Mac Pro Refesh in March

1679111219

Comments

  • Reply 161 of 374
    MarvinMarvin Posts: 15,309moderator
    Quote:
    Originally Posted by Lemon Bon Bon. View Post


    And if the iMac reaches the point of acceptable ram, gpu and cpu (as I feel the current top end is...ie as you say, you're no longer in physical pain due to the limitations of said machine) surely you plug into Raids/GPU stacks/Modular Render stacks? And/Or the 're-factored' Pro could.



    IMO, that's the ideal setup. Get a specialised render box like the Boxx Pro rendering machine but with GPUs and get the front-end to be as nicely designed and lean as possible.



    Quote:
    Originally Posted by Lemon Bon Bon. View Post


    Great link by the way, Marvin. Enjoyed reading about Nv and Weta working together to take Avator and move fx to the next level.) Yes. A fascinating read. Made me think of the implications, if any for APple's software and hardware.



    One of the phrases that stood out for me was "The complexity of Avatar motivated us to think about rendering differently". They didn't take a linear approach of adding more CPUs like everyone else, the limitations forced them to think outside the box - something Apple has promoted for as long as it has existed.



    Apple could give a serious boost to the high-end rendering scene if they just connect a few dots together. NexT used to ship Renderman in their OS to do 3D graphics. It is pretty much the benchmark for 3D post-production. What better way to supplement it than offer the same capabilities Weta describes.



    They have Thunderbolt to connect external boxes and they have OpenCL to run on them along with their Core software architectures. All they have to do is get Pixar to develop a piece of core raytracing tech e.g Core Illumination and allow scanline engines to access the pre-computed data.



    Motion, Final Cut, Maya, Lightwave, you name it, they just get the shadow, reflection, refraction, whatever data from the pre-compute step and do the final render on the CPU, which takes a fraction of the time.



    Quote:
    Originally Posted by Lemon Bon Bon. View Post


    The numbers you point to show quite an immense time saving going GPU over traditional cpu render farms by a factor of 6-7 to 1!!! That's pretty immense. Where will computer/Apple computers be in ten years time?



    GPUs are going up by a staggering amount. NVidia has a roadmap suggesting this year's Kepler GPUs will do 5 DP GFLOPs per watt and Maxwell (2013-2014) will be 15 DP GFLOPs per watt. This means that a 10W mobile Maxwell GPU in 2013-2014 will outperform the current high-end 6-core Xeon X5870. This year's mobile Kepler GPUs will be half of a Xeon chip.



    Current Xeons can do about 100GFLOPs in 100Watts i.e 1 DP GFLOP per watt. NVidia has a white paper noting that the Tesla C1060 is 8x faster than the Xeon W5590 (p15) with around the same power draw:



    http://www.nvidia.com/content/PDF/fe...chitecture.pdf



    The S1070 used on Avatar is a 1U rackmount server with 4x high-end GPUs connecting to a host via PCIe. This link could be replaced by Thunderbolt as it's just a compute engine (not real-time).



    We can see this dramatic leap of power year after year with devices like the iPad. You never hear that the CPU jumps up to 9x performance in a single year.
  • Reply 162 of 374
    wizard69wizard69 Posts: 13,377member
    Rather whatever Apple does in the future it has to appeal to the wider desktop audience. mac pro already has a significant sales disadvantage due to its limited appeal. Specializing a box even more will not help out Apple desktop sales.



    Quote:
    Originally Posted by Marvin View Post


    IMO, that's the ideal setup. Get a specialised render box like the Boxx Pro rendering machine but with GPUs and get the front-end to be as nicely designed and lean as possible.



    You are assuming that everybody with a desire for a desktop has a work load that fits into that sort of arrangement. I suspect that is a small number of pros using Apple hardware right now.

    Quote:

    One of the phrases that stood out for me was "The complexity of Avatar motivated us to think about rendering differently". They didn't take a linear approach of adding more CPUs like everyone else, the limitations forced them to think outside the box - something Apple has promoted for as long as it has existed.



    That is mostly marketing hype and self promotion. Many industries have been looking at and incorporating GPU processing into their work loads.

    Quote:

    Apple could give a serious boost to the high-end rendering scene if they just connect a few dots together. NexT used to ship Renderman in their OS to do 3D graphics. It is pretty much the benchmark for 3D post-production. What better way to supplement it than offer the same capabilities Weta describes.



    The volume isn't there. This is why a generalized approach of building desktops that support OpenCL is a priority at Apple. By doing so they can ship hardware to a wider audience that allows for reasonable pricing. If apple where to start building specialized GPU processing hardware the cost of this hardware would make the Mac Pros look cheap.



    Thankfully Apples approach is paying off and they had the influence to get Intel to build in OpenCL compatibility on their new hardware. So Apple will soon have coverage of OpenCL from the extreme low end to the high end. Notably at reasonable prices.

