So what. AIX and HPUX are essentially dead due to Linux. Solaris is half dead. And while linux and unix runs web servers and databases the enterprise runs on Exchange, Active Directory, Sharepoint and Win Server.
Trust me, AIX and HPUX are far from dead, especially on PowerPC based servers from IBM. Top end enterprise are more likely to use either Windows or a Combination of Windows with AIX or HPUX. Linux is more for the small and mid range market.
Really? OS X server can provide the same level of service as Active Directory? Nice. The mail server does everything that Exchange can do? Cool. At the enterprise level? Right.
If the department is small enough perhaps. At best Apple servers are really only SOHO and small business capable.
So what. Apple has a VERY easy to implement server for small businesses. Yeah, and it's a decent size market too.
Yeah, and Apple replacing cash registers with iPhones, iPads, iMacs or MacMinis. Do you know how many cash registers there are in the retail market? A #$%# LOAD. Nordstrom is replacing IBM cash registers slowly with iPhones. It's hysterical.
I can't imagine what the holdup is. The new processors are available. Thunderbolt is well developed. Better graphics cards are available. Everything is there. So release the new Mac Pro already. What, are they waiting for it to grow legs and deliver itself?
2013 for a new Mac Pro? What's this recent CPU update, a liver transplant? Lies, all lies.
The Mac Pro is dead. Tim just can't own up to it.
ASUS can put out several motherboard iterations in less than a year. Maybe Apple have a solitary school leaver working on the next gen Pro.
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
I can't imagine what the holdup is. The new processors are available. Thunderbolt is well developed. Better graphics cards are available. Everything is there. So release the new Mac Pro already. What, are they waiting for it to grow legs and deliver itself?
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
Yes, that's it. Don't actually answer my question. That's the way to be taken seriously.
Seriously, I don't want to be taken seriously. This is a Mac forum on the WWW. Just keep putting funny little animated icons at the end of your sentences and then the readership will know to take your comments seriously.
Meanwhile, all the components in PC land available to engineers exist today in sufficiency to make the new (old) Mac Pro truly a thing of the past.
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
It's not just the motherboard they have to worry about, it the cooling, the case may have to have some changes. NOW, here's why you shouldn't compare Apple to Dell, HP, Asus, etc.
Apple likes to work efficiently and they also like to be able to make a profit. ASUS has cheaper labor for designing motherboards. Apple has CHIP designers, which they MIGHT have to have design chips. Apple has a profit margin they would like to keep on their hardware products. ALL of the Windows PC mfg aren't making much money which is why THEY are not doing so well.
Apple doesn't want to end up being like these other PC mfg that spit out product every month unless the product line is getting a compelling upgrade and they can make a decent return on investment to cover all of the costs of R&D, Marketing, Advertisements, Support, etc. And Apple is not going to use some generic PC motherboard, YUCK.
As to the machine you outlined below, I was actually hoping to see such a box replace the Mac Pro this year. The technology is there to do so.
Not quite. Y'see by holding onto the architecture from 2010, it means that a single CPU 10-core 2013 Mac Pro will be faster than the last version. Apple only competes with themselves in performance.
If they brought out a single CPU Sandy Bridge, it would look bad vs a DP 2010 model.
They did this with the Core 2 architecture. PCs moved to Core-i and Apple stuck with Core 2 likely because of NVidia. Then, when the new models arrived, they did a 100% jump in a single generation and the MBA didn't look so bad, even with a ULV processor.
A 10-core Haswell should be 2.5x the performance of a 10-core 2010 Pro (if it existed obviously), Ivy Bridge would be 2x. Either one allows them to drop to a single CPU, cut the price and still be faster than the last generation. Sandy Bridge doesn't.
HP (the biggest desktop manufacturer in the world) will be relieved to hear that after narrowly ditching the entire market but they obviously made their choice to abandon it with very good reasoning.
Now regarding Intel, are you suggesting that dropping Ivy Bridge might actually allow them to push the Haswell version out next year while accounting for any hiccups in implementation.
I don't see why they can't skip Ivy Bridge Xeons. It's just a fab that they already have in production. They can start on the Haswell Xeon architecture design now alongside the consumer chips and target production late 2013. It solves everybody's problems. It seems nonsensical to me to permanently keep their best CPUs a full generation behind the consumer CPUs.
