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.
It's probably common in trading due to times when more than one market is on trading hours. I typically use secondary displays for reference data and client notes, so they don't have to be of the same quality. If an older display is just no longer optimal and the backlight still holds, I retire it to secondary use.
Quote:
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.
Ah I think you mean 24 core machines from other vendors. The other vendors tend to segregate things differently. Most have a cheaper line that only uses the 1600s variants. Those would max out at 6 cores. On lines that support dual socket configurations, they have to use 2600s. In these configurations a 12 core would typically be 2 x 6, unless it somehow works out cheaper for the individual to populate one socket with a 12 core than 2 x 6. Looking at intel's lists, it is probably not that far apart, but you come out a bit better with 2 x 6, and of course they can ship before February.
Quote:
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.
That part is highly debatable. Installing a roc or a simple host card to a box with an embedded controller isn't really difficult, and you may have a more stable range of options. Given the way they're outfitted, I don't see why they chose to claim Raid 5 support. It is weird to do that without the shorter firmware timings. On a side note I wonder if these will be updated to support OpenCL 2.0, as it's now finalized. I ask because these mac pros will probably be on sale into 2015, and OpenCL 2.0 supposedly adds a lot to the toolset. I haven't looked through the entire specification. I'm merely annoyed they still don't support OpenCL on iOS. It seems like a good way to attract a greater range of productivity apps.
All the Mac Pro circuit boards look so beautifully designed.
The move to black circuit boards was a good one; it makes for gorgeous internals. And it fits right in with:
You have to make the back of the fence that nobody will see just as good looking as the front of the fence. Even though nobody will see it, you will know, and that will show that you're dedicated to making something perfect.
Open CL reminds me of that skit in Monty Python and the Holy Grail, where Sir Lancelot is running to storm the castle and every time the guards look up, Lancelot’s approach loops back to a long way off in the distance, then all of a sudden he skewers one of the guards with an Aha!
Hopefully Apple’s dual GPU decision is going to cause that Aha to come from the developers of software that could use a leap forward like 3D rendering.
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.
Cpspro, are you planning on putting dual Nvidia Quadro or AMD FirePro cards in your linux server? If so good luck with the pricing!
Also, GPUs are not just for graphics anymore. There is this new thing called OpenCL. So in reality the new Mac Pro already has 3 very powerful processing units!
It's taken almost 6 years for Apple to push out a redesigned Mac Pro that's a whopping twice as fast and supports less memory than last year's model.
I am impressed.
Quote:
Originally Posted by Cpsro
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 2nd quote was essentially announced by the 1st. Should I be impressed, anyways?
BTW, you refer to "Dysan" in a couple of posts. Do you mean Dyson? Or was the error intentional, as Dysan is flagged by my spell-checker, whereas Dyson is not.
The former is a media company that no longer exists.
Fanboys can be strange and myopic. I am an Apple zealot extraordinaire, but I don't let that get in the way of expressing my needs and disappointment in the direction the Mac Pro has been taken. Sorry, folks!
Repeating... I don't need or want GPUs (at least at this time), hence the lack of GPUs in my custom configured linux system which has twice the performance (where I need it) compared to the least-expensive 12-core Mac Pro and for about the same price. I could have received the parts last week or earlier--not wait a couple months for when they should be even cheaper (and the profit margins greater ;-)
Fanboys can be strange and myopic. I am an Apple zealot extraordinaire, but I don't let that get in the way of expressing my needs and disappointment in the direction the Mac Pro has been taken. Sorry, folks!
Repeating... I don't need or want GPUs (at least at this time), hence the lack of GPUs in my custom configured linux system which has twice the performance (where I need it) compared to the least-expensive 12-core Mac Pro and for about the same price. I could have received the parts last week or earlier--not wait a couple months for when they should be even cheaper (and the profit margins greater ;-)
I wouldn't buy a Dyson either.
It makes no sense to build a fictional machine and then say "x is doing it wrong because they don't offer y." It's fine to be disappointed that a vendor isn't servicing your specific needs and it's fine to point what would be ideal for you, but your comments are going well beyond that.
Fanboys can be strange and myopic. I am an Apple zealot extraordinaire, but I don't let that get in the way of expressing my needs and disappointment in the direction the Mac Pro has been taken. Sorry, folks!
