Scientific compute servers (from Sun, HP and Dell) have supported 256 GB and more for over 10 years (albeit at a hefty price). In 2013, a fully redesigned Mac Pro handles only 64 GB--just double that of a 2006 Mac Pro. For professionals who don't need graphics computation, it would have been far better to include a second processor to reach 24 cores (48 hyperthreads) and more DIMM slots, while ditching the superfluous second graphics card. As indicated above, I for one am very disappointed in the new design. As for FCPX users, have fun. As for everyone else, meh.
Your emphasis of "professionals" seems to suggest that only people who don't need graphics computation are professionals, and then your final point again dismisses some of Apple's big users.
Personally, I don't like the plastic iPhone (and I absolutely cannot stand iOS7). That's life. I don't have to buy it. But, it is selling, which is good for Apple, which is their business. They are not in business to create a custom, made-to-order phone just for me. I can guarantee it would not sell well! The new MP may not be for for you. That's life. But if your comments about RAM are correct, then the MP has never been the machine for you. That doesn't make it a bad machine.
But, regarding those GPUs, they run OpenCL. Apple has already reworked (yes, reworked) Final Cut to utilize the GPUs. I know of at least two other developers (one has publicly committed to doing so) who are reworking their apps to use the GPUs to milk the machines for all they've got. Would it be possible for the "professionals" in your line of work to look into the possibility of reworking apps so that they could use the GPUs and the MP would be a very useful machine for you?
I confess to not being a techie and I have no idea what OpenCl really is or what it can and cannot do.
3) Same as with point 2, I think it's a poor argument to claim the 2nd GPU is superfluous. Why would any company do that? They have a deal with AMD? Tim Cook is trying to Brewster's Millions Apple? The new Mac Pro might ultimately end up being a bad move for Apple (but based on how quickly the time to delivery was pushed back that seems doubtful) but it's clear they researched this in ways we can't imagine so I don't know what additional evidence we would have that would show it need be de duel socket and only have 1 GPU for it to be a success.
I missed the nerdy references during your absence.
1) I don't understand why the Mac Pro is all of a sudden being negatively compared against specialized computers that need a lot more RAM than its ever had. Why wasn't this "it sucks because it does't have 32 DIMM slots on 8 risers"-style argument not ever brought up before? Why now is there such a comparison when it's never been designed for the specialized tasks it's being compared to. It's like looking at a brand new Rolls Royce Phantom and complaining that it's a horrible cargo plane.
OMG Why the false pretenses? Who is complaining that the new Mac Pro doesn't have 32 DIMM slots?????
I for one am complaining that the new Mac Pro lost an entire CPU (up to 12 cores and 24 hyperthreads) and lost 4 DIMM slots.
It also lost the option of having just one graphics processing card and lost virtually all internal expansion.
IMHO this is not progress.
I am certain the new design fits well enough with the needs of video professionals, particularly since FCPX takes good advantage of the graphics cards.
But rather than fully embracing today's technology to conquer the domain of scientific/engineering compute servers of 10 years ago and maintain relevance in areas the Mac Pro has been losing relevance, the new Mac Pro hunkers down in the world of graphics professionals and reinforces this position by abandoning the second CPU found in all earlier generations of the Mac Pro.
2) Apple knows who is buying what Mac Pros at a certain price point. With the cost of that 12-core CPU I would doubt there are enough consumers to warrant having an entirely new design to suit those buyers. I don't know if that's true but I would say Apple likely did their research in the years it took to bring the Mac Pro to market to make sure they were targeting their Mac Pro customers.
I've owned several Mac Pros (and PowerMacs). Apple isn't targeting my field's needs. IMHO Apple chose a design that it felt could be marketed more successfully (re: "Can't innovate, my ass!") rather than maintain relevance to a broader range of disciplines.
3) Same as with point 2, I think it's a poor argument to claim the 2nd GPU is superfluous. Why would any company do that? They have a deal with AMD? Tim Cook is trying to Brewster's Millions Apple? The new Mac Pro might ultimately end up being a bad move for Apple (but based on how quickly the time to delivery was pushed back that seems doubtful) but it's clear they researched this in ways we can't imagine so I don't know what additional evidence we would have that would show it need be de duel socket and only have 1 GPU for it to be a success.
More false arguments. The 2nd GPU is fine for FCPX. You clearly don't understand the varied needs of scientific/engineering computing applications, which don't necessarily need a second GPU (or any GPU at all) but could use a 2nd CPU and more RAM.
Or perhaps they know how many people from various fields are buying their machines and made a logical decision to make a machine that would make them good money.
OMG Why the false pretenses? Who is complaining that the new Mac Pro doesn't have 32 DIMM slots?????
That was some hyperbole at an attempt at humour, but not that in this thread someone stated 64GB was the bare MINIMUM and 256GiB was the AVERAGE which makes 32x16GiB=512GiB not an unreasonable comment since it's already being stated that 256GiB is some standard for workstations.