What we're so up in arms about is the priority (or lack thereof) of the Mac Pro. The video card that the brand new Mac Pro is shipping with is three years old. The "new" processors that we got yesterday were released by Intel over a year ago. The actually-new, Sandy Bridge E5 processors that Intel released (and the chipset that also allows faster RAM) was released in April. Other companies are shipping machines based on that architecture, but not Apple. I'm not TOO upset by the lack of USB 3 and Thunderbolt. Apple has never gone out of their way to add anything to the reference Intel motherboards, and the chipset that supports those functions isn't yet available.
The bottom line is that Apple updated this machine to early-to-mid 2011 specs with a 2009 video card -- in July 2012. What's not to be irate about?
I like the fact that you completely ignored his question, and thereby proved his point. Most people bitching about the Mac Pro don't even need one based on what they do, and are bitching for the hell of it. I know people who do nothing but watch porn and needless customization to their computers and call themselves 'power users'. His question is what the current Mac Pro can't do, that you need it to do.
I've been trying to put two and two together here and have this idea in my head now. Apple will be introducing a Mac Pro replacement that makes use of Knights Corner, one of Intels many core chips. The timing is right with Knight Corner rumored to go into production late this year.
Apparently Knights Corner is implemented as sort of a co processor. This means it could potentially be accessed with the OpenCL infrastructure Apple already has in place. If so applications developers could very quickly update apps to leverage the co processor. So a future Pro Mac might have a single Xeon socket along with a site for Knights Corner. In effect the Mac becomes a 50 core machine. "In effect" because of the co processor access scheme and the fact that Knights Corner processors aren't your traditional i86's.
Yeah this would be bleeding edge but it might explain the strange behavior around the current Mac Pro. I've searched around for other technology explanations and this is the best I can come up with. The approach would seem to mesh well with some of Apples preferred markets too.
Something very subtle has been happening at Apple since Tim took control. Apple seems to be more aware of the needs of the Pro and Business users -- especially pre-announcing strategic directions. This allows those whose paycheck/career/business depends on products and features and needs to plan and budget for their adoption.
This [partial] opening of the kimono, if done effectively, should not impact the secrecy benefit that Apple enjoys in the consumer market.
I would not be surprised to see large enterprises enter into co-development efforts with Apple (with appropriate NDAs).
I remember complaining to Guy Kawasaki (a few years before Steve left) that Apple didn't understand "businesses" and was missing a great opportunity. I always believed that Steve was the primary reason, and that he prided himself in it!
Tim, has come from the "other side" and understands how enterprise works and what they need -- and I expect he will engage that business without sacrificing the consumer needs.
Nothing against Steve -- Steve provided the platform (Apple) from which Tim can move the enterprise world.
I like it!
They may not want it to come back and bite them. I don't subscribe to the myths that ending the mac pro would necessarily kill Apple, but they do need to know that certain markets are covered. As an example their developers need machines that are conducive to OSX and iOS development. It doesn't matter what they look like or what they call it. It just matters that both the small and large developers are serviced reasonably well by hardware options, and this includes their staff members who develop applications in house. In the case of content creation, Macs have been popular for a very long time, and I do think that ultimately benefits Apple. Basically anything that helps bring cool stuff to iOS and OSX benefits Apple's long term corporate health.
Quote:
Originally Posted by Bloodshotrollin'red
2013 for a new Mac Pro? What's this recent CPU update, a liver transplant? Lies, all lies.
The Mac Pro is dead. Tim just can't own up to it.
ASUS can put out several motherboard iterations in less than a year. Maybe Apple have a solitary school leaver working on the next gen Pro.
They don't really do that with their workstation level hardware, but you're right. We should have seen a Sandy Bridge E announcement rather than this.
Quote:
Originally Posted by Marvin
Not quite. Y'see by holding onto the architecture from 2010, it means that a single CPU 10-core 2013 Mac Pro will be faster than the last version. Apple only competes with themselves in performance.
If they brought out a single CPU Sandy Bridge, it would look bad vs a DP 2010 model.
They did this with the Core 2 architecture. PCs moved to Core-i and Apple stuck with Core 2 likely because of NVidia. Then, when the new models arrived, they did a 100% jump in a single generation and the MBA didn't look so bad, even with a ULV processor.
