The Mac Pro is the "Kwisatz Haderach" of all computers, and Apple is considering its relevance as a viable product. Come on Steve enforce your will in the after life! Make them not just continue it but make it better, faster, stronger!
Mac Daily has a poll - should the Mac Pro line be discontinued ? :
There are 9,000 votes - and the "Yes - discontinue" votes are winning 51% to 44% for keeping the line. Most of the voters probably own nothing but an iPod or an iPhone ....
you know I've been thinking perhaps the conventional workstation PC is really old hat now no pun to linux intended but new Arm developements are beginning to impress and having said multinode transputational computing may be the way of the future then I came accross HPs Moonshot server using quad card Arm system on a chip design and realised that would also make a really fast video render workstation very much in the fashion of tesla GPU processing.
OK so the HP design is very large and IP server centric but the principal is sound.
Its could be a bit like taking 8 macMinis on cards and integrating them in one Macpro X case. the result would be a local MacOS server to satify professional use that could also be a lone workstation or indeed serve several iMacs in a studio multiclient setup at a greatly reduced low energy cost alternative to using Intel Zeons.
It is sort of like the occupy wall street gang. Protest something you don't have the mental capacity to understand. The parallels are more than minor here.
Quote:
Originally Posted by BigPhotos
Mac Daily has a poll - should the Mac Pro line be discontinued ? :
There are 9,000 votes - and the "Yes - discontinue" votes are winning 51% to 44% for keeping the line. Most of the voters probably own nothing but an iPod or an iPhone ....
There is one problem with this poll. It really doesn't ask the question of what to replace it with. In my estimation the Pro is a dead platform in its current configuration. Apple still needs a High performance machine to serve the professional markets but they need to deliver that machine at the right price point. The platform needs to attract a wider array of users to maintain viability.
Obviously this HP approach is a bit to focused on server Workloads but many of the concepts reflect future capabilities.
Quote:
Originally Posted by AllanMc
you know I've been thinking perhaps the conventional workstation PC is really old hat now no pun to linux intended but new Arm developements are beginning to impress and having said multinode transputational computing may be the way of the future
Just realize that most Pro users are more focused on high performance computing rather than server duties. A massive number of ARM processors actually makes good sense for many server workloads.
Quote:
then I came accross HPs Moonshot server using quad card Arm system on a chip design and realised that would also make a really fast video render workstation very much in the fashion of tesla GPU processing.
OK so the HP design is very large and IP server centric but the principal is sound.
Yes for a server it is a sound approach. For a Mac Pro it would need modification as Arm performance simply isn't good enough for the workloads a Mac Pro sees. The Mac Pro needs to be refactored, there is no doubt in my mind here. The problem is the modular approach can be expensive so Apple would have to over come that issue.
Quote:
Its could be a bit like taking 8 macMinis on cards and integrating them in one Macpro X case. the result would be a local MacOS server to satify professional use that could also be a lone workstation or indeed serve several iMacs in a studio multiclient setup at a greatly reduced low energy cost alternative to using Intel Zeons.
8 cards implies a big heavy chassis that is also expensive thus one machine doesn't address the problems Apple is having on the desktop. In stead imagine a family of desktop machines that all take the same compute care. The Mini might take one card, the XMac two and the Pro eight. With one design effort and careful selection of clock rates Apple could cover most desktop needs.
The thing here isn't the specifics though it is rather the fact that Apple needs to innovate once again on the desktop. Traditional PCs may be seeing reduced demand, but that might be an indication of hardware not meeting user needs.
Obviously this HP approach is a bit to focused on server Workloads but many of the concepts reflect future capabilities.
Just realize that most Pro users are more focused on high performance computing rather than server duties. A massive number of ARM processors actually makes good sense for many server workloads.
Yes for a server it is a sound approach. For a Mac Pro it would need modification as Arm performance simply isn't good enough for the workloads a Mac Pro sees..
In the HPvideo http://youtu.be/4PIajg_Htx0 at imin24 the presenter shows a card about 1/2hight PCIexpress size containing 4Arm processors presumably quad core or at least by A15 it will be sometime next year I've read sugested, complete with independent memory modules, thats 16cores on one card and 8core processsors arn't too far off either, at 22nm thats a cool processing stack that Zeons will have trouble matching up to in future
Quote:
The Mac Pro needs to be refactored, there is no doubt in my mind here. The problem is the modular approach can be expensive so Apple would have to over come that issue.