Repeating... I don't need or want GPUs (at least at this time), hence the lack of GPUs in my custom configured linux system which has twice the performance (where I need it) compared to the least-expensive 12-core Mac Pro and for about the same price. I could have received the parts last week or earlier--not wait a couple months for when they should be even cheaper (and the profit margins greater ;-)
I wouldn't buy a Dyson either.
I actually believe that GPUs are a very cost effective way to get bang for buck. But not the Quadro/FirePro. Nvidia GTX titan is around $1000 and performs at near Tesla levels. Or, of course, there are the actual Tesla boards. Problem is, if you configure your machine with two xeons and two GPU boards, you need a 1.2Kw power supply, and forget about the 11lb weight. Many people couldn't care less. But not all.
The decision to include 1 CPU and 2 GPUs, as against the traditional workstation model of 2 CPUs and 1 GPU is a daring move. But it's also a recognition that todays GPUs are really general purpose stream processors, that can do things that are nothing to do with graphics.
For example, searching through an array of data for a particular byte value: 16 CPU cores, with 64-bit wide registers can check 16 x 8 = 128 array slots per unit time. Two D700 GPUs, with 2048 32-bit stream processors each, can check 4096 x 4 = 16,384 array slots per unit time.
The first time you see this is real life - the first time you see a "gamer" GPU do something (nothing to do with graphics) 128 times faster than that Xeon you paid all that money for (this has happened to me) it's a paradigm shift in your mind.
I never even thought about the color of the logic boards.
It's like the hardware designer equivalent of wearing clean underwear.
Quote:
Originally Posted by marubeni
I actually believe that GPUs are a very cost effective way to get bang for buck. But not the Quadro/FirePro. Nvidia GTX titan is around $1000 and performs at near Tesla levels. Or, of course, there are the actual Tesla boards. Problem is, if you configure your machine with two xeons and two GPU boards, you need a 1.2Kw power supply, and forget about the 11lb weight. Many people couldn't care less. But not all.
A Dyson is a VERY nice vacuum cleaner.
The 780s are also quite good. The biggest thing in favor of titan is probably the 6GB. If it wasn't for that I would say the 780s have a better price to performance ratio.
Quote:
Originally Posted by ascii
For example, searching through an array of data for a particular byte value: 16 CPU cores, with 64-bit wide registers can check 16 x 8 = 128 array slots per unit time. Two D700 GPUs, with 2048 32-bit stream processors each, can check 4096 x 4 = 16,384 array slots per unit time.
It's kind of like I said once before. They're great for really really parallel workloads. Interestingly there are some areas within graphics and visualization where that could really be exploited. I suspect it's an issue of existing code bases, older algorithms, and lack of a clear future in terms of what framework ends up being dominant. That being said, I just picked up a book on OpenCL and heterogeneous programming.
As others have noted, all that says is that the storage device can be removed by the user. It doesn't mean anyone, including Apple, is gonna sell you something else to put there.
While the move to solid-state storage is obviously a good one, I'm not yet sure whether the relatively small speed advantage likely to be realized by the PCIe interface offers enough benefit to offset the huge disadvantage of being incompatible with worldwide standards. Crucial and Samsung are constantly improving their storage offerings while also lowering prices. Apple computers don't reap those benefits. Is Apple's approach enough better to make it worth being a lone voice in the wilderness?
Quote:
Originally Posted by PhilBoogie
Good grief man! Get a grip. If Apple isn't going to do some serious forward-thinking, then who is? Otherwise we'd be stuck with floppies, user-removable batteries, physical keyboards and DVD's, to name just a few.
With due respect, different use cases make certain moves more or less desirable. For some, the benefits of certain changes are outweighed by the liabilities. For example, to the guy who spends 12 hours a day in the field, a user-replaceable battery matters more than a mm of thinness. That saved mm is gonna cost him DOUBLING the thickness of his device through the use of an external battery.
Enjoy the benefits of new approaches but don't wear blinders. Sometimes newer ain't necessarily better, or is better in some ways at the expense of being worse in others. By staying aware of which babies were thrown out with the bathwater, hopefully we can add them back in as technology advances (resulting in physical keyboards that can appear/disappear and change shape on demand, for example).