    Quote:

    They have Thunderbolt to connect external boxes and they have OpenCL to run on them along with their Core software architectures. All they have to do is get Pixar to develop a piece of core raytracing tech e.g Core Illumination and allow scanline engines to access the pre-computed data.



    Thunderbolt isn't exactly the best interface for this sort of thing, at least not in the large scale. What TB might do for Apple is to allow for small scale clusters that could potentially be very useful in large computational projects. Remember at this point if mac Pro is replace, the new machine will likely be close to twice as fast CPU wise and leery interesting GPU wise.

    Quote:

    Motion, Final Cut, Maya, Lightwave, you name it, they just get the shadow, reflection, refraction, whatever data from the pre-compute step and do the final render on the CPU, which takes a fraction of the time.



    Again this is all interesting but Apple can't be in the business of special purpose boxes. If they really want to offer up something like this it has to fall out of a more general purpose architecture that can ship in volume.

    Quote:

    GPUs are going up by a staggering amount. NVidia has a roadmap suggesting this year's Kepler GPUs will do 5 DP GFLOPs per watt and Maxwell (2013-2014) will be 15 DP GFLOPs per watt. This means that a 10W mobile Maxwell GPU in 2013-2014 will outperform the current high-end 6-core Xeon X5870. This year's mobile Kepler GPUs will be half of a Xeon chip.



    This isn't lost at Apple, in fact I'd say they have had an eye on this reality for some time. Think about it, OpenCL and some of the other Apple initiatives have been with us for awhile now. The pay off in good GPU performance is recognized and if some of the rumors about AMD's LLano in Apple hardware where true Apple has been struggling to get the optimal hardware in place. Good things don't happen overnight.

    Quote:

    Current Xeons can do about 100GFLOPs in 100Watts i.e 1 DP GFLOP per watt. NVidia has a white paper noting that the Tesla C1060 is 8x faster than the Xeon W5590 (p15) with around the same power draw:



    Yeah more specs. You have to remember the devil is in the details. GPU's currently are very good at certain types of floating point processing but they aren't the solution to ever problem. So while amazing GFLOP numbers can be thrown around by NVidia and others just realize that GPU's need the right sorts of data to work on to get those numbers.

    Quote:

    http://www.nvidia.com/content/PDF/fe...chitecture.pdf



    The S1070 used on Avatar is a 1U rackmount server with 4x high-end GPUs connecting to a host via PCIe. This link could be replaced by Thunderbolt as it's just a compute engine (not real-time).



    With the right hardware, a Mac Pro, you could just plug in a compute card and bypass the slow TB link. Even better would be a future where every machine comes with a GPU with the resources to do compute.

    Quote:

    We can see this dramatic leap of power year after year with devices like the iPad. You never hear that the CPU jumps up to 9x performance in a single year.



    Yes but you have to understand they whys here. The things that a GPU does in a system are highly parallel. That means one gets almost 100% out of every processor added. You can't get that sort of advantage from most CPU applications no matter what you do. The fact that GPU hardware can be applied to a narrow range of other problems is sort of icing on the cake.
  • Reply 163 of 374
    mcarlingmcarling Posts: 1,106member
    Quote:
    Originally Posted by wizard69 View Post


    Apple can't really build or define hardware in terms of what the movie industry needs. Rather whatever Apple does in the future it has to appeal to the wider desktop audience. mac pro already has a significant sales disadvantage due to its limited appeal. Specializing a box even more will not help out Apple desktop sales.



    It looks like you may have missed the point. Apple do not need to produce specialized hardware. All Apple need to do beyond the current Thunderbolt and OpenCL support is provide some addition software, as Marvin has specified above. Third-party hardware manufacturers would then produce the Thunderbolt accessory hardware desired by the market. Once the solution is in place for rendering, third-party software developers would exploit it for other applications. The result is that an ordinary unspecialized Mac Mini could be used (with the addition of third-party Thunderbolt hardware) to do computations which currently require a Mac Pro.
  • Reply 164 of 374
    wizard69wizard69 Posts: 13,377member
    Quote:
    Originally Posted by mcarling View Post


    It looks like you may have missed the point. Apple do not need to produce specialized hardware. All Apple need to do beyond the current Thunderbolt and OpenCL support is provide some addition software, as Marvin has specified above. Third-party hardware manufacturers would then produce the Thunderbolt accessory hardware desired by the market.



    My point is that Apple nor a third party can sustain on such a product! Or to phrase it as a question how can such a product compete against a PC outfitted with a couple of compute cards. Mind you a company building the specialized hardware would need to be able to finance on going operations from said hardware. At best a company might have a chance if the hardware is general purpose enough to attract a wider audience.

    Quote:

    Once the solution is in place for rendering, third-party software developers would exploit it for other applications. The result is that an ordinary unspecialized Mac Mini could be used (with the addition of third-party Thunderbolt hardware) to do computations which currently require a Mac Pro.