8 cards implies a big heavy chassis that is also expensive thus one machine doesn't address the problems Apple is having on the desktop.
In stead imagine a family of desktop machines that all take the same compute care. The Mini might take one card, the XMac two and the Pro eight. With one design effort and careful selection of clock rates Apple could cover most desktop needs.
It doesn't take such a large leap of imagination to consider a svelt MacPro concept casing to hold a 6or8way PCIe motherboard with a slim down the side 2" PSU ideal for studio rack mounting with 3 or 4 super silent thin fans. a much smaller and simpler unit than todays MacPro and you don't need to exaggerate dimensions to make it seem unreasonably large.
This could have say 3 Arm processor PCIe boards thats (48cores) as described above and a duel PCIeGPU card plus a couple of PCIeSSD cards and an interface card for specialist requirements, remember cost isn't the limiting factor on the companies flagship infact cheap might be sending the wrong message which all reflects on the whole Apple range.
Now consider the exact same footprint but Half hight chassis using the same motherboard PCIexpress design for a MacX semiPro workstation at more reasonable cost.
It doesn't have to be complicated any further, there are 2.5" and 3.5" HD PCIe hybrid cards aswell for the cash strained, and I'd imagine it would sporn many more card designs.
These workstations could be totom poled with TB through connections for the more extreme uses like calculating the latest Gnome or rendering the latest "Avater" movie,
Quote:
The thing here isn't the specifics though it is rather the fact that Apple needs to innovate once again on the desktop. Traditional PCs may be seeing reduced demand, but that might be an indication of hardware not meeting user needs
I imagine eventually PCIe 4 will double the system speed again in due course but professionals will need this flexibility whether it be Mac or PC for quite some time to come.
And such flexibility would give Apple a platform to employ multiple architecture strategies into the future.
AllanMc; where are you from? I ask because your posts are hard to follow.
Quote:
Originally Posted by AllanMc
In the HPvideo http://youtu.be/4PIajg_Htx0 at imin24 the presenter shows a card about 1/2hight PCIexpress size containing 4Arm processors presumably quad core or at least by A15 it will be sometime next year I've read sugested, complete with independent memory modules, thats 16cores on one card and 8core processsors arn't too far off either, at 22nm thats a cool processing stack that Zeons will have trouble matching up to in future
Remember we are talking ARM cores here which often have a hard time keeping up with Intel hardware. Lots of cores work well in servers because you generate lots of threads and each processor can handle a single thread. desktop apps often rely more on single core performance.
Quote:
It doesn't take such a large leap of imagination to consider a svelt MacPro concept casing to hold a 6or8way PCIe motherboard with a slim down the side 2" PSU ideal for studio rack mounting with 3 or 4 super silent thin fans. a much smaller and simpler unit than todays MacPro and you don't need to exaggerate dimensions to make it seem unreasonably large.
I'm all for smaller hardware. That being said I still think Apple needs a broader array of hardware.
Quote:
This could have say 3 Arm processor PCIe boards thats (48cores) as described above and a duel PCIeGPU card plus a couple of PCIeSSD cards and an interface card for specialist requirements, remember cost isn't the limiting factor on the companies flagship infact cheap might be sending the wrong message which all reflects on the whole Apple range.
NO - expensive for no reason sends the wrong message as does forcing people to buy configurations they don't want. In fact Apple has been sending the wrong message for years now. Basically that message is this if you need slots (RAM or PCI) or drive bays, please bend over.
Quote:
Now consider the exact same footprint but Half hight chassis using the same motherboard PCIexpress design for a MacX semiPro workstation at more reasonable cost.
It doesn't have to be complicated any further, there are 2.5" and 3.5" HD PCIe hybrid cards aswell for the cash strained, and I'd imagine it would sporn many more card designs.
These workstations could be totom poled with TB through connections for the more extreme uses like calculating the latest Gnome or rendering the latest "Avater" movie,
I imagine eventually PCIe 4 will double the system speed again in due course but professionals will need this flexibility whether it be Mac or PC for quite some time to come.
And such flexibility would give Apple a platform to employ multiple architecture strategies into the future.
Yes, but with a matte screen, or your own screen. I have a mini server in the office, if it's running something in the background I like being able to turn off the monitor and walk away, no worries about something moving my mouse and turning it back on. I like that fact it's matte too.