I for one am complaining that the new Mac Pro lost an entire CPU (up to 12 cores and 24 hyperthreads) and lost 4 DIMM slots.
It did, just as the MBP lost the 2.5" drive slot as well as the maximum total storage that was possible, especially when you consider adding a drive in the ODD bay.
Is the new Mac Pro slower as a result? If the previous Mac Pro could only use 32GiB and new one 64GiB who are these Mac Pro buyers that were wanting to have 128GiB this year?
It also lost the option of having just one graphics processing card and lost virtually all internal expansion.
IMHO this is not progress.
Again, the MBPs lost the ODD and the 2.5" HDD as well the ability to use off-the-shelf 2.5" drives but I see the new Retiina MBPs as a huge leap in progress, not a regression.
I am certain the new design fits well enough with the needs of video professionals, particularly since FCPX takes good advantage of the graphics cards.
But rather than fully embracing today's technology to conquer the domain of scientific/engineering compute servers of 10 years ago and maintain relevance in areas the Mac Pro has been losing relevance, the new Mac Pro hunkers down in the world of graphics professionals and reinforces this position by abandoning the second CPU found in all earlier generations of the Mac Pro.
I'm still not getting why it's a failure with only one CPU or why Apple should focus on very specific use cases which, frankly, I've never even heard of ever using Mac Pros. Also, I have yet to see any tests yet that show how the 2nd GPU and OpenCL work in conjunction with the CPU.Would you really rather have 4x 3.5" HDD running on disk speeds over SATA II over a PCIe SSD that eschews SATA altogether? I wouldn't.
If you really an expandable Mac tower buy the new Mac Pro and stick it inside a computer case from Newegg along with an ODD and a TB RAID. You can also put an alien sticker and some neon lights on it. You could even put a second Mac Pro in there.
Or perhaps they know how many people from various fields are buying their machines and made a logical decision to make a machine that would make them good money.
It sucks if 'our' product needs aren't being met exactly how we want them to but it's not 'our' call to tell a company how they should make a product just as 'we' have the right not to buy a product that doesn't suit our needs. If we don't like it we can go elsewhere to get our needs met, or if there is nothing else out there 'we' have the option to fill what 'we' conceive is a profitable void in the market. If Apple sees they missed the mark they will try to capture it or they will lose out, but that's business. I'm sure buyers for a 2x12-core with 256GiB RAM with 8x3.5" HDDs exist (somewhere) but I don't any in this thread who had their CC out and ready to drop $40k on a new Mac that are now shaking their fist at Tim Cook.
Frankly I'll never understand the concern for simple numbers like the number of cores or how much RAM it has. We see this all the time Android having 4 cores and 2-3GB of RAM which somehow translates to it being faster. Why does any of that matter over the actual performance results. if the new Mac Pro is faster than the old one for a given price point isn't that good enough for the Mac Pro customer?
It sucks if 'our' product needs aren't being met exactly how we want them to but it's not 'our' call to tell a company how they should make a product just as 'we' have the right not to buy a product that doesn't suit our needs. If we don't like it we can go elsewhere to get our needs met, or if there is nothing else out there 'we' have the option to fill what 'we' conceive is a profitable void in the market. If Apple sees they missed the mark they will try to capture it or they will lose out, but that's business. I'm sure buyers for a 2x12-core with 256GiB RAM with 8x3.5" HDDs exist (somewhere) but I don't any in this thread who had their CC out and ready to drop $40k on a new Mac that are now shaking their fist at Tim Cook.
Frankly I'll never understand the concern for simple numbers like the number of cores or how much RAM it has. We see this all the time Android having 4 cores and 2-3GB of RAM which somehow translates to it being faster. Why does any of that matter over the actual performance results. if the new Mac Pro is faster than the old one for a given price point isn't that good enough for the Mac Pro customer?
I am not really sure what you are on about. The raw numbers (not vs android, but vs the previous generation [3 years old] Mac Pro) really don't look so hot. Now, it can be argued (and has been) that the new MP is faster once the software is tweaked, but it certainly is not a quantum leap in performance. Presumably losing half of the CPU cores might have something to with this.
I am not really sure what you are on about. The raw numbers (not vs android, but vs the previous generation [3 years old] Mac Pro) really don't look so hot. Now, it can be argued (and has been) that the new MP is faster once the software is tweaked, but it certainly is not a quantum leap in performance. Presumably losing half of the CPU cores might have something to with this.
1) Is there a price disparity in the cost of 12-core Xeons used in the Mac Pro in those 3 years?
2) When you say "once the software is tweaked" you're referring to the benchmarks right? You're not saying the OS and Apple's apps designed for the Mac Pro are slower now than they were 3 years ago for a comparatively priced machine?
Is there a price disparity in the cost of 12-core Xeons used in the Mac Pro in those 3 years?