    I can't ever see this being a success. You already have companies making compute optimized servers, some with GPU acceleration, in rack type clusters. An alternative as described would need to be cheaper and high performance, something I don't see happening. Even for a single box implementation I don't see a big advantage over a Pro with GPU acceleration cards. Maybe I'm missing something but I just don't see the economics.
  • Reply 165 of 374
    hmmhmm Posts: 3,405member
    Quote:
    Originally Posted by Marvin View Post


    Apple could give a serious boost to the high-end rendering scene if they just connect a few dots together. NexT used to ship Renderman in their OS to do 3D graphics. It is pretty much the benchmark for 3D post-production. What better way to supplement it than offer the same capabilities Weta describes.



    They have Thunderbolt to connect external boxes and they have OpenCL to run on them along with their Core software architectures. All they have to do is get Pixar to develop a piece of core raytracing tech e.g Core Illumination and allow scanline engines to access the pre-computed data.



    Motion, Final Cut, Maya, Lightwave, you name it, they just get the shadow, reflection, refraction, whatever data from the pre-compute step and do the final render on the CPU, which takes a fraction of the time.







    GPUs are going up by a staggering amount. NVidia has a roadmap suggesting this year's Kepler GPUs will do 5 DP GFLOPs per watt and Maxwell (2013-2014) will be 15 DP GFLOPs per watt. This means that a 10W mobile Maxwell GPU in 2013-2014 will outperform the current high-end 6-core Xeon X5870. This year's mobile Kepler GPUs will be half of a Xeon chip.



    Current Xeons can do about 100GFLOPs in 100Watts i.e 1 DP GFLOP per watt. NVidia has a white paper noting that the Tesla C1060 is 8x faster than the Xeon W5590 (p15) with around the same power draw:



    http://www.nvidia.com/content/PDF/fe...chitecture.pdf



    The S1070 used on Avatar is a 1U rackmount server with 4x high-end GPUs connecting to a host via PCIe. This link could be replaced by Thunderbolt as it's just a compute engine (not real-time).




    Edit: You aren't suggesting that they rendered Avatar on a single 1U server are you?



    Avatar had a ton of cpus there too. Some of what you're mentioning requires clarification, and editing down of the NVidia marketing kool aid that they licensed from Apple (okay I couldn't resist suggesting that, the concept makes me laugh).



    I don't know where gpu based rendering and number crunching will be in a couple years. It's still in its infancy. I can't recall if Renderman supports it at all. Mental Images incorporated some gpu functions into Iray with select gpus (although I haven't used iray). VRay RT incorporates it as well for draft renders. That's used more in visualization and advertising (as is Mental Ray) with film seemingly split between Mental Ray and Renderman. Not every studio uses Renderman or uses it exclusively. Maxwell Render also implemented a real time gpu rendering preview in some of their stuff, and I think Modo grabbed a couple gpu features. Anyway a lot of this stuff started to pop up in 2010 or so. It's still in its infancy today, and there are a number of limitations into what the gpu can run. I haven't found enough information on it to understand the progression entirely, but regarding Apple, they've never supported that market very well.



    My real issue with Apple isn't that the only solution must be the mac pro. It's that some of the heavier hardware options aren't fully supported under other solutions, and the quality control in the imacs is atrocious (especially displays). Apple seems to go cheap on things that the general consumer doesn't understand. They figure out what people will tolerate.



    I don't think they'll see many gains here. If anything Linux is a better point of migration for that industry. As of right now half of the plugins and scripts don't run on OSX versions of stuff compared to the Windows side, and many studio pipelines aren't built to support macs.





    Quote:
    Originally Posted by mcarling View Post


    It looks like you may have missed the point. Apple do not need to produce specialized hardware. All Apple need to do beyond the current Thunderbolt and OpenCL support is provide some addition software, as Marvin has specified above. Third-party hardware manufacturers would then produce the Thunderbolt accessory hardware desired by the market. Once the solution is in place for rendering, third-party software developers would exploit it for other applications. The result is that an ordinary unspecialized Mac Mini could be used (with the addition of third-party Thunderbolt hardware) to do computations which currently require a Mac Pro.



    Ugh... you really need to look at your details. GPUs can receive data much faster than they would over thunderbolt currently, and i'm not sure what protocols thunderbolt supports anyway. Apple likes to tell us that it supports everything, but there are already a bunch of fringe examples where it simply doesn't work. They're lying or leaving out information. As to Marvin's suggestion, not even a proof of concept exists at this point, much less a functioning rig. You're basically turning the mini into a slim client, which is pointless. I don't know how to hammer this in any better. The point of a workstation is for work that can't be centralized onto a server. You simply don't have that bandwidth with thunderbolt, and thunderbolt in itself draws it bandwidth specifically from available PCI lanes meaning that you'll never exceed that limit on a given machine. Further no thunderbolt standard that matches the current transfer rate an internal gpu can receive is coming up in the near future, and when it does, there's no guarantee that such bandwidth would run off the board type used by the mini. It's more likely that it would be throttled to a slim chip like the macbook air.