I thought about a Pro for expandability, but was worried about the noise level, are they quite noisy (compared to virtually silent mini?) I have never heard one running.
The loudest part of a Mac Pro is the hard drive. It's not like the PowerMac G5, I'll tell you.
I think the time is over for consumer towers but imagine a Mac Pro without any HDDs, but instead just a row of the SSD cards used in the MBAs. That would be both impressive in speed and price.
AllanMc; where are you from? I ask because your posts are hard to follow.
The Uk, ie where Jonathon Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
Quote:
Remember we are talking ARM cores here which often have a hard time keeping up with Intel hardware. Lots of cores work well in servers because you generate lots of threads and each processor can handle a single thread. desktop apps often rely more on single core performance.
True... at the moment, but ARM A15 is not far around the corner and ProApps are using multiple threads now and will increase in the working future,
Quote:
I'm all for smaller hardware. That being said I still think Apple needs a broader array of hardware.
Cost to develop is the problem, if one design concept can be flexible enough to encompass all Pro requirements then apple would have something to move forward with, at the moment Optical drives and Hard-drives are an outgoing technology, SATA is a slow interface for such outgoing drives but PCIe3 is a fast flexible architecture and also directly connects fast TB connectivity. why invent more when a suitable already supported structure already exists.
Quote:
NO - expensive for no reason sends the wrong message as does forcing people to buy configurations they don't want. In fact Apple has been sending the wrong message for years now. Basically that message is this if you need slots (RAM or PCI) or drive bays, please bend over.
Its my turn not to understand your meaning here? PCs are ultimately flexible, you can build whatever you want, only the MacPro offers pro's the same relative flexibility on the MacOS platform, otherwise there's no competition and all Processionals will emigrate to the platform that provides what they need, and that is flexibility to spend whatever they want to spend on their high end business machine.
Quote:
Simple is all most always the best way. But what you describe isn't simple
You can't get more simpler than a Case - PSU and a motherboard, and 3rd party PCI cards , the up and coming Window8 accommodates up to 160 cores compatible parallel processing, MacOS does multiple cores too and there is a good amount of highend workstation class software that uses them too.
The question is really - Do Apple want to competitively support a provisional market or not?
The Uk, ie where Steve Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
True, but there is a very real future for certain server farms using ARM processors. There is just too many avenues in which ARM can save money for an equivalent setup for certain usage types.
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
What if they design a chipset that can have eight processors on a board?
And my emoticon for the above question is "/2" to show that I'm only half joking about that. With the insanely lower power draws of ARM chips, would something like that truly be impractical in the future?
What if they design a chipset that can have eight processors on a board?
And my emoticon for the above question is "/2" to show that I'm only half joking about that. With the insanely lower power draws of ARM chips, would something like that truly be impractical in the future?
Intel processors are greater than an order of magnitude more powerful, are we going to have 100 core Mac Pros?
There are diminishing returns for adding more cores in most desktop activities.
ARM processors use so much less power, because they are much less powerful.
Intel processors are greater than an order of magnitude more powerful, are we going to have 100 core Mac Pros?
Well, since Intel currently has no plans for quantum computers and will reach the minimum physical size of transistor gateways sometime near the end of this decade, I'd say we'll have do something different.
I do see your point, though.
Intel's biggest fear is ARM in laptops and low-end desktops, though. Will Intel eventually be relegated to become the next IBM? Making only super-high-end chips and customized stuff?
Well, since Intel currently has no plans for quantum computers and will reach the minimum physical size of transistor gateways sometime near the end of this decade, I'd say we'll have do something different.
I do see your point, though.
Intel's biggest fear is ARM in laptops and low-end desktops, though. Will Intel eventually be relegated to become the next IBM? Making only super-high-end chips and customized stuff?
The thing is though, IBM (at least via their stock price) seems to be doing fairly well. The future will be very interesting.
The Uk, ie where Steve Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
You lost a bit of credibility there as Apple didn't even have a gaming play until the ARM hardware arrived.
Quote:
True... at the moment, but ARM A15 is not far around the corner and ProApps are using multiple threads now and will increase in the working future,
I'm not convinced A15 will be all that attractive.