A 10-core Haswell should be 2.5x the performance of a 10-core 2010 Pro (if it existed obviously), Ivy Bridge would be 2x. Either one allows them to drop to a single CPU, cut the price and still be faster than the last generation. Sandy Bridge doesn't.
HP (the biggest desktop manufacturer in the world) will be relieved to hear that after narrowly ditching the entire market but they obviously made their choice to abandon it with very good reasoning.
14nm is Broadwell, the die-shrink of Haswell. 14nm will be in development in 2013, deploying 2014:
I don't see why they can't skip Ivy Bridge Xeons. It's just a fab that they already have in production. They can start on the Haswell Xeon architecture design now alongside the consumer chips and target production late 2013. It solves everybody's problems. It seems nonsensical to me to permanently keep their best CPUs a full generation behind the consumer CPUs.
If they were dropping to single package only, they could keep the dual ones on the older process, but it would make more sense to just update both. With workstations board designs tend to be a two cycle thing. Ivy Bridge isn't likely to see much in the way of board revisions beyond minor tweaks compared to a Sandy Bridge E reference design. It's not just Apple. The other oems do the same thing both for cost and stability reasons on workstation products. In the case of the core2 products, that was different. Apple was using NVidia chipsets for gpu reasons, and they stayed on Core2 on some laptops and the mini for that reason. It's weird doing that on the mac pro, especially when it requires feature updates such as usb3. I guess Intel could skip Ivy, but the rumors do not presently suggest this. Workstation and server implementations are more complex in some ways given the tight tolerance. Like I was saying, stepping issues + sata implementation pushed it way back this last time. They could focus solely on tock cycles for their workstation lines. This would theoretically help with the scheduling problems, but I don't know if we'll actually see the big iron Haswell Xeons prior to 2014. Right now the rumor is that Ivy Bridge E will surface mid next year. If Apple is waiting for that, this means you'd have some of the same nehalem cpus in the machine (the quads) over 4 years later, and Nehalem/Westmere would be in the mac pros when the consumer lines are pushing haswell. All signs so far suggest that Haswell consumer cpus will remain capped at quad core models. Intel is obviously allocating as much space as possible to the gpu rather than cpu cores at the consumer end.
Quote:
Originally Posted by wizard69
I've been trying to put two and two together here and have this idea in my head now. Apple will be introducing a Mac Pro replacement that makes use of Knights Corner, one of Intels many core chips. The timing is right with Knight Corner rumored to go into production late this year.
Apparently Knights Corner is implemented as sort of a co processor. This means it could potentially be accessed with the OpenCL infrastructure Apple already has in place. If so applications developers could very quickly update apps to leverage the co processor. So a future Pro Mac might have a single Xeon socket along with a site for Knights Corner. In effect the Mac becomes a 50 core machine. "In effect" because of the co processor access scheme and the fact that Knights Corner processors aren't your traditional i86's.
Yeah this would be bleeding edge but it might explain the strange behavior around the current Mac Pro. I've searched around for other technology explanations and this is the best I can come up with. The approach would seem to mesh well with some of Apples preferred markets too.
Apple is adventurous when it carries a coolness factor. Other times (like usb3) they tend to be a little conservative. I wonder how well it would mesh with programming. We haven't seen much in the way of N-core scaling at this point.
My 2006 Mac Pro will not run Boot Camp and the RAM isn't the fastest. It's still running fine, though and will hold me until next year's models since I put in a Radeon 5770. I am hoping SSD drives will be more reasonable next year too. One thing about Mac Pros - they LAST.
I've been trying to put two and two together here and have this idea in my head now. Apple will be introducing a Mac Pro replacement that makes use of Knights Corner, one of Intels many core chips. The timing is right with Knight Corner rumored to go into production late this year.
I agree, I figured it was a possibility for WWDC but someone said they won't be making them in volume. I found this article:
This supercomputer is due late 2012 and will have 3200 machines. $2.5m is to be spent on Knights Corner for 8 petaflops of DP computation. That would mean 8000 Knights Corner chips, between 2 and 3 per machine. This means $312 for each individual Knights Corner chip.
A 6-core E5 processor will be around 150GFLOPs DP. Knights Corner would be about 6.5x faster.