I don't know, is there? And why does it matter? If I have the penultimate Mac Pro, what compelling reason is there to upgrade to the new one, given that I am apparently not getting a major performance bump? (this is _not_ a rhetorical question, I am genuinely curious).
What I'm likely to do is skip the Mac Dysan, use my MBP for a graphics workstation (with dual external monitors) and build a dual-processor (24-core) compute server (linux) for about the same money as the least-expensive 12-core Mac Pro, then add 128 GB RAM for another couple grand (or 256 GB if I can find another two grand). I can always add one or more GPUs later as needed. Much of this will be funded by retiring old Mac Pros that Apple refuses to support any longer.
Outlook: geekbench scores above 43K, versus 24K for the new 12-core Mac Pro and 22K for the previous generation 12-core Mac Pro. (yes, that's my primary business, running geekbench! ;-)
I don't know, is there? And why does it matter? If I have the penultimate Mac Pro, what compelling reason is there to upgrade to the new one, given that I am apparently not getting a major performance bump? (this is _not_ a rhetorical question, I am genuinely curious).
Think about it for a second from a PC vendor and not a consumer. Why would it matter to Apple if the cost of a single 12-core Xeon from Intel had dramatically increased in price since their update 3 years ago?
I only see 2 12-core Xeons on Intel's price list sheet:
If these base costs are, say, $1000 more than what their price list was 3 years ago and Apple already knows that their Mac Pro customers simply aren't paying for the extra core when the price gets to a certain point then they have to make a decision: try to make the update worthwhile (hopefully so they don't go 3 years before updates) -or- focus on trying to make everyone on internet forums happy armchair engineers, marketers, CEOs, CTOs, and everyone else involved in bringing the new Mac Pro to market.
The new Mac Pro would never have had dual 12-core CPUs. The old one maxed out at dual 6-core when there were dual 8-core options because of the price. Apple also never offered 128GB RAM options. A dual 12-core from HP is $10k with no GPUs - the CPUs retail for $3500 each. Dual D700 is $1000 so the extra top-end GPU is $500. For OpenCL tasks, that D700 will match or exceed the 12-core so you get $3500 of compute performance for $500 for tasks that use them.
If CPU performance is the priority, an option is to get two (or more) 12-cores with D300s for $6500 each. It's not as easy to manage networked machines vs a single CPU but people who need the most CPU cores possible, will most likely have these kind of setups anyway along with a queue manager.
OWC confirmed the nMP supports at least 6x 27" 1440p displays. This is something someone asked about for stock trading:
If these base costs are, say, $1000 more than what their price list was 3 years ago and Apple already knows that their Mac Pro customers simply aren't paying for the extra core when the price gets to a certain point then they have to make a decision
Intel didn't have a 12-core chip before, it was dual 6-core. The dual 6-core was 2x $1440 = $2880 so this new single chip 12-core is slightly lower priced. It would have been nicer if they'd priced it like a single chip though because that would mean better performance per dollar. Now it's around the same price, same performance and it's just in one chip vs two.
As far as Apple's decision to go with a single CPU, it doesn't make them less competitive than their competition in the higher volume price range because they're using the same chips. HP's E5-1650v2 is $4030, Apple's is $3999, slightly varied options but mostly the same. Buying two 12-core Mac Pros at $13k vs a single 24-core HP for $10k + GPU/SSD/PSU cost is not really a problem as it doesn't really affect that many people.
Think about it for a second from a PC vendor and not a consumer. Why would it matter to Apple if the cost of a single 12-core Xeon from Intel had dramatically increased in price since their update 3 years ago?
I only see 2 12-core Xeons on Intel's price list sheet:
If these base costs are, say, $1000 more than what their price list was 3 years ago and Apple already knows that their Mac Pro customers simply aren't paying for the extra core when the price gets to a certain point then they have to make a decision: try to make the update worthwhile (hopefully so they don't go 3 years before updates) -or- focus on trying to make everyone on internet forums happy armchair engineers, marketers, CEOs, CTOs, and everyone else involved in bringing the new Mac Pro to market.
Part of the reason (presumably a nontrivial part) of why Apple is bringing out a new Mac Pro is for their existing pro user base. What I am saying is that while I am not part of that user base (in the sense that I don't have the 2010-12 MacPro, and don't do video editing), if I were, I don't see why I would want to spend a lot of money for a machine giving me NO performance advantage (as per TheVerge benchmarks in my previous post). As I said, since I am not in that group, I might be missing something obvious.
The new Mac Pro would never have had dual 12-core CPUs. The old one maxed out at dual 6-core when there were dual 8-core options because of the price. Apple also never offered 128GB RAM options. A dual 12-core from HP is $10k with no GPUs - the CPUs retail for $3500 each. Dual D700 is $1000 so the extra top-end GPU is $500. For OpenCL tasks, that D700 will match or exceed the 12-core so you get $3500 of compute performance for $500 for tasks that use them.