Quote:
Originally Posted by SolipsismX
Apple knows who is buying what Mac Pros at a certain price point. [...] 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 don't know if that's true. We tend to give Apple an awful lot of credit regarding how well they know their customer base, but do they really? If so, how? I own a whole houseful of Apple products and manage a business that uses them, and Apple has never once asked me how I use them or how they could make my experience better.
All Apple knows for sure is which models sell and which don't, with no confirmed way of knowing WHY. For example, they might see that the 64GB version of the iPhone doesn't sell worth a damn and conclude that people don't need or want that much storage, when the reality is that buyers would LOVE to have 64GB or more but just can't justify or afford the extra $200! Apple has no better source of such knowledge than anyone else, and my own experience leads me to wonder if their market data is even as good as some of their competitors?
Quote:
Originally Posted by SolipsismX
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 don't think the current design has nearly as much to do with market research as with what best satisfies the demands of Apple's software product, specifically FCPX. Just look at the test results so far: with Adobe Premiere, meh. With FCPX, wow! Coincidence?
That aside, I don't think anyone besides CS is saying that Apple should have gone with 2xCPU/1xGPU *instead* of 1xCPU/2xGPU, but could very easily have offered such configurations *as well.* I haven't bothered looking into it since it's not an option, but I suspect our needs here would probably be better served with more CPU per dollar and less GPU expense.
Quote:
Originally Posted by SolipsismX
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.
I disagree. I think it's perfectly reasonable for the consumers of these items to give feedback to the supplier about what they want and need. THAT'S where Apple can get some genuine, useful market data.
My clients give me real-time feedback about what they want. Without that I'd have only my best guesses based on return customers on which to base my decision making. Obviously giving/getting feedback is better for both of us.
If Apple doesn't want to listen, that's their prerogative, but I don't see anything wrong with the buyer telling the vendor what they want.
While the move to solid-state storage is obviously a good one, I'm not yet sure whether the relatively small speed advantage likely to be realized by the PCIe interface offers enough benefit to offset the huge disadvantage of being incompatible with worldwide standards. Crucial and Samsung are constantly improving their storage offerings while also lowering prices. Apple computers don't reap those benefits. Is Apple's approach enough better to make it worth being a lone voice in the wilderness?
"We also ran the popular Black Magic disk test on our review unit .. the result was a write speed of 952.8 MB/sec and a read speed of 920.5 MB/sec"
SATA 3 interfaces max out at about 550 MB/sec. I have the latest Samsung EVO SSD in my PC and *wish* it was using PCI connectivity.
Oh. Okay, that's a bigger advantage than I thought. That makes the question even harder to answer!
Do speeds like that really happen in real life, or do other factors limit the rate at which the system can read and write? Would this configuration actually BE faster than a SATA3 interface or only have a "theoretical" advantage?
I'm confused - what kind of computing are you doing? From the memory requirements that you are quoting it sounds like large 3D numerical simulation, but that field has not been dominated by workstations for many years. We use clusters of Mac Pros for small problems to avoid the hassle of getting on the big machines, but not for anything serious. What kind of workstation setups are you referring to with 256 GB?
That is exactly what I was wondering as well. Almost all of the scientific computing work that I have done in the last two decades has been done on clusters. I still remember my first time using a Beowulf setup which cobbled together a bunch of older PC motherboards. Then again, I haven't used Mathematica/Maple for anything to this level. We pretty much make our own programs to reduce as much overhead as possible.
Oh. Okay, that's a bigger advantage than I thought. That makes the question even harder to answer!
Do speeds like that really happen in real life, or do other factors limit the rate at which the system can read and write? Would this configuration actually BE faster than a SATA3 interface or only have a "theoretical" advantage?
I think the theoretical limit is 1200MB/sec, so the 950MB/sec figure is what you can really expect. As for whether it would be noticeable in real life I guess that depends on the size of file you're working with. A 10GB video file might take 10s to load on a Mac Pro but 20s on a PC, but a 1MB Word file would be instantaneous on both.
Comments
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.