The look on the faces of old-school Mac Pro buyers persuaded to buy a Thunderbolt-chainable $2999 toy box with over 2 TFLOPs of double precision computation - priceless
It may not be the Mac Pro as we know it now, but knowing they are going to continue a professional line of desktop machines makes me happy. In 2013 I was actually going to sell my 2010 iMac (Apple Care would've run out by that point) and get a Pro tower. Nice to know I can still do that. Weather its a "typical" tower, so to speak, remains to be seen.
What makes you think that Apple's Mac Pro replacement will meet your needs?
Apple thinks most people do not need optical disks, spinning hard drives, wired ethernet, nor internal expansion.
My guess is that the Mac Pro replacement will be a small box (like the mini) with a fast processor, a few extra USB 3 ports, and perhaps an extra Thunderbolt port or two. If we are lucky, the internal RAM and SSD won't be soldered and/or glued in place. Of course, it will have a new form factor (perhaps machined aluminum) and look really cool.
Such a machine would serve the needs of 99% of the high end market. Apple doesn't need the other 1%.
Remember, Apple doesn't build products to meet the desires of their customers. They build their products, and leave their customers little choice but to move to the new technology.
A Windows computer bought today can run DOS software developed in 1990.
A Mac bought today won't even run Adobe software from early 2007.
I am not saying the Apple's policies are good or bad. I am just pointing out that this is how Apple behaves, it has been a very profitable strategy, and it is a strategy that has made the vast majority of their customers very happy.
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
I meant to respond to this before. They've had three years on this since the last revision. It doesn't surprise me that they left the design alone. What surprises me is that they didn't work with Foxconn on an X79 board revision. These ones could have been the same thing with up to date internals. It's not that these are no longer powerful. It's more a case that when it comes to dropping that much on a new machine, I'd like something that's fully up to date rather than having to take a lot of upgrade options to compensate for the use of an older hardware generation. I was hoping for one of the cheaper E5 hex cores (can't remember the model, but they retail for roughly $600) within a $3k package + 7XXX gpu options rather than a 5770 and 5870.
Quote:
Originally Posted by Marvin
I agree, I figured it was a possibility for WWDC but someone said they won't be making them in volume. I found this article:
This supercomputer is due late 2012 and will have 3200 machines. $2.5m is to be spent on Knights Corner for 8 petaflops of DP computation. That would mean 8000 Knights Corner chips, between 2 and 3 per machine. This means $312 for each individual Knights Corner chip.
A 6-core E5 processor will be around 150GFLOPs DP. Knights Corner would be about 6.5x faster.
8" Cube
Xeon E5-2630 (150GFLOPs) - $612
Knight Corner (1000GFLOPs) - $312
AMD 7970 (947GFLOPs) - $400
The look on the faces of old-school Mac Pro buyers persuaded to buy a Thunderbolt-chainable $2999 toy box with over 2 TFLOPs of double precision computation - priceless
I think you misunderstand the behavior. You read about smudgy screens on imacs or see problems with logic board failures on anorexic machines. You see ports missing. Of course it's easier to grab a mac pro. If Apple designs with function and efficiency prioritized over aesthetics, the whining will subside. People just want something that can work well with their existing needs. You don't just look at the base machine. If you have any specific requirements that are serviced by means of PCI hardware or other things, they're all examined prior to a major purchase so that $4k doesn't become $7k after the fact. Can you really blame anyone for being skeptical? Also isn't Knights Corner kind of a response to NVidia's tesla cards? I will have to read more about them, but general computing is a bit different from HPCs running highly customized code. I don't know how one integrates with the other or how long such an infiltration would take if it really does trickle into the workstation market. Bleh... now I want to find more articles. Damn you Marvin.
Comments
Quote:
Originally Posted by nht
So what. AIX and HPUX are essentially dead due to Linux. Solaris is half dead. And while linux and unix runs web servers and databases the enterprise runs on Exchange, Active Directory, Sharepoint and Win Server.
Trust me, AIX and HPUX are far from dead, especially on PowerPC based servers from IBM. Top end enterprise are more likely to use either Windows or a Combination of Windows with AIX or HPUX. Linux is more for the small and mid range market.