    Quote:
    Originally Posted by wizard69 View Post


    You are assuming that everybody with a desire for a desktop has a work load that fits into that sort of arrangement. I suspect that is a small number of pros using Apple hardware right now.



    That is mostly marketing hype and self promotion. Many industries have been looking at and incorporating GPU processing into their work loads.



    The volume isn't there. This is why a generalized approach of building desktops that support OpenCL is a priority at Apple. By doing so they can ship hardware to a wider audience that allows for reasonable pricing. If apple where to start building specialized GPU processing hardware the cost of this hardware would make the Mac Pros look cheap.



    Thankfully Apples approach is paying off and they had the influence to get Intel to build in OpenCL compatibility on their new hardware. So Apple will soon have coverage of OpenCL from the extreme low end to the high end. Notably at reasonable prices.



    Thunderbolt isn't exactly the best interface for this sort of thing, at least not in the large scale. What TB might do for Apple is to allow for small scale clusters that could potentially be very useful in large computational projects. Remember at this point if mac Pro is replace, the new machine will likely be close to twice as fast CPU wise and leery interesting GPU wise.






    Bleh the point being missed by him was largely that server potential was there for non localized functions already. Thunderbolt doesn't really change this, and the bandwidth isn't high enough to approximate that of a local connection type anyway. I don't see where the misunderstanding lies between workstations and servers here, but it's definitely there. People only use workstations for long server like number crunching when it benefits them. This means that if they own one box already, it can use either unused cpu cycles to crunch numbers in the background or be allocated to this purpose when not in real use. This is a completely separate thing from functions which must be run in a localized manner due to IO bandwidth and protocol constraints. If it wouldn't have been feasible in the past to offload something via fibre channel, what makes thunderbolt a game changer there? I'm not sure where the misunderstanding lies, but putting together a clunky solution isn't going to attract more power users to the platform and promote sales growth of anything. It doesn't even make remote sense. You can do the same thing with a typical i7 PC desktop and simply have a moderately powerful desktop gpu there. Why downgrade to a mini for the same money, and if you're powering a stronger gpu elsewhere, your power savings on desktops won't be much considering the heavier external gpu.



    Also in his reference to Boxx, they sell a lot of overclocked single socket workstations. They don't run with minis or laptop like devices. I think they have a slim client or two. Anyway the point was that externalized hardware only works for part of the process, and Apple isn't even a huge contender there. In desktops the mac pro has remained in use simply because it lacks some of the issues of the others, especially in integrating with existing hardware. Apple goes really cheap on ports, and thunderbolt is still lacking much development a year later. I just don't see the case to chop everything down to a mini when there are better solutions even to the problems suggested by Marvin which aren't commonly0 run on Macs anyway.
  • Reply 166 of 374
    mcarlingmcarling Posts: 1,106member
    Quote:
    Originally Posted by wizard69 View Post


    My point is that Apple nor a third party can sustain on such a product! Or to phrase it as a question how can such a product compete against a PC outfitted with a couple of compute cards. Mind you a company building the specialized hardware would need to be able to finance on going operations from said hardware. At best a company might have a chance if the hardware is general purpose enough to attract a wider audience.





    I can't ever see this being a success. You already have companies making compute optimized servers, some with GPU acceleration, in rack type clusters. An alternative as described would need to be cheaper and high performance, something I don't see happening. Even for a single box implementation I don't see a big advantage over a Pro with GPU acceleration cards. Maybe I'm missing something but I just don't see the economics.



    What you seem to be missing is software. People don't buy a Mac Pro rather than a PeeCee just because they like Apple hardware. People buy Macs because they want OSX and the software solutions which it provides. If the software is sufficiently better, the hardware doesn't need to be the cheapest.



    Quote:
    Originally Posted by hmm View Post


    Ugh... you really need to look at your details. GPUs can receive data much faster than they would over thunderbolt currently, and i'm not sure what protocols thunderbolt supports anyway. Apple likes to tell us that it supports everything, but there are already a bunch of fringe examples where it simply doesn't work. They're lying or leaving out information. As to Marvin's suggestion, not even a proof of concept exists at this point, much less a functioning rig. You're basically turning the mini into a slim client, which is pointless. I don't know how to hammer this in any better. The point of a workstation is for work that can't be centralized onto a server. You simply don't have that bandwidth with thunderbolt, and thunderbolt in itself draws it bandwidth specifically from available PCI lanes meaning that you'll never exceed that limit on a given machine. Further no thunderbolt standard that matches the current transfer rate an internal gpu can receive is coming up in the near future, and when it does, there's no guarantee that such bandwidth would run off the board type used by the mini. It's more likely that it would be throttled to a slim chip like the macbook air.







