooting in that, rendering speeds will only get faster.
Now for CGI scenes, that's always going to need faster crap.
Seriously I'm not obsessed with the form factor of the device, I'm more concerned about getting the right features at the right price.
Now that doesn't mean there aren't preferred dimensions or form factors. Ideally the device would be more of a rectangle that sits well on a shelf or desk. A ball would really sit well at all. This rectangle should be compatible with standard Rack mount heights and a fraction of a standard width. The idea being to be able to sit two or three of them side by side in a rack. So we are talking something around 5 or 8 or so inches in width, 3U in height and maybe 12 inches deep.
The limitation on depth is there to allow thIs XMac to sit on a book shelf without fear of it falling off. The whole idea with these dimensions is one machine that can be easily deployed in a number of different situations. Hell add a few clips to make bolting it to a wall easy.
Most importantly they should put in the effort to make it easy to service. The most pathetic thing about Apple right now is that their laptops are easier to work on that things like the Mini or iMac. Desktop machines by their nature should be very easy to access, update or repair.
In the end I'd be happy with just about anything at the right price point and capability level.
Quote:
Originally Posted by ceti331
I'll second the 'beefed up mac mini' thoughts here, how about another cube shaped computer with the same footprint,
I would have bought another one long ago. Sure they are slow compared to iMacs. But, that's only because the iMacs have been kept up to date.
I need an easily expandable desktop computer for our photography/video business. Maybe they can do it on a smaller footprint. I really don't care. The size doesn't bug me. And I love the ease of access to its internals.
Apple's stalling in updating the Mac Pro is making decreasing sales a self-fufilling prophecy.
I would have bought another one long ago. Sure they are slow compared to iMacs. But, that's only because the iMacs have been kept up to date.
Most likely in a single threaded app but the Pro can be much faster with heavily threaded apps. Not to mention winning hands down multitasking.
Quote:
I need an easily expandable desktop computer for our photography/video business. Maybe they can do it on a smaller footprint. I really don't care. The size doesn't bug me. And I love the ease of access to its internals.
Apple's stalling in updating the Mac Pro is making decreasing sales a self-fufilling prophecy.
Yep. However realize that it does appear in this case that Intel is screwing up. Either way the lack of an update hurts. However the self fulfilling prophecy is a reality, the Pro just doesn't meet customer needs.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
you do realize the potential throughput for a thunderbolt lane running over opitcal wiring is 100Gps? This is according to Intel. Granted, Apple's current implementation uses Thunderbolt over copper wiring but as the tech advances it may be highly unlikely you'll need internal expansion cards, even for GPU's.
you do realize the potential throughput for a thunderbolt lane running over opitcal wiring is 100Gps? This is according to Intel. Granted, Apple's current implementation uses Thunderbolt over copper wiring but as the tech advances it may be highly unlikely you'll need internal expansion cards, even for GPU's.
Intel said it could reach 100 gigabits per second in a decade from now. I mention this just for clarification as too many people get messed up with gigabits versus gigabytes. Many other things will change within ten years as well, and this is assuming they don't run into problems. It's not an instant bandwidth change just by using fibre cables, and Intel has warned us they will be expensive. They've mentioned things like it's a matter of how much bandwidth people are willing to pay for.
Dadi Perlmutter of Intel's Architecture Group said copper wires are working much better than expected, and that fiber was expensive. 'It's going to be way out,' Perlmutter said. 'At the end of the day it's all about how much speed people need versus how much they would be willing to pay.'"
There are articles like that all over the internet about this. It will not be instant and it will cost real money. They are not going to integrate it into the chipsets next year. It will not automatically make it into every PC. Even in its current form it's an excellent step up from something like usb, but it's not a magical solution to all of your problems. The solution of may work well enough as a catch all two to three years from now doesn't really help if you need a computer before then.
Yep. However realize that it does appear in this case that Intel is screwing up. Either way the lack of an update hurts. However the self fulfilling prophecy is a reality, the Pro just doesn't meet customer needs.
While Intel is not behind in their Xeon server class processor line, they are still making faster chips. A single chip mac pro running a quad core I7 at 3.4Ghz would be a decent upgrade from the current entry level tower, and be more affordable. Apple's obsession with streamlining is actually hurting them in this instance.
Comments
I agree that even if building and selling MP's is a net loss for Apple, it is important, at the very least, to maintain professional/workstation Operating System market share.