It's probably common in trading due to times when more than one market is on trading hours. I typically use secondary displays for reference data and client notes, so they don't have to be of the same quality. If an older display is just no longer optimal and the backlight still holds, I retire it to secondary use.
Quote:
Ah I think you mean 24 core machines from other vendors. The other vendors tend to segregate things differently. Most have a cheaper line that only uses the 1600s variants. Those would max out at 6 cores. On lines that support dual socket configurations, they have to use 2600s. In these configurations a 12 core would typically be 2 x 6, unless it somehow works out cheaper for the individual to populate one socket with a 12 core than 2 x 6. Looking at intel's lists, it is probably not that far apart, but you come out a bit better with 2 x 6, and of course they can ship before February.
That part is highly debatable. Installing a roc or a simple host card to a box with an embedded controller isn't really difficult, and you may have a more stable range of options. Given the way they're outfitted, I don't see why they chose to claim Raid 5 support. It is weird to do that without the shorter firmware timings. On a side note I wonder if these will be updated to support OpenCL 2.0, as it's now finalized. I ask because these mac pros will probably be on sale into 2015, and OpenCL 2.0 supposedly adds a lot to the toolset. I haven't looked through the entire specification. I'm merely annoyed they still don't support OpenCL on iOS. It seems like a good way to attract a greater range of productivity apps.
The move to black circuit boards was a good one; it makes for gorgeous internals. And it fits right in with:
I never even thought about the color of the logic boards.
Open CL reminds me of that skit in Monty Python and the Holy Grail, where Sir Lancelot is running to storm the castle and every time the guards look up, Lancelot’s approach loops back to a long way off in the distance, then all of a sudden he skewers one of the guards with an Aha!
Hopefully Apple’s dual GPU decision is going to cause that Aha to come from the developers of software that could use a leap forward like 3D rendering.
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.
Cpspro, are you planning on putting dual Nvidia Quadro or AMD FirePro cards in your linux server? If so good luck with the pricing!
Also, GPUs are not just for graphics anymore. There is this new thing called OpenCL. So in reality the new Mac Pro already has 3 very powerful processing units!
http://en.wikipedia.org/wiki/OpenCL
http://www.khronos.org/opencl/
P. S. Don't forget to look at who is the original creator/author of Open CL. I think you may be underestimating Apple.
It's taken almost 6 years for Apple to push out a redesigned Mac Pro that's a whopping twice as fast and supports less memory than last year's model.
I am impressed.
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 2nd quote was essentially announced by the 1st. Should I be impressed, anyways?
BTW, you refer to "Dysan" in a couple of posts. Do you mean Dyson? Or was the error intentional, as Dysan is flagged by my spell-checker, whereas Dyson is not.
The former is a media company that no longer exists.
http://en.wikipedia.org/wiki/Dysan
The latter is the vacuum cleaner (among other things) maker.
http://www.dyson.com/default.aspx
Fanboys can be strange and myopic. I am an Apple zealot extraordinaire, but I don't let that get in the way of expressing my needs and disappointment in the direction the Mac Pro has been taken. Sorry, folks!
Repeating... I don't need or want GPUs (at least at this time), hence the lack of GPUs in my custom configured linux system which has twice the performance (where I need it) compared to the least-expensive 12-core Mac Pro and for about the same price. I could have received the parts last week or earlier--not wait a couple months for when they should be even cheaper (and the profit margins greater ;-)
I wouldn't buy a Dyson either.
It makes no sense to build a fictional machine and then say "x is doing it wrong because they don't offer y." It's fine to be disappointed that a vendor isn't servicing your specific needs and it's fine to point what would be ideal for you, but your comments are going well beyond that.
Fanboys can be strange and myopic. I am an Apple zealot extraordinaire, but I don't let that get in the way of expressing my needs and disappointment in the direction the Mac Pro has been taken. Sorry, folks!
Repeating... I don't need or want GPUs (at least at this time), hence the lack of GPUs in my custom configured linux system which has twice the performance (where I need it) compared to the least-expensive 12-core Mac Pro and for about the same price. I could have received the parts last week or earlier--not wait a couple months for when they should be even cheaper (and the profit margins greater ;-)
I wouldn't buy a Dyson either.