    Bleh the point being missed by him was largely that server potential was there for non localized functions already. Thunderbolt doesn't really change this, and the bandwidth isn't high enough to approximate that of a local connection type anyway. I don't see where the misunderstanding lies between workstations and servers here, but it's definitely there. People only use workstations for long server like number crunching when it benefits them. This means that if they own one box already, it can use either unused cpu cycles to crunch numbers in the background or be allocated to this purpose when not in real use. This is a completely separate thing from functions which must be run in a localized manner due to IO bandwidth and protocol constraints. If it wouldn't have been feasible in the past to offload something via fibre channel, what makes thunderbolt a game changer there? I'm not sure where the misunderstanding lies, but putting together a clunky solution isn't going to attract more power users to the platform and promote sales growth of anything. It doesn't even make remote sense. You can do the same thing with a typical i7 PC desktop and simply have a moderately powerful desktop gpu there. Why downgrade to a mini for the same money, and if you're powering a stronger gpu elsewhere, your power savings on desktops won't be much considering the heavier external gpu.



    There are some computational applications for which Thunderbolt has ample bandwidth and the tighter integration of a Thunderbolt connection may offer advantages over an Ethernet connection. It doesn't have to be the best solution for all customers in order to be the best solution for some customers.



    Again, I'm not suggesting that Apple make any specialized hardware. I believe it's enough for Apple to provide some software. If there is sufficient market demand, someone will build Thunderbolt GPU compute engines. If demand, is insufficient, then they won't be built.
  • Reply 167 of 374
    mccrabmccrab Posts: 201member
    Quote:
    Originally Posted by Lemon Bon Bon. View Post


    While I'd like a bigger 'mini'...(I'd buy a Cube tomorrow...) I don't see Apple doing that. (I'm not sure what it can give me that an iMac can't.) Otherwise it wouldn't be a mini. The iMac isn't going anywhere...it firmly occupies the 'mid-tower' ground from £995 to £2k. So a mini-Pro? That would be overlap. Hey, they USED to sell the entry tower at £1495 (something like that...) but now they don't. Again. I don't think that's an accident.



    Are they really going to develop a new desktop when they just canned their cheaper laptop entry model, the 'Macbook?' (A good servant for the laptop sales...but they canned it all the same.) Looks like the iPad will take care of the 'netbook'/cheap laptop area.



    If they redesign the Pro...to a mini-Pro...where's it going to sit, price wise?





    I'd gladly pay up for a cube that DOESN'T max out at 16GB RAM (4-64GB please), single processor (dual processor option please), average graphics ('nough said) and difficult to replace HDD/SSDD (3/4 bay with hardware RAID please). Price is less relevant. Please give us some more headroom Apple! A Mini-Pro (in cube form) would be awesome.
  • Reply 168 of 374
    macroninmacronin Posts: 1,174member
    Inserting Renderman into OS X as a Core component would be pretty sweet…



    And the simple fact of it being a bundle-in with the older NeXT OS, which became Mac OS X, which became OS X; could be a sort of leverage…



    Petition to the Mouse…?!? ;^p
  • Reply 169 of 374
    hmmhmm Posts: 3,405member
    Quote:
    Originally Posted by MacRonin View Post


    Inserting Renderman into OS X as a Core component would be pretty sweet?



    And the simple fact of it being a bundle-in with the older NeXT OS, which became Mac OS X, which became OS X; could be a sort of leverage?



    Petition to the Mouse??!? ;^p



    You know it's something like $2000/seat right? Yes... bundled into OSX.
  • Reply 170 of 374
    wizard69wizard69 Posts: 13,377member
    Quote:
    Originally Posted by mcarling View Post


    What you seem to be missing is software. People don't buy a Mac Pro rather than a PeeCee just because they like Apple hardware. People buy Macs because they want OSX and the software solutions which it provides. If the software is sufficiently better, the hardware doesn't need to be the cheapest.



    My point is that systems like this, that is computational clusters, always have a cost consideration. That is why so many are built out of generic hardware and Linux.



    Now could Apple create a captive market around niche hardware and software? It is certainly possible but unlikely to be profitable unless they can successfully charge a great deal for said hardware. I don't see it happening.



    Now ohers have suggested leading with software attached to the hardware to I guess wedge themselves into the high performance computing market. That is start with video and make sure the tools are there to support more general needs. This isn't impossible but this market is already well served by others and it is still a niche market. Apple can play in this market much easier by simply equipping their machines with OpecCL hardware and by having a slot to support another GPU card. I just don't see the feasibility or attraction in a headless GPU computation box.

    Quote:



    There are some computational applications for which Thunderbolt has ample bandwidth and the tighter integration of a Thunderbolt connection may offer advantages over an Ethernet connection. It doesn't have to be the best solution for all customers in order to be the best solution for some customers.



    There well may be in the future TB advantages and for small systems that may be now. The problem is that Ethernet and other standards have a tremendous amount of read to go infrastructure that can scale. Depending upon the tech you might have hundreds of nodes on an Erhernet based system. It isn't clear to me that TB can support such systems. In fact I'm pretty sure it can't.