Quote:
Originally Posted by nht
Really? OS X server can provide the same level of service as Active Directory? Nice. The mail server does everything that Exchange can do? Cool. At the enterprise level? Right.
If the department is small enough perhaps. At best Apple servers are really only SOHO and small business capable.
So what. Apple has a VERY easy to implement server for small businesses. Yeah, and it's a decent size market too.
Yeah, and Apple replacing cash registers with iPhones, iPads, iMacs or MacMinis. Do you know how many cash registers there are in the retail market? A #$%# LOAD. Nordstrom is replacing IBM cash registers slowly with iPhones. It's hysterical.
2013 for a new Mac Pro? What's this recent CPU update, a liver transplant? Lies, all lies.
The Mac Pro is dead. Tim just can't own up to it.
ASUS can put out several motherboard iterations in less than a year. Maybe Apple have a solitary school leaver working on the next gen Pro.
I can't imagine what the holdup is. The new processors are available. Thunderbolt is well developed. Better graphics cards are available. Everything is there. So release the new Mac Pro already. What, are they waiting for it to grow legs and deliver itself?
Having what to do with the Mac Pro?
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
Quote:
Originally Posted by Tallest Skil
Having what to do with the Mac Pro?
Given this is a thread about the next Mac Pro my comment obviously has absolutely nothing to do with the topic. I was referring to faeries.
Yes, that's it. Don't actually answer my question. That's the way to be taken seriously.
Quote:
Originally Posted by Tallest Skil
Yes, that's it. Don't actually answer my question. That's the way to be taken seriously.
Seriously, I don't want to be taken seriously. This is a Mac forum on the WWW. Just keep putting funny little animated icons at the end of your sentences and then the readership will know to take your comments seriously.
Meanwhile, all the components in PC land available to engineers exist today in sufficiency to make the new (old) Mac Pro truly a thing of the past.
Quote:
Originally Posted by jragosta
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
It's not just the motherboard they have to worry about, it the cooling, the case may have to have some changes. NOW, here's why you shouldn't compare Apple to Dell, HP, Asus, etc.
Apple likes to work efficiently and they also like to be able to make a profit. ASUS has cheaper labor for designing motherboards. Apple has CHIP designers, which they MIGHT have to have design chips. Apple has a profit margin they would like to keep on their hardware products. ALL of the Windows PC mfg aren't making much money which is why THEY are not doing so well.
Apple doesn't want to end up being like these other PC mfg that spit out product every month unless the product line is getting a compelling upgrade and they can make a decent return on investment to cover all of the costs of R&D, Marketing, Advertisements, Support, etc. And Apple is not going to use some generic PC motherboard, YUCK.
Not quite. Y'see by holding onto the architecture from 2010, it means that a single CPU 10-core 2013 Mac Pro will be faster than the last version. Apple only competes with themselves in performance.
If they brought out a single CPU Sandy Bridge, it would look bad vs a DP 2010 model.
They did this with the Core 2 architecture. PCs moved to Core-i and Apple stuck with Core 2 likely because of NVidia. Then, when the new models arrived, they did a 100% jump in a single generation and the MBA didn't look so bad, even with a ULV processor.
A 10-core Haswell should be 2.5x the performance of a 10-core 2010 Pro (if it existed obviously), Ivy Bridge would be 2x. Either one allows them to drop to a single CPU, cut the price and still be faster than the last generation. Sandy Bridge doesn't.
HP (the biggest desktop manufacturer in the world) will be relieved to hear that after narrowly ditching the entire market but they obviously made their choice to abandon it with very good reasoning.
14nm is Broadwell, the die-shrink of Haswell. 14nm will be in development in 2013, deploying 2014:
http://www.engadget.com/2012/05/14/intel-sets-sights-on-5nm-chip/
2013 will still be 22nm Haswell.
I don't see why they can't skip Ivy Bridge Xeons. It's just a fab that they already have in production. They can start on the Haswell Xeon architecture design now alongside the consumer chips and target production late 2013. It solves everybody's problems. It seems nonsensical to me to permanently keep their best CPUs a full generation behind the consumer CPUs.