Quote:
Cost to develop is the problem, if one design concept can be flexible enough to encompass all Pro requirements then apple would have something to move forward with, at the moment Optical drives and Hard-drives are an outgoing technology, SATA is a slow interface for such outgoing drives but PCIe3 is a fast flexible architecture and also directly connects fast TB connectivity. why invent more when a suitable already supported structure already exists.
The Mac Pro is a single platform that trys to appeal to everybody and has as a result been a failure. I don't think Apple has a chance in hell with a one size fits all machine. Modules usable in different machines is another discussion though.
Quote:
Its my turn not to understand your meaning here? PCs are ultimately flexible, you can build whatever you want, only the MacPro offers pro's the same relative flexibility on the MacOS platform, otherwise there's no competition and all Processionals will emigrate to the platform that provides what they need, and that is flexibility to spend whatever they want to spend on their high end business machine.
You are wrong there, corporations don't give people free reign over what they can spend for computing hardware. When you have Apple hardware that is often three times the cost of PC hardware capable of doing the job you don't have a chance to even suggest Apple hardware.
The problem is Apple doesn't have the right box, expandable box, at the right price point.
Quote:
You can't get more simpler than a Case - PSU and a motherboard, and 3rd party PCI cards , the up and coming Window8 accommodates up to 160 cores compatible parallel processing, MacOS does multiple cores too and there is a good amount of highend workstation class software that uses them too.
The problem is that is what the Mini is. The Mini is a fine machine as long as your needs are simple.
Quote:
The question is really - Do Apple want to competitively support a provisional market or not?
Professional? Yes I want them to support professional markets but there is a whole array of professional markets that don't need nor can the justify the expense of a Mac Pro. Because the Pros market is so small Apple risks loosing the Pro market anyways due to the lack of sales. In a nut shell they have to offer a better platform to move forward. Technology can do a lot to make that better platform, but Apple seems to be unwilling to spend the time and effort on the desktop.
You guys are doing an awful lot of mental gymnastics to avoid having an actual workstation class machine. I'd hate to see what you'd come up with if asked to design a bulldozer
Comments
The Mac Pro is the "Kwisatz Haderach" of all computers, and Apple is considering its relevance as a viable product. Come on Steve enforce your will in the after life! Make them not just continue it but make it better, faster, stronger!
Mac Daily has a poll - should the Mac Pro line be discontinued ? :
There are 9,000 votes - and the "Yes - discontinue" votes are winning 51% to 44% for keeping the line. Most of the voters probably own nothing but an iPod or an iPhone ....
Vote here:
http://macdailynews.com/
http://youtu.be/4PIajg_Htx0
OK so the HP design is very large and IP server centric but the principal is sound.
Its could be a bit like taking 8 macMinis on cards and integrating them in one Macpro X case. the result would be a local MacOS server to satify professional use that could also be a lone workstation or indeed serve several iMacs in a studio multiclient setup at a greatly reduced low energy cost alternative to using Intel Zeons.
just thinking out loud guys...
http://www.engadget.com/2011/11/02/h...l-the-boys-to/
Mac Daily has a poll - should the Mac Pro line be discontinued ? :
There are 9,000 votes - and the "Yes - discontinue" votes are winning 51% to 44% for keeping the line. Most of the voters probably own nothing but an iPod or an iPhone ....
Vote here:
http://macdailynews.com/
There is one problem with this poll. It really doesn't ask the question of what to replace it with. In my estimation the Pro is a dead platform in its current configuration. Apple still needs a High performance machine to serve the professional markets but they need to deliver that machine at the right price point. The platform needs to attract a wider array of users to maintain viability.
you know I've been thinking perhaps the conventional workstation PC is really old hat now no pun to linux intended but new Arm developements are beginning to impress and having said multinode transputational computing may be the way of the future
Just realize that most Pro users are more focused on high performance computing rather than server duties. A massive number of ARM processors actually makes good sense for many server workloads.
then I came accross HPs Moonshot server using quad card Arm system on a chip design and realised that would also make a really fast video render workstation very much in the fashion of tesla GPU processing.
http://youtu.be/4PIajg_Htx0
OK so the HP design is very large and IP server centric but the principal is sound.
Yes for a server it is a sound approach. For a Mac Pro it would need modification as Arm performance simply isn't good enough for the workloads a Mac Pro sees. The Mac Pro needs to be refactored, there is no doubt in my mind here. The problem is the modular approach can be expensive so Apple would have to over come that issue.