    Now TB might effectively interface you to a couple of boxes or a small cluster if the "hub" MAc had several TB ports. That has significant potential, however you wouldn't be doing Avatar on it.

    Quote:

    Again, I'm not suggesting that Apple make any specialized hardware. I believe it's enough for Apple to provide some software. If there is sufficient market demand, someone will build Thunderbolt GPU compute engines. If demand, is insufficient, then they won't be built.



    Apple already provides much of the software required. The problem is nobody builds clusters or compute engines around Apple hardware and software. At least not recently. Even if somebody where to go after this market Agressively there would be a huge cost advantage to simply supporting a GPU compute card on the PCI Express bus.
  • Reply 171 of 374
    wizard69wizard69 Posts: 13,377member
    Your post boils down what is needed in a desktop to a few essentials. This is really key to a viable desktop from Apple.



    Quote:
    Originally Posted by McCrab View Post


    I'd gladly pay up for a cube that DOESN'T max out at 16GB RAM (4-64GB please),



    I don't currently have need of RAM past 16GB but I'm not at all pleased with artificial restrictions here. So support for more slots would be very nice.

    Quote:

    single processor (dual processor option please),



    This I debate with myself a bit. I see dual socket up port as a bad idea. It adds size and expense that we are trying to get away from.

    Quote:

    average graphics ('nough said)



    Average means different things to different people. Ideally the machine would support a descrete GPU working in conjunction with an integrated GPU. At least on a low end model that uses an Ivy Bridge type chip.



    Why? Well Intel GPUs really aren't that great and probably never will be. However the combination of a integrated and nielsen of the road descrete GPU could lead to very nice performance and be easily leveraged for compute usage. I just see a modest amount of hardware doing very well for the price. Especially if the GPU is AMDs or NVidias latest architectures which they are now debuting in lower performance versions.

    Quote:

    and difficult to replace HDD/SSDD (3/4 bay with hardware RAID please).



    Difficult to replace? No way, sadly apple does better with their laptops these days. Apple needs to pull its head out of you know where and actually make a serviceable desktop machine.



    As to disk drives hardware RAID really isn't needed anymore. With all the cores in today's processors software RAID is fine. While drive bays are still needed I want to see a move to a high speed PCI Express port for solid state storage cards. Today that would be a boot and app drive but the pay off is huge.

    Quote:

    Price is less relevant. Please give us some more headroom Apple! A Mini-Pro (in cube form) would be awesome.



    Yeah something like that. It would likely be a bit bugger and it would have to support a slot in at least one form. The important thing is that it must depart from legacy devices and be as minimalist as possible. Price is important as we don't want a failure on our hands like the Mac Pro.
  • Reply 172 of 374
    Quote:
    Originally Posted by wizard69 View Post


    At least people are recognizing the Pros very old design these days. Take a ten year old laptop and compare it to today's, things have changed a lot. Not so much with the Mac Pro where each revision just freshens the architecture while not innovating.



    Good question which one needs to have confidence in guesses about current Mac Pro sales. I could see sales of less than 10k a month. Obviously internal information would be required to say for sure, and frankly the upper tier models are the only ones some would consider. So sales actually could be biased towards the high end.



    Please get a grip here the MacBook was replaced by something that better serves the same market. Every time I see comments like this I shake my head, it was pretty clear where the Mac Book was going the first time Apple debuted the refactored AIRs at a more reasonable price.



    This constant falling back to performance is a waste of time, because for every case that one can find for an iMac smacking a Pro around a like number can be found where the Pro smacks the iMac. Notably at lot harder than the iMac can manage to smack the Pro.



    More importantly I don't really see a lot of Pro sales going to people simply focusing on performance, at least not CPU performance. They are instead buying for the other capacities that the machine offers.





    It all depends upon how you look at the numbers. Apples Mac sales are cooling off and the desktop market for Apple in the US is absolutely flat. Also look at the stores and what do you see most directly - the low cost products. It isn't so much that devices are selling well because of portability but rather because of cost.



    Yes but look back to all of through all of those years and realize that Apple did have reasonable desktop offerings. Even the Mini was a more reasonable offering. Not that the Mini is extremely bad just that one could go to Apple and get a complete system at a reasonable cost. Today theMini is more expensive and the monitor solution is terrible. It is no surprise then that the Mini is the sales darling of the online industry where it is packaged with reasonable screen solutions.



    PPC became a joke performance wise, so they got beaten up simply due to real world performance. It is the pricing of the low end that really bothers me. Apple simply doesn't have anything remotely competitive Asa midrange desktop machine.



    How in the hell can you make any of those statements???. There is nothing about an XMac that would be unprofitable in the least. The Pro is likely the least profitable machine in Apples line up. Further a properly built XMac would be more of a successor to the Apple 2.



    Exact specs - how about using your imagination? There are many ways for Apple to realize an XMac. What you need to see though is that this in part would be the total overhaul of the Mac lineup. Gone would be the current Pro at the high end and the Mini at the low end. In stead we would have a scalable platform that could cover the entire performance spectrum.