Quote:
Originally Posted by Maltz
What we're so up in arms about is the priority (or lack thereof) of the Mac Pro. The video card that the brand new Mac Pro is shipping with is three years old. The "new" processors that we got yesterday were released by Intel over a year ago. The actually-new, Sandy Bridge E5 processors that Intel released (and the chipset that also allows faster RAM) was released in April. Other companies are shipping machines based on that architecture, but not Apple. I'm not TOO upset by the lack of USB 3 and Thunderbolt. Apple has never gone out of their way to add anything to the reference Intel motherboards, and the chipset that supports those functions isn't yet available.
The bottom line is that Apple updated this machine to early-to-mid 2011 specs with a 2009 video card -- in July 2012. What's not to be irate about?
I like the fact that you completely ignored his question, and thereby proved his point. Most people bitching about the Mac Pro don't even need one based on what they do, and are bitching for the hell of it. I know people who do nothing but watch porn and needless customization to their computers and call themselves 'power users'. His question is what the current Mac Pro can't do, that you need it to do.
Apparently Knights Corner is implemented as sort of a co processor. This means it could potentially be accessed with the OpenCL infrastructure Apple already has in place. If so applications developers could very quickly update apps to leverage the co processor. So a future Pro Mac might have a single Xeon socket along with a site for Knights Corner. In effect the Mac becomes a 50 core machine. "In effect" because of the co processor access scheme and the fact that Knights Corner processors aren't your traditional i86's.
Yeah this would be bleeding edge but it might explain the strange behavior around the current Mac Pro. I've searched around for other technology explanations and this is the best I can come up with. The approach would seem to mesh well with some of Apples preferred markets too.
Quote:
Originally Posted by Dick Applebaum
Something very subtle has been happening at Apple since Tim took control. Apple seems to be more aware of the needs of the Pro and Business users -- especially pre-announcing strategic directions. This allows those whose paycheck/career/business depends on products and features and needs to plan and budget for their adoption.
This [partial] opening of the kimono, if done effectively, should not impact the secrecy benefit that Apple enjoys in the consumer market.
I would not be surprised to see large enterprises enter into co-development efforts with Apple (with appropriate NDAs).
I remember complaining to Guy Kawasaki (a few years before Steve left) that Apple didn't understand "businesses" and was missing a great opportunity. I always believed that Steve was the primary reason, and that he prided himself in it!
Tim, has come from the "other side" and understands how enterprise works and what they need -- and I expect he will engage that business without sacrificing the consumer needs.
Nothing against Steve -- Steve provided the platform (Apple) from which Tim can move the enterprise world.
I like it!
They may not want it to come back and bite them. I don't subscribe to the myths that ending the mac pro would necessarily kill Apple, but they do need to know that certain markets are covered. As an example their developers need machines that are conducive to OSX and iOS development. It doesn't matter what they look like or what they call it. It just matters that both the small and large developers are serviced reasonably well by hardware options, and this includes their staff members who develop applications in house. In the case of content creation, Macs have been popular for a very long time, and I do think that ultimately benefits Apple. Basically anything that helps bring cool stuff to iOS and OSX benefits Apple's long term corporate health.
Quote:
Originally Posted by Bloodshotrollin'red
2013 for a new Mac Pro? What's this recent CPU update, a liver transplant? Lies, all lies.
The Mac Pro is dead. Tim just can't own up to it.
ASUS can put out several motherboard iterations in less than a year. Maybe Apple have a solitary school leaver working on the next gen Pro.
They don't really do that with their workstation level hardware, but you're right. We should have seen a Sandy Bridge E announcement rather than this.
Quote:
Originally Posted by Marvin
Not quite. Y'see by holding onto the architecture from 2010, it means that a single CPU 10-core 2013 Mac Pro will be faster than the last version. Apple only competes with themselves in performance.
If they brought out a single CPU Sandy Bridge, it would look bad vs a DP 2010 model.
They did this with the Core 2 architecture. PCs moved to Core-i and Apple stuck with Core 2 likely because of NVidia. Then, when the new models arrived, they did a 100% jump in a single generation and the MBA didn't look so bad, even with a ULV processor.
A 10-core Haswell should be 2.5x the performance of a 10-core 2010 Pro (if it existed obviously), Ivy Bridge would be 2x. Either one allows them to drop to a single CPU, cut the price and still be faster than the last generation. Sandy Bridge doesn't.