If CPU performance is the priority, an option is to get two (or more) 12-cores with D300s for $6500 each. It's not as easy to manage networked machines vs a single CPU but people who need the most CPU cores possible, will most likely have these kind of setups anyway along with a queue manager.
OWC confirmed the nMP supports at least 6x 27" 1440p displays. This is something someone asked about for stock trading:
Re the last item, that's very interesting, but very strangely misguided: I do a lot of stock trading, and all of that world is very solidly windows based (I would not have a windows machines if it weren't, but it is, so I do). Since trading workstations are generally chained to the floor, portability is not really a plus, and neither is dual-booting.
As for performance, the benchmarks I have mentioned is for performance, period. Whether the machine diddles its bits on CPU or the GPU is not so relevant. I think downgrading to D300 will degrade system performance a lot (though since everyone seems to be benchmarking high end machines, this seems a bit hard to tell).
Part of the reason (presumably a nontrivial part) of why Apple is bringing out a new Mac Pro is for their existing pro user base. What I am saying is that while I am not part of that user base (in the sense that I don't have the 2010-12 MacPro, and don't do video editing), if I were, I don't see why I would want to spend a lot of money for a machine giving me NO performance advantage (as per TheVerge benchmarks in my previous post). As I said, since I am not in that group, I might be missing something obvious.
Oh I completely agree that is there is no performance advantage for their intended usage between the 2010 and 2013 version then there is no reason to upgrade or if one is looking to buy a Mac Pro should consider the previous model.
Note that they didn't really do any reasonable testing of apps, which is a key point one should look for when making this kind of buying decision. They do write the below quoted paragraph, but without any comparable times between new and old Mac Pros, and the new iMac to give it any relevance. My guess is this will show it's much faster than the 2010 version and with Adobe updates their apps it will show the same thing.
"The Verge is a Premiere house. (Once Final Cut Pro 7 was discontinued, FCP X didn't look like it was going to satisfy our needs.) However, since FCP X was specifically optimized for the new Mac Pro, we tested our RED footage with the app and it handled native footage from the Epic shockingly well. For this test, I turned off auto-render and set the playback quality to "better performance." I was able to layer four streams, resized and composed on top of each other with color correction on each clip, and FCP X played the composite back without stuttering or dropping frames."
PS: These discussions feel like a conversion about the Tesla Roadster but looking at a chart that's judging the number of pistons and how much trunk space it has.
What I'm likely to do is skip the Mac Dysan, use my MBP for a graphics workstation (with dual external monitors) and build a dual-processor (24-core) compute server (linux) for about the same money as the least-expensive 12-core Mac Pro, then add 128 GB RAM for another couple grand (or 256 GB if I can find another two grand). I can always add one or more GPUs later as needed. Much of this will be funded by retiring old Mac Pros that Apple refuses to support any longer.
Outlook: geekbench scores above 43K, versus 24K for the new 12-core Mac Pro and 22K for the previous generation 12-core Mac Pro. (yes, that's my primary business, running geekbench! ;-)
Actually, I don't see you building a dual processor linux server for so cheap. Dual processor means xeon, and the cheapest 12 core xeons ar $2.6K.
Oh I completely agree that is there is no performance advantage for their intended usage between the 2010 and 2013 version then there is no reason to upgrade or if one is looking to buy a Mac Pro should consider the previous model.
Note that they didn't really do any reasonable testing of apps, which is a key point one should look for when making this kind of buying decision. They do write the below quoted paragraph, but without any comparable times between new and old Mac Pros, and the new iMac to give it any relevance. My guess is this will show it's much faster than the 2010 version and with Adobe updates their apps it will show the same thing.
"The Verge is a Premiere house. (Once Final Cut Pro 7 was discontinued, FCP X didn't look like it was going to satisfy our needs.) However, since FCP X was specifically optimized for the new Mac Pro, we tested our RED footage with the app and it handled native footage from the Epic shockingly well. For this test, I turned off auto-render and set the playback quality to "better performance." I was able to layer four streams, resized and composed on top of each other with color correction on each clip, and FCP X played the composite back without stuttering or dropping frames."
PS: These discussions feel like a conversion about the Tesla Roadster but looking at a chart that's judging the number of pistons and how much trunk space it has.
Well, re apps we will see what we see, but Adobe has not exactly been fast on the draw of late (umm, late being last ten years). As for the Tesla Roadster, interesting analogy, since the TR began and ended its days as a bleeding edge rich man's plaything (partly due to the sad lack of trunk space), so let's hope that the nMP does not suffer the same fate.
Actually, I don't see you building a dual processor linux server for so cheap. Dual processor means xeon, and the cheapest 12 core xeons ar $2.6K.
The least-expensive 12-core Mac Pro is $6,499 plus tax. Dual E5-2697 v2 processors with dual-processor SuperMicro motherboard can be had from newegg for $5500. That leaves $1K for additional components, including two CPU cooling fans, a case, a high quality power supply, and a cheap graphics card.