    Beyond that Macs don't have anymore momentum portability wise. The surge or better yet the ramp will be biased towards iOS devices more and more every quarter. If momentum was there Apple wouldn't be sending me ads for AIRs and other Mac hardware.



    A model range that replaces both the Pro and Mini and takes a few iMac sales along with it could easily sell 2 million or more a quarter.



    Why would you bring up the Cube which back then was a terrible value relative to anything else just like most of today's Mac Pros? If anything Apple learned nothing from the Cube as they repeated that insanity with the original AIRs. At least with the AIRs they pulled their collective heads out of a dark place and produced something the market could eventually accept.



    tides go in and out, currently the tide is receding.

    Let me count the ways.
    1. A much larger screen that fits my needs.

    2. The ability to actually add disk storage to the base configuration.

    3. Serviceability.

    4. Speed, a desktop can easily exceed the performance of any portable.

    5. Expandability or configurability. And no I'm not talking GPUs here.

    6. A desktop actually reduces clutter. Throw a laptop on a desk and you have to connect a bunch of wires to it.

    7. I need someplace to store everything and to backup all of those portable devices.

    This makes no sense at all. Average consumers shop and thus are more sensitive to value than your so called die hards. They can perceive a bargain and a screwing.



    Do you really understand business at all? I can't tell you how many time I've been pulled off primary tasks to solve the problem of the moment. We aren't even talking consummer electronics here.



    Apples condition is no more different than John Dears, they sell big and little tractors and everything in between. They do not give up on one segment just because another is hot.



    Just because you can constantly repeat something does not make it true. IMac can't possibly smack the Pro unless you are very narrowly focused on optimal benchmarks. Your constant obsession here tells me you have no idea or appreciation of the Pros advantages.



    I'm with you on pricing though. The whole discussion about XMac revolves around the idea of a far more cost effective desktop line up.



    It would be very bad voodoo for Apple to abandon the desktop.



    If this was true I'd almost be happy. The fact is iMac has soaked up nothing. Apple is in fact loosing sales. People that want a desktop are simply going elsewhere.



    The iMac is not a replacement for either the Mini nor the Pro, it is as simple as that.



    Considering that the iMac is in a category of its own and is the only machine with some success there is no reason to can it. The Mac Pro and Mini on the other hand could die tomorrow and no one would bat an eye.



    Again you have a very narrow view of reality. There is still a strong demand for performance that can't be had in a portable. That is the driver for desktop machines. Apple doesn't need a Tower in the sense of the Mac Pro to deliver such performance any more, that we can agree upon. However neither the iMac nor the Mini can deliver such performance. The obvious response is a midrange capable machine.



    Interestingly you are half way there to a XMac. Just throw in modern technology and you would be all set. In fact you seem to have pretty much verified in my mind why an XMac is needed. You dismiss XMac one moment but then turn around and realize that some of the points offered up are important and suggest what could be considered to be an XMac.



    The more thought that people put into this the more obvious it should become that Apple needs to do something about the Mini and the Pro. XMac is simply a concept that they can build a new family around.



    2 million desktop sales? Fantasy. Turn the tide from 4:1 to 2:1? Sure. And Apple won't charge a premium on the said line like they currently do with everything you want in it?!? You think the extra processor will add cost? Ya think?



    Unlikely. *Passes a 'reality' sandwich for Wizard to get a 'grip' on.



    Build their desktop around your 'X-Mac.' A complete 180 in design direction? Very unlikely.



    *Apple don't want you to service your machine.



    *Apple don't want to sell dirt cheap desktop machines.



    *Apple don't want them too expandable.



    *The market is buying laptops. (They sit onto desks much easier! And you can even carry them!

    They have most of the desktop's power these days! Want a bigger screen? Buy one. Apple will sell you one too!)



    *That's the 'desktop' they're selling most of these days. It's not your desktop. But Apple doesn't care because they're selling 5 million Macs a quarter. ie Record numbers. ie without this fabled X-Mac.



    This is how they make their money. Buy one every 3 years and buy another.



    So keep dreaming for those features. That's where you'll find them.



    www.apple.com Check out the store. That's what they're selling. There is no X-Mac.



    You'll *maybe' get a choice of gpu and HD at point of sale. That's your expansion internally.



    If you want to add some more? You can just about do a ram module yourself. That's about as much as they want a user to do.



    I guess you can add an external hard drive. There's Firewire, USB and Thunderbolt.



    Ironic that you're clinging onto a Macbook Pro. Keeping clinging to your 'desktop'. Because you'll be waiting for the X-Mac far longer than I did...



    No harm in wishing for this 'X-Mac.' You'll be waiting in hope.



    The 'desktop' is changing. Just not in the direction you want.



    'It doesn't matter how many times you say it...it doesn't make it...' happen.