Its could be a bit like taking 8 macMinis on cards and integrating them in one Macpro X case. the result would be a local MacOS server to satify professional use that could also be a lone workstation or indeed serve several iMacs in a studio multiclient setup at a greatly reduced low energy cost alternative to using Intel Zeons.
just thinking out loud guys...
http://www.engadget.com/2011/11/02/h...l-the-boys-to/
8 cards implies a big heavy chassis that is also expensive thus one machine doesn't address the problems Apple is having on the desktop. In stead imagine a family of desktop machines that all take the same compute care. The Mini might take one card, the XMac two and the Pro eight. With one design effort and careful selection of clock rates Apple could cover most desktop needs.
The thing here isn't the specifics though it is rather the fact that Apple needs to innovate once again on the desktop. Traditional PCs may be seeing reduced demand, but that might be an indication of hardware not meeting user needs.
Obviously this HP approach is a bit to focused on server Workloads but many of the concepts reflect future capabilities.
Just realize that most Pro users are more focused on high performance computing rather than server duties. A massive number of ARM processors actually makes good sense for many server workloads.
Yes for a server it is a sound approach. For a Mac Pro it would need modification as Arm performance simply isn't good enough for the workloads a Mac Pro sees..
In the HPvideo http://youtu.be/4PIajg_Htx0 at imin24 the presenter shows a card about 1/2hight PCIexpress size containing 4Arm processors presumably quad core or at least by A15 it will be sometime next year I've read sugested, complete with independent memory modules, thats 16cores on one card and 8core processsors arn't too far off either, at 22nm thats a cool processing stack that Zeons will have trouble matching up to in future
The Mac Pro needs to be refactored, there is no doubt in my mind here. The problem is the modular approach can be expensive so Apple would have to over come that issue.
8 cards implies a big heavy chassis that is also expensive thus one machine doesn't address the problems Apple is having on the desktop.
In stead imagine a family of desktop machines that all take the same compute care. The Mini might take one card, the XMac two and the Pro eight. With one design effort and careful selection of clock rates Apple could cover most desktop needs.
It doesn't take such a large leap of imagination to consider a svelt MacPro concept casing to hold a 6or8way PCIe motherboard with a slim down the side 2" PSU ideal for studio rack mounting with 3 or 4 super silent thin fans. a much smaller and simpler unit than todays MacPro and you don't need to exaggerate dimensions to make it seem unreasonably large.
This could have say 3 Arm processor PCIe boards thats (48cores) as described above and a duel PCIeGPU card plus a couple of PCIeSSD cards and an interface card for specialist requirements, remember cost isn't the limiting factor on the companies flagship infact cheap might be sending the wrong message which all reflects on the whole Apple range.
Now consider the exact same footprint but Half hight chassis using the same motherboard PCIexpress design for a MacX semiPro workstation at more reasonable cost.
It doesn't have to be complicated any further, there are 2.5" and 3.5" HD PCIe hybrid cards aswell for the cash strained, and I'd imagine it would sporn many more card designs.
These workstations could be totom poled with TB through connections for the more extreme uses like calculating the latest Gnome or rendering the latest "Avater" movie,
The thing here isn't the specifics though it is rather the fact that Apple needs to innovate once again on the desktop. Traditional PCs may be seeing reduced demand, but that might be an indication of hardware not meeting user needs
I imagine eventually PCIe 4 will double the system speed again in due course but professionals will need this flexibility whether it be Mac or PC for quite some time to come.
And such flexibility would give Apple a platform to employ multiple architecture strategies into the future.
Simple is often the best way you know. http://forums.appleinsider.com/image...s/1smoking.gif
In the HPvideo http://youtu.be/4PIajg_Htx0 at imin24 the presenter shows a card about 1/2hight PCIexpress size containing 4Arm processors presumably quad core or at least by A15 it will be sometime next year I've read sugested, complete with independent memory modules, thats 16cores on one card and 8core processsors arn't too far off either, at 22nm thats a cool processing stack that Zeons will have trouble matching up to in future
Remember we are talking ARM cores here which often have a hard time keeping up with Intel hardware. Lots of cores work well in servers because you generate lots of threads and each processor can handle a single thread. desktop apps often rely more on single core performance.
It doesn't take such a large leap of imagination to consider a svelt MacPro concept casing to hold a 6or8way PCIe motherboard with a slim down the side 2" PSU ideal for studio rack mounting with 3 or 4 super silent thin fans. a much smaller and simpler unit than todays MacPro and you don't need to exaggerate dimensions to make it seem unreasonably large.