Edit: oops, I forgot to consider the cost of linux.
Edit 2: even if it winds up being a little more expensive than the least-expensive 12-core Mac Pro, Apple doesn't offer anything like it. Call it priceless.
The least-expensive 12-core Mac Pro is $6,499 plus tax. Dual E5-2697 v2 processors with dual-processor SuperMicro motherboard can be had from newegg for $5500. That leaves $1K for additional components, including two CPU cooling fans, a case, a high quality power supply, and a cheap graphics card.
Edit: oops, I forgot to consider the cost of linux.
Interesting, though since I do a lot of GPU computing, you would have to throw in a couple of GTX titans. Still, not a bad price.
Re the last item, that's very interesting, but very strangely misguided: I do a lot of stock trading, and all of that world is very solidly windows based (I would not have a windows machines if it weren't, but it is, so I do). Since trading workstations are generally chained to the floor, portability is not really a plus, and neither is dual-booting.
There are more uses for multiple displays than stocks, I just mentioned it because someone on the forum specifically asked if it would support 6 Thunderbolt displays as they would consider it for trading. They might use Windows too.
As for performance, the benchmarks I have mentioned is for performance, period. Whether the machine diddles its bits on CPU or the GPU is not so relevant. I think downgrading to D300 will degrade system performance a lot (though since everyone seems to be benchmarking high end machines, this seems a bit hard to tell).
The benchmarks you linked to are for the 8-core model and they use Adobe's software without the GPUs being added to the whitelist of supported GPUs. I don't think anyone has a 12-core model because they won't ship until later, possibly February.
I don't see why I would want to spend a lot of money for a machine giving me NO performance advantage (as per TheVerge benchmarks in my previous post). As I said, since I am not in that group, I might be missing something obvious.
This isn't different from any other manufacturer. You can say exactly the same about Dell and HP. At the same prices points as a slightly older machine, you don't get much performance advantage. Why reserve the criticism for Apple when it's Intel's options that are the issue and affect everyone?
The advantages with the new Mac Pro are better GPUs supporting OpenCL 1.2 and OpenGL 4 and dual GPUs where the old model only supported a single high-end GPU, quieter operation, PCIe SSD and you get Thunderbolt support with 4K display support as well as HDMI out and USB3. While Thunderbolt support is seen by some as irrelevant vs PCIe, it makes it easier to setup fast hardware RAID storage arrays that are plug and play.
Comments
Scientific compute servers (from Sun, HP and Dell) have supported 256 GB and more for over 10 years (albeit at a hefty price). In 2013, a fully redesigned Mac Pro handles only 64 GB--just double that of a 2006 Mac Pro. For professionals who don't need graphics computation, it would have been far better to include a second processor to reach 24 cores (48 hyperthreads) and more DIMM slots, while ditching the superfluous second graphics card. As indicated above, I for one am very disappointed in the new design. As for FCPX users, have fun. As for everyone else, meh.
Your emphasis of "professionals" seems to suggest that only people who don't need graphics computation are professionals, and then your final point again dismisses some of Apple's big users.
Personally, I don't like the plastic iPhone (and I absolutely cannot stand iOS7). That's life. I don't have to buy it. But, it is selling, which is good for Apple, which is their business. They are not in business to create a custom, made-to-order phone just for me. I can guarantee it would not sell well! The new MP may not be for for you. That's life. But if your comments about RAM are correct, then the MP has never been the machine for you. That doesn't make it a bad machine.
But, regarding those GPUs, they run OpenCL. Apple has already reworked (yes, reworked) Final Cut to utilize the GPUs. I know of at least two other developers (one has publicly committed to doing so) who are reworking their apps to use the GPUs to milk the machines for all they've got. Would it be possible for the "professionals" in your line of work to look into the possibility of reworking apps so that they could use the GPUs and the MP would be a very useful machine for you?
I confess to not being a techie and I have no idea what OpenCl really is or what it can and cannot do.
3) Same as with point 2, I think it's a poor argument to claim the 2nd GPU is superfluous. Why would any company do that? They have a deal with AMD? Tim Cook is trying to Brewster's Millions Apple? The new Mac Pro might ultimately end up being a bad move for Apple (but based on how quickly the time to delivery was pushed back that seems doubtful) but it's clear they researched this in ways we can't imagine so I don't know what additional evidence we would have that would show it need be de duel socket and only have 1 GPU for it to be a success.
I missed the nerdy references during your absence.
1) I don't understand why the Mac Pro is all of a sudden being negatively compared against specialized computers that need a lot more RAM than its ever had. Why wasn't this "it sucks because it does't have 32 DIMM slots on 8 risers"-style argument not ever brought up before? Why now is there such a comparison when it's never been designed for the specialized tasks it's being compared to. It's like looking at a brand new Rolls Royce Phantom and complaining that it's a horrible cargo plane.
OMG Why the false pretenses? Who is complaining that the new Mac Pro doesn't have 32 DIMM slots?????