And, as much as I hesitate to EVER suggest that Apple would allow 3P hardware manufacturers again, I can see a possibility that Apple may partner with another producer of High End workstations (i.e. HP, etc.) to make the machines. They may even call it an apple and sell it on the Apple store. This would probably alleviate many of the operating losses caused by the MP.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
A great deal of that computationally intensive processing takes place on the internal PCIe cards, such as NVidia Cuda graphics cards for Adobe, etc, etc.. Video I/O is on these cards. Where are there 16x Thunderbolt expansion options? The cruel fact is that most of that work will move to PC's. Heaven forbid! It will just be easier than Thunderbolt kludges connected to underpowered laptops and consumer level desktops. There is a halo effect from the high end creative work done on Mac Pros. I just hope Apple's new management realizes that. Not to mention that many people's livelihoods depend on them.
Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
A quick list of things indicating a troubled platform.
- When updates do come they are uninspired.
- The lack of TB support on the platform. Not that I find it compelling just that they used a laptop to introduce the technology.
- Massive price bumps for no apparent reason.
- The general lack of focus on new technology.
- There is almost to public acknowledgement of the machines existence any more. Apple just doesn't promote the machine like it has in the past.
Now given this I suspect that Apple more than breaks even on hardware costs, that is the cost to assemble the machine. However I really doubt that they make enough money from the platform to justify the engineering expense to build the next iteration. In fact each new release looks like an attempt to limit engineering costs as much as possible.Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
There are a number of reasons why one shouldn't rely upon external expansion or maybe better said why such isn't desirable. I'm not even convinced that the current approach with PC based PCI-Express cards is all that great either. Unfortunately it is what we have as a standard right now. It wouldn't hurt Apple all that much to transition to a new technology here too. PCI-E 3 is soon to be here as is the transition to alternative storage tech, it is a really good time to focus on alternative board form factors. The right form factor could be applicable to the PRO, and the Mini. It could possibly be extended to the iMac and laptops.
Now I can see people already crying in their milk, why why why. Well it is simple, the industry needs a common format moving forward that supports PCI-E 3 and can be very low cost for the transition to solid state storage. In other words a primary motivator here is a card that effectively replaces the disk drive bay and the current PCI card slots.
Let's keep things under perspective. Despite all the legitimate complains from the professional users demanding the highest performance in a workstation, it is a very small market. In fact there wasn't enough money in this market to sustain SGI, SUN and all the other major workstation makers catering to this market. And the market using Apple workstations is probably even smaller since it seems its mostly graphics/video users, since lots of scientific/engineering software simply doesn't run in OSX.
Actually I think Apples market is probably bigger but the fact that it is bigger doesn't mean it is big enough. This is one of the reasons why I suggest that Apple needs to significantly expand the appeal of the Pro. It needs to be a platform that can be had in at least two different performance configurations. One should be built with a mainstream processor the other with a dual socket Xeon type mother board. They need to pull in sales from people that need the other features of the Pro besides hard thread count.
In my area of work, all scientific simulation software that used to run exclusively on unix and required a very expensive branded workstation, was eventually ported to windows for desktop use, and linux to run in clusters for big jobs that a workstation simply cannot handle. OSX will never be supported, since it doesn't offer anything superior above linux for this purpose.
Actually I do see Apple getting some traction here. In many ways Mac OS/X is still more appealing than going the Windows route. Clusters though will likely stay with Linux.
This situation is not new. so I doubt Apple is going to simply discontinue the MP now when they would have done so long ago when it was clear it had extremely limited commercial appeal. It seems they are content to have it available despite the low sales. In fact, HP Dell etc offerings are in the same price range, so I don't understand why people are complaining that they are a rip off.
The ripoff comes from the relatively high cost if you are looking for some of the expansion capability. For example I could effectively get buy with four cores and a solid OpenCL supporting video solution, something that could be had today in an iMac and probably in an Ivy Bridge Mini. However neither of those two platforms take care of other factors that are important, these are:
- Ram expansion.
- Storage expansion.
- Maintainability.
- I/O expansion.
You would have a very hard time convincing me that the Pro is any thing less than a ripoff if you are trying to answer these needs. For all of its power and usefulness the Mini falls flat on its face with all of the items above. It is especially shameful of Apple to realize that the MBPs are far more maintainable that the iMac or Mini.The arguments supporting the Mac Pro because of the computational performance are very valid but that isn't the only reason to desire a Mac Pro. I suspect many mis this point when dismissing the Pro. For example the Pro is currently the only piece of Apple hardware that could even be considered as a storage server.