HP (the biggest desktop manufacturer in the world) will be relieved to hear that after narrowly ditching the entire market but they obviously made their choice to abandon it with very good reasoning.
14nm is Broadwell, the die-shrink of Haswell. 14nm will be in development in 2013, deploying 2014:
http://www.engadget.com/2012/05/14/intel-sets-sights-on-5nm-chip/
2013 will still be 22nm Haswell.
I don't see why they can't skip Ivy Bridge Xeons. It's just a fab that they already have in production. They can start on the Haswell Xeon architecture design now alongside the consumer chips and target production late 2013. It solves everybody's problems. It seems nonsensical to me to permanently keep their best CPUs a full generation behind the consumer CPUs.
If they were dropping to single package only, they could keep the dual ones on the older process, but it would make more sense to just update both. With workstations board designs tend to be a two cycle thing. Ivy Bridge isn't likely to see much in the way of board revisions beyond minor tweaks compared to a Sandy Bridge E reference design. It's not just Apple. The other oems do the same thing both for cost and stability reasons on workstation products. In the case of the core2 products, that was different. Apple was using NVidia chipsets for gpu reasons, and they stayed on Core2 on some laptops and the mini for that reason. It's weird doing that on the mac pro, especially when it requires feature updates such as usb3. I guess Intel could skip Ivy, but the rumors do not presently suggest this. Workstation and server implementations are more complex in some ways given the tight tolerance. Like I was saying, stepping issues + sata implementation pushed it way back this last time. They could focus solely on tock cycles for their workstation lines. This would theoretically help with the scheduling problems, but I don't know if we'll actually see the big iron Haswell Xeons prior to 2014. Right now the rumor is that Ivy Bridge E will surface mid next year. If Apple is waiting for that, this means you'd have some of the same nehalem cpus in the machine (the quads) over 4 years later, and Nehalem/Westmere would be in the mac pros when the consumer lines are pushing haswell. All signs so far suggest that Haswell consumer cpus will remain capped at quad core models. Intel is obviously allocating as much space as possible to the gpu rather than cpu cores at the consumer end.
Quote:
Originally Posted by wizard69
I've been trying to put two and two together here and have this idea in my head now. Apple will be introducing a Mac Pro replacement that makes use of Knights Corner, one of Intels many core chips. The timing is right with Knight Corner rumored to go into production late this year.
Apparently Knights Corner is implemented as sort of a co processor. This means it could potentially be accessed with the OpenCL infrastructure Apple already has in place. If so applications developers could very quickly update apps to leverage the co processor. So a future Pro Mac might have a single Xeon socket along with a site for Knights Corner. In effect the Mac becomes a 50 core machine. "In effect" because of the co processor access scheme and the fact that Knights Corner processors aren't your traditional i86's.
Yeah this would be bleeding edge but it might explain the strange behavior around the current Mac Pro. I've searched around for other technology explanations and this is the best I can come up with. The approach would seem to mesh well with some of Apples preferred markets too.
Apple is adventurous when it carries a coolness factor. Other times (like usb3) they tend to be a little conservative. I wonder how well it would mesh with programming. We haven't seen much in the way of N-core scaling at this point.
My 2006 Mac Pro will not run Boot Camp and the RAM isn't the fastest. It's still running fine, though and will hold me until next year's models since I put in a Radeon 5770. I am hoping SSD drives will be more reasonable next year too. One thing about Mac Pros - they LAST.
Sure it will. Are you getting error messages?
I agree, I figured it was a possibility for WWDC but someone said they won't be making them in volume. I found this article:
http://www.hpcwire.com/hpcwire/2011-09-22/dell_to_build_10-petaflop_supercomputer_for_science.html
This supercomputer is due late 2012 and will have 3200 machines. $2.5m is to be spent on Knights Corner for 8 petaflops of DP computation. That would mean 8000 Knights Corner chips, between 2 and 3 per machine. This means $312 for each individual Knights Corner chip.
A 6-core E5 processor will be around 150GFLOPs DP. Knights Corner would be about 6.5x faster.
8" Cube
Xeon E5-2630 (150GFLOPs) - $612
Knight Corner (1000GFLOPs) - $312
AMD 7970 (947GFLOPs) - $400
The look on the faces of old-school Mac Pro buyers persuaded to buy a Thunderbolt-chainable $2999 toy box with over 2 TFLOPs of double precision computation - priceless
Quote:
Originally Posted by benanderson89
Oh thank christ!