I for one am complaining that the new Mac Pro lost an entire CPU (up to 12 cores and 24 hyperthreads) and lost 4 DIMM slots.
It also lost the option of having just one graphics processing card and lost virtually all internal expansion.
IMHO this is not progress.
I am certain the new design fits well enough with the needs of video professionals, particularly since FCPX takes good advantage of the graphics cards.
But rather than fully embracing today's technology to conquer the domain of scientific/engineering compute servers of 10 years ago and maintain relevance in areas the Mac Pro has been losing relevance, the new Mac Pro hunkers down in the world of graphics professionals and reinforces this position by abandoning the second CPU found in all earlier generations of the Mac Pro.
Or perhaps they know how many people from various fields are buying their machines and made a logical decision to make a machine that would make them good money.
That was some hyperbole at an attempt at humour, but not that in this thread someone stated 64GB was the bare MINIMUM and 256GiB was the AVERAGE which makes 32x16GiB=512GiB not an unreasonable comment since it's already being stated that 256GiB is some standard for workstations.
It did, just as the MBP lost the 2.5" drive slot as well as the maximum total storage that was possible, especially when you consider adding a drive in the ODD bay.
Is the new Mac Pro slower as a result? If the previous Mac Pro could only use 32GiB and new one 64GiB who are these Mac Pro buyers that were wanting to have 128GiB this year?
Again, the MBPs lost the ODD and the 2.5" HDD as well the ability to use off-the-shelf 2.5" drives but I see the new Retiina MBPs as a huge leap in progress, not a regression.
I'm still not getting why it's a failure with only one CPU or why Apple should focus on very specific use cases which, frankly, I've never even heard of ever using Mac Pros. Also, I have yet to see any tests yet that show how the 2nd GPU and OpenCL work in conjunction with the CPU.Would you really rather have 4x 3.5" HDD running on disk speeds over SATA II over a PCIe SSD that eschews SATA altogether? I wouldn't.
If you really an expandable Mac tower buy the new Mac Pro and stick it inside a computer case from Newegg along with an ODD and a TB RAID. You can also put an alien sticker and some neon lights on it. You could even put a second Mac Pro in there.
It sucks if 'our' product needs aren't being met exactly how we want them to but it's not 'our' call to tell a company how they should make a product just as 'we' have the right not to buy a product that doesn't suit our needs. If we don't like it we can go elsewhere to get our needs met, or if there is nothing else out there 'we' have the option to fill what 'we' conceive is a profitable void in the market. If Apple sees they missed the mark they will try to capture it or they will lose out, but that's business. I'm sure buyers for a 2x12-core with 256GiB RAM with 8x3.5" HDDs exist (somewhere) but I don't any in this thread who had their CC out and ready to drop $40k on a new Mac that are now shaking their fist at Tim Cook.
Frankly I'll never understand the concern for simple numbers like the number of cores or how much RAM it has. We see this all the time Android having 4 cores and 2-3GB of RAM which somehow translates to it being faster. Why does any of that matter over the actual performance results. if the new Mac Pro is faster than the old one for a given price point isn't that good enough for the Mac Pro customer?
It sucks if 'our' product needs aren't being met exactly how we want them to but it's not 'our' call to tell a company how they should make a product just as 'we' have the right not to buy a product that doesn't suit our needs. If we don't like it we can go elsewhere to get our needs met, or if there is nothing else out there 'we' have the option to fill what 'we' conceive is a profitable void in the market. If Apple sees they missed the mark they will try to capture it or they will lose out, but that's business. I'm sure buyers for a 2x12-core with 256GiB RAM with 8x3.5" HDDs exist (somewhere) but I don't any in this thread who had their CC out and ready to drop $40k on a new Mac that are now shaking their fist at Tim Cook.
Frankly I'll never understand the concern for simple numbers like the number of cores or how much RAM it has. We see this all the time Android having 4 cores and 2-3GB of RAM which somehow translates to it being faster. Why does any of that matter over the actual performance results. if the new Mac Pro is faster than the old one for a given price point isn't that good enough for the Mac Pro customer?
I am not really sure what you are on about. The raw numbers (not vs android, but vs the previous generation [3 years old] Mac Pro) really don't look so hot. Now, it can be argued (and has been) that the new MP is faster once the software is tweaked, but it certainly is not a quantum leap in performance. Presumably losing half of the CPU cores might have something to with this.
1) Is there a price disparity in the cost of 12-core Xeons used in the Mac Pro in those 3 years?
2) When you say "once the software is tweaked" you're referring to the benchmarks right? You're not saying the OS and Apple's apps designed for the Mac Pro are slower now than they were 3 years ago for a comparatively priced machine?
Is there a price disparity in the cost of 12-core Xeons used in the Mac Pro in those 3 years?