    As for performance? Why do you want this X-Mac then? Don't cite performance of a desktop over a laptop and then say it isn't important. Apple will sell you decent performance at a premium price. They don't offer bleeding edge performance machines. They offer compact design. Elegance. Zen. But you'll pay for it. 'It doesn't matter how many times...etc.' you rail against it Wizard. That's what they're offering.



    Try their feedback page and see how long it takes for the X-Mac to hit the stores.



    Get the Ivy Bridge Mac Mini when it lands and be happy. What kind of work are you doing anyhow?



    Lemon Bon Bon.
  • Reply 173 of 374
    Quote:

    Beyond that Macs don't have anymore momentum portability wise. The surge or better yet the ramp will be biased towards iOS devices more and more every quarter. If momentum was there Apple wouldn't be sending me ads for AIRs and other Mac hardware.



    Sure. That's why they're selling more Mac laptops than ever before in RECORD NUMBERS. *(Dragged by the halo of iOS and Apple Store.)



    Lemon Bon Bon.
  • Reply 174 of 374
    Quote:

    The fact is iMac has soaked up nothing. Apple is in fact loosing sales. People that want a desktop are simply going elsewhere.



    Back that up with some facts. That's pure conjecture because you don't like the iMac. An absurd statement. Apple is losing sales? From who? They're selling Macs in record numbers. (Oh. I guess, they're losing a sale...uhm...to you..?)



    Yes. The 'people' who used to want a desktop are going elsewhere. Laptops.



    Uhm...if Apple were to drop the prices on their overpriced desktops they'd sell more? Well, yeah...



    Apple used to sell about 1 million Macs a quarter once upon a time. It used to be in favour of desktops. But that changed over time to laptops. That's because laptops over enough power for 9/10 uses for most people.



    Once the affordable iMac came on the scene it impacted PowerMac sales. As the iMac grew in power and occupied the mid-tower range and the Mac Pro moved on up...iMac sales impacted it more. There's a reason why Apple don't do sales breakouts anymore. Because certain lines would look weak. In this case, I'm guessing the Pro would.



    The World Economy is in the crapper. Newsflash. People want value. As far as desktops are concerned? That's the iMac. For the desktop, it's the Air.



    Lemon Bon Bon.
  • Reply 175 of 374
    Apple just can't offer a screen less desktop at a reasonable price. Look at the inflated price of the mini. No k/b. Mouse or screen. £529 to get started. Add in the 27 inch Apple monitor and you get that squeezy feeling in your wallet.



    It's *almost* always been this way. *(Remembers the Candy iMac getting down to £545 and the PPC Mac Mini at sub £400 notable examples of a cheap desktops Apple offered a long time ago.) Their all in ones whether desktop or laptop offer 'better' value...relative to the Pro or Mini...and if you're happy with Apple's premium pricing to start with.)



    Lemon Bon Bon.
  • Reply 176 of 374
    Quote:

    Again you have a very narrow view of reality. There is still a strong demand for performance that can't be had in a portable. That is the driver for desktop machines. Apple doesn't need a Tower in the sense of the Mac Pro to deliver such performance any more, that we can agree upon. However neither the iMac nor the Mini can deliver such performance. The obvious response is a midrange capable machine.



    www.apple.com. That's Apple's reality. And their 5 million in sales of Macs in record numbers make your 'narrow' view of reality precisely that. A minority as far as Apple customers are concerned.



    Apple are a portable computing company. All their promo shots show laptops even with Pro(sumer) software like Final Cut. 4 Million laptops+.



    Apple will release the iPad tomorrow. Zillions of ipads sold already.



    That's Apple's reality. Not so narrow.



    Lemon Bon Bon.
  • Reply 177 of 374
    Quote:

    I need someplace to store everything and to backup all of those portable devices.



    Try buying an external hard drive.



    Lemon Bon Bon.
  • Reply 178 of 374
    Quote:

    Apples Mac sales are cooling off



    Ridiculous. They've out grown the PC market for how many quarters now?



    Lemon Bon Bon.
  • Reply 179 of 374
    hmmhmm Posts: 3,405member
    Quote:
    Originally Posted by Lemon Bon Bon. View Post


    Ridiculous. They've out grown the PC market for how many quarters now?



    Lemon Bon Bon.



    That's the one are where I think you need a little perspective. Their numbers have been good there, but the Apple fan sites like to spin them in whatever way sounds best. I doubt they are picking up much in imac growth, and studies seem to include the ipad in laptops whenever it specifically suits them (again it would be more appropriate if it was a functional standalone machine).
  • Reply 180 of 374
    tallest skiltallest skil Posts: 43,388member
    Quote:
    Originally Posted by hmm View Post


    That's the one are where I think you need a little perspective. Their numbers have been good there, but the Apple fan sites like to spin them in whatever way sounds best. I doubt they are picking up much in imac growth, and studies seem to include the ipad in laptops whenever it specifically suits them (again it would be more appropriate if it was a functional standalone machine).



    So you're saying Apple is lying about selling more Macs every quarter than previous?
Sign In or Register to comment.