Seriously??
What I posted was relevant and representative of my personal experience. Why don't you move on to a different forum.
His request was rational in the context of this thread. To many have already posted going on and on about how well their G4 performs. The reality is if your hardware is fast enough for you or has the other capabilities you need then this forum isn't about your needs.
Also, everyone, don't forget that just because Apple discontinues the MP, doesn't mean they are out of the workstation market for good. The MP's design, while elegant and extremely well engineered, is very dated.
Well this I agree with. The whole platform is in need of an overhaul.
I agree that even if building and selling MP's is a net loss for Apple, it is important, at the very least, to maintain professional/workstation Operating System market share.
Trying to sell hardware at a lost has killed many a company. In Apples position though they would open themselves up to lawsuits.
And, as much as I hesitate to EVER suggest that Apple would allow 3P hardware manufacturers again, I can see a possibility that Apple may partner with another producer of High End workstations (i.e. HP, etc.) to make the machines. They may even call it an apple and sell it on the Apple store. This would probably alleviate many of the operating losses caused by the MP.
What advantage would there be for Apple. All of this hardware is made in China, often in the same plants.
Doing something like this for the iMac or the Mac Mini would be unthinkable, given the tight engineering specs and proprietary designs used in both. BUt the Mac Pro is pretty much just a shiny PC built with a really well designed case. There's no particular reason they couldn't license the case to HP, and HP couldn't obtain the same components through its sources and build an identical computer that would come preinstalled with OSX.
again there is no rational reason to do this. More importantly Apple needs to innovate on the desktop, you can't do that by relabeling somebody else's hardware.
Again, I hesitate to suggest that Apple would ever again experiment with 3P hardware licenses. BUt if they were going to, the Mac Pro seems a logical place to do it.
The problem with the Mac Pro is that Apple is building a computer for the past not one that is suitable for the future. There are many technologies coming on line that will permit a radically different Mac Pro that allows the machine to focus on high performance. Apple just needs to focus on the hardware a bit to get there.
You have made some interesting points, especially in respect of SGI and Sun. A colleague and I however, ported a LINUX only scientific package to OS X with very little effort. The advantage? To be fair, this applies in my case if not in general but it relieved me of the need to run that awful OS, especially the great pretender Ubuntu. Honestly, to me, this was a great relief and gave me a complete processing chain under the one OS (the one to rule them all lol)! Others would disagree.
It's nice to see 8 instances of the application running in 8 cores, each one fully taxed.
All the best.
I agree that porting to OSX is not a big deal (depending on the libraries involved and licensing of any proprietary stuff used in the application ). However the market is already so small that it's not worth to offer full commercial support. Perhaps the market can grow if these companies were to target the academic market, selling a OSX version for university students in math/science/engineering ) like Mathematica and MatLab do) and this would increase demand for workstation quality hardware for the mac platform.
Now I can see people already crying in their milk, why why why. Well it is simple, the industry needs a common format moving forward that supports PCI-E 3 and can be very low cost for the transition to solid state storage. In other words a primary motivator here is a card that effectively replaces the disk drive bay and the current PCI card slots.
Considering the push for lower wattage consumption across basically all components and the push from mobile devices upward and servers down, it's pretty reasonable to expect some updates on standards for these things. What I hope doesn't catch on or continue is Apple's trend toward junk technology where things that used to be drop-in type replacements all become one week turnaround repair items. Overall longevity on things like batteries has improved, but component reliability as a whole is all over the place. Premature display backlight failure isn't as common these days as it was a few years ago, but hard drive reliability is in many cases down (both ssds and seagate which amuses me because it has "gate" in the name). Pushing everything into breakout boxes seems like a very interim design in an industry that has been focused on consolidation.
The overall dissection of the mac pro design by a few others has shown a lack of understanding over what would influence the price and audience. If we're talking about people using it for work, it comes down to TCO and what you get for it. Some of these guys have $10,000 or more in accompanying hardware addons, software, and peripheral devices needed to make it a fully usable machine. As for thunderbolt, if you need something today, it will probably cost more via a thunderbolt item for lesser performance, and Intel really needs to help push it. They keep hitting snags on cost or development and end up backing away each time.
I'm using a current 12 core Mac Pro and I have render times of up to 20 minutes a frame, sometimes longer. So I don't want to wait a few years for an iMac or Mac Mini to have the power of what is needed NOW.