I don't know, is there? And why does it matter? If I have the penultimate Mac Pro, what compelling reason is there to upgrade to the new one, given that I am apparently not getting a major performance bump? (this is _not_ a rhetorical question, I am genuinely curious).
What I'm likely to do is skip the Mac Dysan, use my MBP for a graphics workstation (with dual external monitors) and build a dual-processor (24-core) compute server (linux) for about the same money as the least-expensive 12-core Mac Pro, then add 128 GB RAM for another couple grand (or 256 GB if I can find another two grand). I can always add one or more GPUs later as needed. Much of this will be funded by retiring old Mac Pros that Apple refuses to support any longer.
Outlook: geekbench scores above 43K, versus 24K for the new 12-core Mac Pro and 22K for the previous generation 12-core Mac Pro. (yes, that's my primary business, running geekbench! ;-)
Think about it for a second from a PC vendor and not a consumer. Why would it matter to Apple if the cost of a single 12-core Xeon from Intel had dramatically increased in price since their update 3 years ago?
I only see 2 12-core Xeons on Intel's price list sheet:
E5-2695 v2 (30M cache, 12 Cores, 24 Threads, 2.40 GHz (115W) 8.00 GT/sec Intel® QPI, 22nm) $2,336
If these base costs are, say, $1000 more than what their price list was 3 years ago and Apple already knows that their Mac Pro customers simply aren't paying for the extra core when the price gets to a certain point then they have to make a decision: try to make the update worthwhile (hopefully so they don't go 3 years before updates) -or- focus on trying to make everyone on internet forums happy armchair engineers, marketers, CEOs, CTOs, and everyone else involved in bringing the new Mac Pro to market.
If CPU performance is the priority, an option is to get two (or more) 12-cores with D300s for $6500 each. It's not as easy to manage networked machines vs a single CPU but people who need the most CPU cores possible, will most likely have these kind of setups anyway along with a queue manager.
OWC confirmed the nMP supports at least 6x 27" 1440p displays. This is something someone asked about for stock trading:
http://blog.macsales.com/22073-what-do-you-call-6-displays-and-a-2013-mac-pro-fun
Intel didn't have a 12-core chip before, it was dual 6-core. The dual 6-core was 2x $1440 = $2880 so this new single chip 12-core is slightly lower priced. It would have been nicer if they'd priced it like a single chip though because that would mean better performance per dollar. Now it's around the same price, same performance and it's just in one chip vs two.
As far as Apple's decision to go with a single CPU, it doesn't make them less competitive than their competition in the higher volume price range because they're using the same chips. HP's E5-1650v2 is $4030, Apple's is $3999, slightly varied options but mostly the same. Buying two 12-core Mac Pros at $13k vs a single 24-core HP for $10k + GPU/SSD/PSU cost is not really a problem as it doesn't really affect that many people.
Think about it for a second from a PC vendor and not a consumer. Why would it matter to Apple if the cost of a single 12-core Xeon from Intel had dramatically increased in price since their update 3 years ago?
I only see 2 12-core Xeons on Intel's price list sheet:
E5-2695 v2 (30M cache, 12 Cores, 24 Threads, 2.40 GHz (115W) 8.00 GT/sec Intel® QPI, 22nm) $2,336
If these base costs are, say, $1000 more than what their price list was 3 years ago and Apple already knows that their Mac Pro customers simply aren't paying for the extra core when the price gets to a certain point then they have to make a decision: try to make the update worthwhile (hopefully so they don't go 3 years before updates) -or- focus on trying to make everyone on internet forums happy armchair engineers, marketers, CEOs, CTOs, and everyone else involved in bringing the new Mac Pro to market.
Part of the reason (presumably a nontrivial part) of why Apple is bringing out a new Mac Pro is for their existing pro user base. What I am saying is that while I am not part of that user base (in the sense that I don't have the 2010-12 MacPro, and don't do video editing), if I were, I don't see why I would want to spend a lot of money for a machine giving me NO performance advantage (as per TheVerge benchmarks in my previous post). As I said, since I am not in that group, I might be missing something obvious.
The new Mac Pro would never have had dual 12-core CPUs. The old one maxed out at dual 6-core when there were dual 8-core options because of the price. Apple also never offered 128GB RAM options. A dual 12-core from HP is $10k with no GPUs - the CPUs retail for $3500 each. Dual D700 is $1000 so the extra top-end GPU is $500. For OpenCL tasks, that D700 will match or exceed the 12-core so you get $3500 of compute performance for $500 for tasks that use them.
If CPU performance is the priority, an option is to get two (or more) 12-cores with D300s for $6500 each. It's not as easy to manage networked machines vs a single CPU but people who need the most CPU cores possible, will most likely have these kind of setups anyway along with a queue manager.
OWC confirmed the nMP supports at least 6x 27" 1440p displays. This is something someone asked about for stock trading:
http://blog.macsales.com/22073-what-do-you-call-6-displays-and-a-2013-mac-pro-fun
Re the last item, that's very interesting, but very strangely misguided: I do a lot of stock trading, and all of that world is very solidly windows based (I would not have a windows machines if it weren't, but it is, so I do). Since trading workstations are generally chained to the floor, portability is not really a plus, and neither is dual-booting.