I actually believe that GPUs are a very cost effective way to get bang for buck. But not the Quadro/FirePro. Nvidia GTX titan is around $1000 and performs at near Tesla levels. Or, of course, there are the actual Tesla boards. Problem is, if you configure your machine with two xeons and two GPU boards, you need a 1.2Kw power supply, and forget about the 11lb weight. Many people couldn't care less. But not all.
A Dyson is a VERY nice vacuum cleaner.
The decision to include 1 CPU and 2 GPUs, as against the traditional workstation model of 2 CPUs and 1 GPU is a daring move. But it's also a recognition that todays GPUs are really general purpose stream processors, that can do things that are nothing to do with graphics.
For example, searching through an array of data for a particular byte value: 16 CPU cores, with 64-bit wide registers can check 16 x 8 = 128 array slots per unit time. Two D700 GPUs, with 2048 32-bit stream processors each, can check 4096 x 4 = 16,384 array slots per unit time.
The first time you see this is real life - the first time you see a "gamer" GPU do something (nothing to do with graphics) 128 times faster than that Xeon you paid all that money for (this has happened to me) it's a paradigm shift in your mind.
I never even thought about the color of the logic boards.
It's like the hardware designer equivalent of wearing clean underwear.
I actually believe that GPUs are a very cost effective way to get bang for buck. But not the Quadro/FirePro. Nvidia GTX titan is around $1000 and performs at near Tesla levels. Or, of course, there are the actual Tesla boards. Problem is, if you configure your machine with two xeons and two GPU boards, you need a 1.2Kw power supply, and forget about the 11lb weight. Many people couldn't care less. But not all.
A Dyson is a VERY nice vacuum cleaner.
The 780s are also quite good. The biggest thing in favor of titan is probably the 6GB. If it wasn't for that I would say the 780s have a better price to performance ratio.
For example, searching through an array of data for a particular byte value: 16 CPU cores, with 64-bit wide registers can check 16 x 8 = 128 array slots per unit time. Two D700 GPUs, with 2048 32-bit stream processors each, can check 4096 x 4 = 16,384 array slots per unit time.
It's kind of like I said once before. They're great for really really parallel workloads. Interestingly there are some areas within graphics and visualization where that could really be exploited. I suspect it's an issue of existing code bases, older algorithms, and lack of a clear future in terms of what framework ends up being dominant. That being said, I just picked up a book on OpenCL and heterogeneous programming.
This is interesting. What does this tell us?
Mac Pro (Late 2013): Removing and installing flash storage - http://support.apple.com/kb/HT6061?viewlocale=en_US&locale=en_US
As others have noted, all that says is that the storage device can be removed by the user. It doesn't mean anyone, including Apple, is gonna sell you something else to put there.
While the move to solid-state storage is obviously a good one, I'm not yet sure whether the relatively small speed advantage likely to be realized by the PCIe interface offers enough benefit to offset the huge disadvantage of being incompatible with worldwide standards. Crucial and Samsung are constantly improving their storage offerings while also lowering prices. Apple computers don't reap those benefits. Is Apple's approach enough better to make it worth being a lone voice in the wilderness?
Good grief man! Get a grip. If Apple isn't going to do some serious forward-thinking, then who is? Otherwise we'd be stuck with floppies, user-removable batteries, physical keyboards and DVD's, to name just a few.
With due respect, different use cases make certain moves more or less desirable. For some, the benefits of certain changes are outweighed by the liabilities. For example, to the guy who spends 12 hours a day in the field, a user-replaceable battery matters more than a mm of thinness. That saved mm is gonna cost him DOUBLING the thickness of his device through the use of an external battery.
Enjoy the benefits of new approaches but don't wear blinders. Sometimes newer ain't necessarily better, or is better in some ways at the expense of being worse in others. By staying aware of which babies were thrown out with the bathwater, hopefully we can add them back in as technology advances (resulting in physical keyboards that can appear/disappear and change shape on demand, for example).
Apple knows who is buying what Mac Pros at a certain price point. [...] 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 don't know if that's true. We tend to give Apple an awful lot of credit regarding how well they know their customer base, but do they really? If so, how? I own a whole houseful of Apple products and manage a business that uses them, and Apple has never once asked me how I use them or how they could make my experience better.