...
. That will not be Apple if they kill the Mac Pro. As I will switch within 3 months of that announcement.
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
Considering the push for lower wattage consumption across basically all components and the push from mobile devices upward and servers down, it's pretty reasonable to expect some updates on standards for these things. What I hope doesn't catch on or continue is Apple's trend toward junk technology where things that used to be drop-in type replacements all become one week turnaround repair items. Overall longevity on things like batteries has improved, but component reliability as a whole is all over the place. Premature display backlight failure isn't as common these days as it was a few years ago, but hard drive reliability is in many cases down (both ssds and seagate which amuses me because it has "gate" in the name). Pushing everything into breakout boxes seems like a very interim design in an industry that has been focused on consolidation.
I can't say I blame you as your concern is justified. The problem is as things get smaller and more highly integrated the ability to effectively repair a device goes down. If the GPU in a Sandy Bridge chip goes north you loose the CPU.
I foresee a day when things like iPad and iPhone are one solid block of material with everything inside a mystery.
The overall dissection of the mac pro design by a few others has shown a lack of understanding over what would influence the price and audience. If we're talking about people using it for work, it comes down to TCO and what you get for it. Some of these guys have $10,000 or more in accompanying hardware addons,
yes I realize that. My point is there isn't enough of those sorts of people to keep the Mac Pro profitable. This is the number one reason why I believe Apple has to find a way to appeal to more users with the next Pro. The idea is to pick up additional sales from users that don't need the flat out performance but rather need a chassis to serve other needs.
software, and peripheral devices needed to make it a fully usable machine. As for thunderbolt, if you need something today, it will probably cost more via a thunderbolt item for lesser performance, and Intel really needs to help push it. They keep hitting snags on cost or development and end up backing away each time.
TB will fail if Intel can't open it up to competing hardware. The only way to get costs down is via micro controllers with build in TB ports. Many peripherals these days run on single chip processors, you can't even do that today with TB. Intel could give the TB chips away and still fail to grab the traction they need in the market place. Right at the moment I don't see TB being cost effective anytime soon. Worst yet we have yet to see an example of any device with more than one TB port, even if TB could replace internal slots right now there are no examples of machines with enough TB ports to replace internal slots.
I foresee a day when things like iPad and iPhone are one solid block of material with everything inside a mystery.
yes I realize that. My point is there isn't enough of those sorts of people to keep the Mac Pro profitable. This is the number one reason why I believe Apple has to find a way to appeal to more users with the next Pro. The idea is to pick up additional sales from users that don't need the flat out performance but rather need a chassis to serve other needs.
TB will fail if Intel can't open it up to competing hardware. The only way to get costs down is via micro controllers with build in TB ports. Many peripherals these days run on single chip processors, you can't even do that today with TB. Intel could give the TB chips away and still fail to grab the traction they need in the market place. Right at the moment I don't see TB being cost effective anytime soon. Worst yet we have yet to see an example of any device with more than one TB port, even if TB could replace internal slots right now there are no examples of machines with enough TB ports to replace internal slots.
27" imac has two thunderbolt ports. I don't know if they share bandwidth. The next generation chips are supposed to have more channels and lowered cost, but unlike usb3 thunderbolt isn't going to be part of the chipset.
I think Apple is reaching a point with this stuff where even if they started work on a successor to that line today, it would be too late. They missed a few features that really hurt their chances of going after windows users early on with the mac pro. I felt at the time to really keep that line going they needed to leverage into windows territory by converting people from Dell and HP workstations to Apple even if they were still running Windows. They had a really strong price point but just lacked a few features. If they got more workstation users using macs regardless of what OS was run, the machine would potentially be in a better spot today.
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
Yes, that is absolutely true! But in 18 months, I better not still be using this 12 core machine, I better be using a 16 or 24 or ??? what ever the next faster revision is.
That's why PRO is attached to the name Mac... Professionals require updates..
If they drop it, your rig will still render at the same speed for the next 18 months. That's the point.
But is he going to be rendering the same size files in 18 months? Hell, if I was still working with MiniDV video, I could still be using my old G5 tower.
But is he going to be rendering the same size files in 18 months? Hell, if I was still working with MiniDV video, I could still be using my old G5 tower.
I don't see why not. The last stop for video quality is Super Hi-Vision, so once we start shooting in that, rendering speeds will only get faster.