As for performance, the benchmarks I have mentioned is for performance, period. Whether the machine diddles its bits on CPU or the GPU is not so relevant. I think downgrading to D300 will degrade system performance a lot (though since everyone seems to be benchmarking high end machines, this seems a bit hard to tell).
Oh I completely agree that is there is no performance advantage for their intended usage between the 2010 and 2013 version then there is no reason to upgrade or if one is looking to buy a Mac Pro should consider the previous model.
Note that they didn't really do any reasonable testing of apps, which is a key point one should look for when making this kind of buying decision. They do write the below quoted paragraph, but without any comparable times between new and old Mac Pros, and the new iMac to give it any relevance. My guess is this will show it's much faster than the 2010 version and with Adobe updates their apps it will show the same thing.
PS: These discussions feel like a conversion about the Tesla Roadster but looking at a chart that's judging the number of pistons and how much trunk space it has.
What I'm likely to do is skip the Mac Dysan, use my MBP for a graphics workstation (with dual external monitors) and build a dual-processor (24-core) compute server (linux) for about the same money as the least-expensive 12-core Mac Pro, then add 128 GB RAM for another couple grand (or 256 GB if I can find another two grand). I can always add one or more GPUs later as needed. Much of this will be funded by retiring old Mac Pros that Apple refuses to support any longer.
Outlook: geekbench scores above 43K, versus 24K for the new 12-core Mac Pro and 22K for the previous generation 12-core Mac Pro. (yes, that's my primary business, running geekbench! ;-)
Actually, I don't see you building a dual processor linux server for so cheap. Dual processor means xeon, and the cheapest 12 core xeons ar $2.6K.
Oh I completely agree that is there is no performance advantage for their intended usage between the 2010 and 2013 version then there is no reason to upgrade or if one is looking to buy a Mac Pro should consider the previous model.
Note that they didn't really do any reasonable testing of apps, which is a key point one should look for when making this kind of buying decision. They do write the below quoted paragraph, but without any comparable times between new and old Mac Pros, and the new iMac to give it any relevance. My guess is this will show it's much faster than the 2010 version and with Adobe updates their apps it will show the same thing.
PS: These discussions feel like a conversion about the Tesla Roadster but looking at a chart that's judging the number of pistons and how much trunk space it has.
Well, re apps we will see what we see, but Adobe has not exactly been fast on the draw of late (umm, late being last ten years). As for the Tesla Roadster, interesting analogy, since the TR began and ended its days as a bleeding edge rich man's plaything (partly due to the sad lack of trunk space), so let's hope that the nMP does not suffer the same fate.
Actually, I don't see you building a dual processor linux server for so cheap. Dual processor means xeon, and the cheapest 12 core xeons ar $2.6K.
The least-expensive 12-core Mac Pro is $6,499 plus tax. Dual E5-2697 v2 processors with dual-processor SuperMicro motherboard can be had from newegg for $5500. That leaves $1K for additional components, including two CPU cooling fans, a case, a high quality power supply, and a cheap graphics card.
Edit: oops, I forgot to consider the cost of linux.
Edit 2: even if it winds up being a little more expensive than the least-expensive 12-core Mac Pro, Apple doesn't offer anything like it. Call it priceless.
The least-expensive 12-core Mac Pro is $6,499 plus tax. Dual E5-2697 v2 processors with dual-processor SuperMicro motherboard can be had from newegg for $5500. That leaves $1K for additional components, including two CPU cooling fans, a case, a high quality power supply, and a cheap graphics card.
Edit: oops, I forgot to consider the cost of linux.
Interesting, though since I do a lot of GPU computing, you would have to throw in a couple of GTX titans. Still, not a bad price.
There are more uses for multiple displays than stocks, I just mentioned it because someone on the forum specifically asked if it would support 6 Thunderbolt displays as they would consider it for trading. They might use Windows too.
The benchmarks you linked to are for the 8-core model and they use Adobe's software without the GPUs being added to the whitelist of supported GPUs. I don't think anyone has a 12-core model because they won't ship until later, possibly February.
This isn't different from any other manufacturer. You can say exactly the same about Dell and HP. At the same prices points as a slightly older machine, you don't get much performance advantage. Why reserve the criticism for Apple when it's Intel's options that are the issue and affect everyone?
The advantages with the new Mac Pro are better GPUs supporting OpenCL 1.2 and OpenGL 4 and dual GPUs where the old model only supported a single high-end GPU, quieter operation, PCIe SSD and you get Thunderbolt support with 4K display support as well as HDMI out and USB3. While Thunderbolt support is seen by some as irrelevant vs PCIe, it makes it easier to setup fast hardware RAID storage arrays that are plug and play.