All Apple knows for sure is which models sell and which don't, with no confirmed way of knowing WHY. For example, they might see that the 64GB version of the iPhone doesn't sell worth a damn and conclude that people don't need or want that much storage, when the reality is that buyers would LOVE to have 64GB or more but just can't justify or afford the extra $200! Apple has no better source of such knowledge than anyone else, and my own experience leads me to wonder if their market data is even as good as some of their competitors?
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 don't think the current design has nearly as much to do with market research as with what best satisfies the demands of Apple's software product, specifically FCPX. Just look at the test results so far: with Adobe Premiere, meh. With FCPX, wow! Coincidence?
That aside, I don't think anyone besides CS is saying that Apple should have gone with 2xCPU/1xGPU *instead* of 1xCPU/2xGPU, but could very easily have offered such configurations *as well.* I haven't bothered looking into it since it's not an option, but I suspect our needs here would probably be better served with more CPU per dollar and less GPU expense.
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.
I disagree. I think it's perfectly reasonable for the consumers of these items to give feedback to the supplier about what they want and need. THAT'S where Apple can get some genuine, useful market data.
My clients give me real-time feedback about what they want. Without that I'd have only my best guesses based on return customers on which to base my decision making. Obviously giving/getting feedback is better for both of us.
If Apple doesn't want to listen, that's their prerogative, but I don't see anything wrong with the buyer telling the vendor what they want.
They do offer a 3.7GHz quad core.
~e
While the move to solid-state storage is obviously a good one, I'm not yet sure whether the relatively small speed advantage likely to be realized by the PCIe interface offers enough benefit to offset the huge disadvantage of being incompatible with worldwide standards. Crucial and Samsung are constantly improving their storage offerings while also lowering prices. Apple computers don't reap those benefits. Is Apple's approach enough better to make it worth being a lone voice in the wilderness?
Why do you say "small" speed advantage?
http://www.macworld.com/article/2082022/the-new-mac-pro-first-impressions.html
"We also ran the popular Black Magic disk test on our review unit .. the result was a write speed of 952.8 MB/sec and a read speed of 920.5 MB/sec"
SATA 3 interfaces max out at about 550 MB/sec. I have the latest Samsung EVO SSD in my PC and *wish* it was using PCI connectivity.
deleted
Why do you say "small" speed advantage?
http://www.macworld.com/article/2082022/the-new-mac-pro-first-impressions.html
"We also ran the popular Black Magic disk test on our review unit .. the result was a write speed of 952.8 MB/sec and a read speed of 920.5 MB/sec"
SATA 3 interfaces max out at about 550 MB/sec. I have the latest Samsung EVO SSD in my PC and *wish* it was using PCI connectivity.
Oh. Okay, that's a bigger advantage than I thought. That makes the question even harder to answer!
Do speeds like that really happen in real life, or do other factors limit the rate at which the system can read and write? Would this configuration actually BE faster than a SATA3 interface or only have a "theoretical" advantage?
I'm confused - what kind of computing are you doing? From the memory requirements that you are quoting it sounds like large 3D numerical simulation, but that field has not been dominated by workstations for many years. We use clusters of Mac Pros for small problems to avoid the hassle of getting on the big machines, but not for anything serious. What kind of workstation setups are you referring to with 256 GB?
That is exactly what I was wondering as well. Almost all of the scientific computing work that I have done in the last two decades has been done on clusters. I still remember my first time using a Beowulf setup which cobbled together a bunch of older PC motherboards. Then again, I haven't used Mathematica/Maple for anything to this level. We pretty much make our own programs to reduce as much overhead as possible.
Interesting.
Oh. Okay, that's a bigger advantage than I thought. That makes the question even harder to answer!
Do speeds like that really happen in real life, or do other factors limit the rate at which the system can read and write? Would this configuration actually BE faster than a SATA3 interface or only have a "theoretical" advantage?
I think the theoretical limit is 1200MB/sec, so the 950MB/sec figure is what you can really expect. As for whether it would be noticeable in real life I guess that depends on the size of file you're working with. A 10GB video file might take 10s to load on a Mac Pro but 20s on a PC, but a 1MB Word file would be instantaneous on both.