Now for CGI scenes, that's always going to need faster crap.
Now that doesn't mean there aren't preferred dimensions or form factors. Ideally the device would be more of a rectangle that sits well on a shelf or desk. A ball would really sit well at all. This rectangle should be compatible with standard Rack mount heights and a fraction of a standard width. The idea being to be able to sit two or three of them side by side in a rack. So we are talking something around 5 or 8 or so inches in width, 3U in height and maybe 12 inches deep.
The limitation on depth is there to allow thIs XMac to sit on a book shelf without fear of it falling off. The whole idea with these dimensions is one machine that can be easily deployed in a number of different situations. Hell add a few clips to make bolting it to a wall easy.
Most importantly they should put in the effort to make it easy to service. The most pathetic thing about Apple right now is that their laptops are easier to work on that things like the Mini or iMac. Desktop machines by their nature should be very easy to access, update or repair.
In the end I'd be happy with just about anything at the right price point and capability level.
I'll second the 'beefed up mac mini' thoughts here, how about another cube shaped computer with the same footprint,
I need an easily expandable desktop computer for our photography/video business. Maybe they can do it on a smaller footprint. I really don't care. The size doesn't bug me. And I love the ease of access to its internals.
Apple's stalling in updating the Mac Pro is making decreasing sales a self-fufilling prophecy.
I would have bought another one long ago. Sure they are slow compared to iMacs. But, that's only because the iMacs have been kept up to date.
Most likely in a single threaded app but the Pro can be much faster with heavily threaded apps. Not to mention winning hands down multitasking.
I need an easily expandable desktop computer for our photography/video business. Maybe they can do it on a smaller footprint. I really don't care. The size doesn't bug me. And I love the ease of access to its internals.
Apple's stalling in updating the Mac Pro is making decreasing sales a self-fufilling prophecy.
Yep. However realize that it does appear in this case that Intel is screwing up. Either way the lack of an update hurts. However the self fulfilling prophecy is a reality, the Pro just doesn't meet customer needs.
It's extremely unlikely that Apple is losing money on the mac pro. Other things have outpaced them in recent years. Lack of consistent updates from intel hasn't helped at all. The longer refresh cycle encourages people to wait. If you've owned a machine for say 2-3 years and it's starting to feel like you need more power, you won't want to buy something that has gone unchanged for more than a year.
Even the 16x cards don't fully saturate that many lanes but if they saturate more than 8 lanes, 16 is the next step. I agree thunderbolt isn't a replacement for that kind of bandwidth. Even if intel pushed it and stepped up the bandwidth, you'd still run into an issue on wattage.
you do realize the potential throughput for a thunderbolt lane running over opitcal wiring is 100Gps? This is according to Intel. Granted, Apple's current implementation uses Thunderbolt over copper wiring but as the tech advances it may be highly unlikely you'll need internal expansion cards, even for GPU's.
you do realize the potential throughput for a thunderbolt lane running over opitcal wiring is 100Gps? This is according to Intel. Granted, Apple's current implementation uses Thunderbolt over copper wiring but as the tech advances it may be highly unlikely you'll need internal expansion cards, even for GPU's.
Intel said it could reach 100 gigabits per second in a decade from now. I mention this just for clarification as too many people get messed up with gigabits versus gigabytes. Many other things will change within ten years as well, and this is assuming they don't run into problems. It's not an instant bandwidth change just by using fibre cables, and Intel has warned us they will be expensive. They've mentioned things like it's a matter of how much bandwidth people are willing to pay for.
Dadi Perlmutter of Intel's Architecture Group said copper wires are working much better than expected, and that fiber was expensive. 'It's going to be way out,' Perlmutter said. 'At the end of the day it's all about how much speed people need versus how much they would be willing to pay.'"
There are articles like that all over the internet about this. It will not be instant and it will cost real money. They are not going to integrate it into the chipsets next year. It will not automatically make it into every PC. Even in its current form it's an excellent step up from something like usb, but it's not a magical solution to all of your problems.
Yep. However realize that it does appear in this case that Intel is screwing up. Either way the lack of an update hurts. However the self fulfilling prophecy is a reality, the Pro just doesn't meet customer needs.
While Intel is not behind in their Xeon server class processor line, they are still making faster chips. A single chip mac pro running a quad core I7 at 3.4Ghz would be a decent upgrade from the current entry level tower, and be more affordable. Apple's obsession with streamlining is actually hurting them in this instance.