I'm all for smaller hardware. That being said I still think Apple needs a broader array of hardware.
This could have say 3 Arm processor PCIe boards thats (48cores) as described above and a duel PCIeGPU card plus a couple of PCIeSSD cards and an interface card for specialist requirements, remember cost isn't the limiting factor on the companies flagship infact cheap might be sending the wrong message which all reflects on the whole Apple range.
NO - expensive for no reason sends the wrong message as does forcing people to buy configurations they don't want. In fact Apple has been sending the wrong message for years now. Basically that message is this if you need slots (RAM or PCI) or drive bays, please bend over.
Now consider the exact same footprint but Half hight chassis using the same motherboard PCIexpress design for a MacX semiPro workstation at more reasonable cost.
It doesn't have to be complicated any further, there are 2.5" and 3.5" HD PCIe hybrid cards aswell for the cash strained, and I'd imagine it would sporn many more card designs.
These workstations could be totom poled with TB through connections for the more extreme uses like calculating the latest Gnome or rendering the latest "Avater" movie,
I imagine eventually PCIe 4 will double the system speed again in due course but professionals will need this flexibility whether it be Mac or PC for quite some time to come.
And such flexibility would give Apple a platform to employ multiple architecture strategies into the future.
Simple is often the best way you know. http://forums.appleinsider.com/image...s/1smoking.gif
Simple is all most always the best way. But what you describe isn't simple.
You mean almost like the 27" iMac?
Yes, but with a matte screen, or your own screen. I have a mini server in the office, if it's running something in the background I like being able to turn off the monitor and walk away, no worries about something moving my mouse and turning it back on. I like that fact it's matte too.
I thought about a Pro for expandability, but was worried about the noise level, are they quite noisy (compared to virtually silent mini?) I have never heard one running.
I thought about a Pro for expandability, but was worried about the noise level
The loudest part of a Mac Pro is the hard drive. It's not like the PowerMac G5, I'll tell you.
The loudest part of a Mac Pro is the hard drive. It's not like the PowerMac G5, I'll tell you.
I think the time is over for consumer towers but imagine a Mac Pro without any HDDs, but instead just a row of the SSD cards used in the MBAs. That would be both impressive in speed and price.
…just a row of the SSD cards used in the MBAs…
Using a nonsensically quick mental calculation, I can see Apple being able to fit 16 of those in a single row where the current hard drives are.
And assuming each is the 120GB SATA III stick from OWC, that's $4,479.84 for JUST those drives.
It's just shy of 2TB, but imagine RAIDing sixteen of those…
Heck, with their size, you could have TWO rows of 16 and do a RAID 0+1. That's like… 14 gigaBYTES per second read/write.
AllanMc; where are you from? I ask because your posts are hard to follow.
The Uk, ie where Jonathon Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
Remember we are talking ARM cores here which often have a hard time keeping up with Intel hardware. Lots of cores work well in servers because you generate lots of threads and each processor can handle a single thread. desktop apps often rely more on single core performance.
True... at the moment, but ARM A15 is not far around the corner and ProApps are using multiple threads now and will increase in the working future,
I'm all for smaller hardware. That being said I still think Apple needs a broader array of hardware.
Cost to develop is the problem, if one design concept can be flexible enough to encompass all Pro requirements then apple would have something to move forward with, at the moment Optical drives and Hard-drives are an outgoing technology, SATA is a slow interface for such outgoing drives but PCIe3 is a fast flexible architecture and also directly connects fast TB connectivity. why invent more when a suitable already supported structure already exists.
NO - expensive for no reason sends the wrong message as does forcing people to buy configurations they don't want. In fact Apple has been sending the wrong message for years now. Basically that message is this if you need slots (RAM or PCI) or drive bays, please bend over.
Its my turn not to understand your meaning here? PCs are ultimately flexible, you can build whatever you want, only the MacPro offers pro's the same relative flexibility on the MacOS platform, otherwise there's no competition and all Processionals will emigrate to the platform that provides what they need, and that is flexibility to spend whatever they want to spend on their high end business machine.
Simple is all most always the best way. But what you describe isn't simple
You can't get more simpler than a Case - PSU and a motherboard, and 3rd party PCI cards , the up and coming Window8 accommodates up to 160 cores compatible parallel processing, MacOS does multiple cores too and there is a good amount of highend workstation class software that uses them too.