It may not be the Mac Pro as we know it now, but knowing they are going to continue a professional line of desktop machines makes me happy. In 2013 I was actually going to sell my 2010 iMac (Apple Care would've run out by that point) and get a Pro tower. Nice to know I can still do that. Weather its a "typical" tower, so to speak, remains to be seen.
What makes you think that Apple's Mac Pro replacement will meet your needs?
Apple thinks most people do not need optical disks, spinning hard drives, wired ethernet, nor internal expansion.
My guess is that the Mac Pro replacement will be a small box (like the mini) with a fast processor, a few extra USB 3 ports, and perhaps an extra Thunderbolt port or two. If we are lucky, the internal RAM and SSD won't be soldered and/or glued in place. Of course, it will have a new form factor (perhaps machined aluminum) and look really cool.
Such a machine would serve the needs of 99% of the high end market. Apple doesn't need the other 1%.
Remember, Apple doesn't build products to meet the desires of their customers. They build their products, and leave their customers little choice but to move to the new technology.
A Windows computer bought today can run DOS software developed in 1990.
A Mac bought today won't even run Adobe software from early 2007.
I am not saying the Apple's policies are good or bad. I am just pointing out that this is how Apple behaves, it has been a very profitable strategy, and it is a strategy that has made the vast majority of their customers very happy.
Quote:
Originally Posted by jragosta
I hope you have some pretty strong evidence since you're accusing Cook of a felony - lying about a material matter.
Even with as much money as Apple has, resources are not unlimited. The critical factor is probably availability of skilled personnel. While hardware and software engineers are a dime a dozen, only a small percentage would fit into Apple's culture and have the skills to contribute at the desired level. In addition, Jony Ive only has so many hours per day.
I meant to respond to this before. They've had three years on this since the last revision. It doesn't surprise me that they left the design alone. What surprises me is that they didn't work with Foxconn on an X79 board revision. These ones could have been the same thing with up to date internals. It's not that these are no longer powerful. It's more a case that when it comes to dropping that much on a new machine, I'd like something that's fully up to date rather than having to take a lot of upgrade options to compensate for the use of an older hardware generation. I was hoping for one of the cheaper E5 hex cores (can't remember the model, but they retail for roughly $600) within a $3k package + 7XXX gpu options rather than a 5770 and 5870.
Quote:
Originally Posted by Marvin
I agree, I figured it was a possibility for WWDC but someone said they won't be making them in volume. I found this article:
http://www.hpcwire.com/hpcwire/2011-09-22/dell_to_build_10-petaflop_supercomputer_for_science.html
This supercomputer is due late 2012 and will have 3200 machines. $2.5m is to be spent on Knights Corner for 8 petaflops of DP computation. That would mean 8000 Knights Corner chips, between 2 and 3 per machine. This means $312 for each individual Knights Corner chip.
A 6-core E5 processor will be around 150GFLOPs DP. Knights Corner would be about 6.5x faster.
8" Cube
Xeon E5-2630 (150GFLOPs) - $612
Knight Corner (1000GFLOPs) - $312
AMD 7970 (947GFLOPs) - $400
The look on the faces of old-school Mac Pro buyers persuaded to buy a Thunderbolt-chainable $2999 toy box with over 2 TFLOPs of double precision computation - priceless
I think you misunderstand the behavior. You read about smudgy screens on imacs or see problems with logic board failures on anorexic machines. You see ports missing. Of course it's easier to grab a mac pro. If Apple designs with function and efficiency prioritized over aesthetics, the whining will subside. People just want something that can work well with their existing needs. You don't just look at the base machine. If you have any specific requirements that are serviced by means of PCI hardware or other things, they're all examined prior to a major purchase so that $4k doesn't become $7k after the fact. Can you really blame anyone for being skeptical? Also isn't Knights Corner kind of a response to NVidia's tesla cards? I will have to read more about them, but general computing is a bit different from HPCs running highly customized code. I don't know how one integrates with the other or how long such an infiltration would take if it really does trickle into the workstation market. Bleh... now I want to find more articles. Damn you Marvin.