The question is really - Do Apple want to competitively support a provisional market or not?
The Uk, ie where Steve Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
True, but there is a very real future for certain server farms using ARM processors. There is just too many avenues in which ARM can save money for an equivalent setup for certain usage types.
Replacing the Mac Pro with ARM smartphone class chips is the most nonsensical thing I have read in this thread.
What if they design a chipset that can have eight processors on a board?
And my emoticon for the above question is "
What if they design a chipset that can have eight processors on a board?
And my emoticon for the above question is "
Intel processors are greater than an order of magnitude more powerful, are we going to have 100 core Mac Pros?
There are diminishing returns for adding more cores in most desktop activities.
ARM processors use so much less power, because they are much less powerful.
There is no magic ARM pixie dust.
Intel processors are greater than an order of magnitude more powerful, are we going to have 100 core Mac Pros?
Well, since Intel currently has no plans for quantum computers and will reach the minimum physical size of transistor gateways sometime near the end of this decade, I'd say we'll have do something different.
I do see your point, though.
Intel's biggest fear is ARM in laptops and low-end desktops, though. Will Intel eventually be relegated to become the next IBM? Making only super-high-end chips and customized stuff?
Well, since Intel currently has no plans for quantum computers and will reach the minimum physical size of transistor gateways sometime near the end of this decade, I'd say we'll have do something different.
I do see your point, though.
Intel's biggest fear is ARM in laptops and low-end desktops, though. Will Intel eventually be relegated to become the next IBM? Making only super-high-end chips and customized stuff?
The thing is though, IBM (at least via their stock price) seems to be doing fairly well. The future will be very interesting.
The Uk, ie where Steve Ives and ARM come from, I was using StrongArm and RiscOS multitasking windows with multiple processors before Apple was weaned off of its 86 based games machines.
and we are talking about future MacPro hardware that will need to be current for the foreseeable future so one has to employ ones imagination to have a little inventive licence in ideas, yes!
You lost a bit of credibility there as Apple didn't even have a gaming play until the ARM hardware arrived.
True... at the moment, but ARM A15 is not far around the corner and ProApps are using multiple threads now and will increase in the working future,
I'm not convinced A15 will be all that attractive.
Cost to develop is the problem, if one design concept can be flexible enough to encompass all Pro requirements then apple would have something to move forward with, at the moment Optical drives and Hard-drives are an outgoing technology, SATA is a slow interface for such outgoing drives but PCIe3 is a fast flexible architecture and also directly connects fast TB connectivity. why invent more when a suitable already supported structure already exists.
The Mac Pro is a single platform that trys to appeal to everybody and has as a result been a failure. I don't think Apple has a chance in hell with a one size fits all machine. Modules usable in different machines is another discussion though.
Its my turn not to understand your meaning here? PCs are ultimately flexible, you can build whatever you want, only the MacPro offers pro's the same relative flexibility on the MacOS platform, otherwise there's no competition and all Processionals will emigrate to the platform that provides what they need, and that is flexibility to spend whatever they want to spend on their high end business machine.
You are wrong there, corporations don't give people free reign over what they can spend for computing hardware. When you have Apple hardware that is often three times the cost of PC hardware capable of doing the job you don't have a chance to even suggest Apple hardware.
The problem is Apple doesn't have the right box, expandable box, at the right price point.
You can't get more simpler than a Case - PSU and a motherboard, and 3rd party PCI cards , the up and coming Window8 accommodates up to 160 cores compatible parallel processing, MacOS does multiple cores too and there is a good amount of highend workstation class software that uses them too.
The problem is that is what the Mini is. The Mini is a fine machine as long as your needs are simple.
The question is really - Do Apple want to competitively support a provisional market or not?
Professional? Yes I want them to support professional markets but there is a whole array of professional markets that don't need nor can the justify the expense of a Mac Pro. Because the Pros market is so small Apple risks loosing the Pro market anyways due to the lack of sales. In a nut shell they have to offer a better platform to move forward. Technology can do a lot to make that better platform, but Apple seems to be unwilling to spend the time and effort on the desktop.
ARM processors use so much less power, because they are much less powerful.
There is no magic ARM pixie dust.
Are you assigning a 1:1 linear relationship between X86 and ARM?