I think they should use the latter desktop grade CPUs. The motherboard will have to be redesigned from scratch anyway. With the past AL iMacs, the Core2 boards were tweaked and modified to accommodate different, but similar components. Now is a good time to reinvent the iMac.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
That's a good point. But if Apple moves the iMac and mini (or an incarnation of the mini) to AIO type CPUs then they can still leverage the cost savings. I think the MacBook and MBP together can get guarantee the same discount the currently enjoy. Plus I'm assuming a huge cost difference between mobile versions and AIO versions that they can make up any price differences via improved margins on the mini and iMac (provided they don't lower the prices dramatically).
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
They could do that if they wanted to, but that is not a feasible solution. Liquid cooling adds cost and a higher potential for problems. There was a patent they put out awhile back for drawing heat away from the components but there has been plenty of time that their engineers may have simply found a better solution of component placement to allow for a a higher wattage CPU. There is also the shrinking of certain chips, a more refined MoBo and maybe even a smaller power supply that would allow for additional space in the system for another fan. Even that may be enough; we really don’t know, but I think we can deduce that Apple’s focus with the iMac is not to make it the fastest machine possible (they have no AIO competition) but to make it appealing to their target market, which isn’t me and likely isn’t many on this forum.
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
There's not enough space for that type of solution under the current case constraints and Apple's had so many issues with liquid cooling (G5 tower) that I don't see them going down that road again.
There's not enough space for that type of solution under the current case constraints and Apple's had so many issues with liquid cooling (G5 tower) that I don't see them going down that road again.
Sorry about that. G5 is before my time. I'm only a recent convert. I didn't know they had already taken a whack at water cooled solutions.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
I'm not convinced that the difference in performance are all that little. So given that what many of us want Apple to do is to address the performance gap between the Mac Pro and the iMac. It doesn't have to be on every iMac but there needs to be a high end model that truly delivers on the performance problem.
Admittedly this is not really what the iMac is designed for and an XMac would be a better product to fill that gap. The problem is many of us are resigned to never seeing an XMac or any thing close to that. A good compromise here would be a high end Mini, which would require a taller cased Mini and a more robust power supply. Even so I can see Apple refactoring the Mini into a power house box, all they need to do is reconsider a few design elements.
That's not the point. He's implying that SL and Grand Central are really only beneficial for quad-core or higher processors and that it's wasted on the current dual-core implementation all of the Mac's except of course for the quad and dual-quad Mac Pro's. I can only assume this argument is to try to reinforce the idea that iMac needs a quad-core to properly utilize SL, which is untrue. Any multi-core processor will benefit from SL's Grand Central, including a dual-core processor.
There is no doubt in my mind that to truly leverage GCD you need more than two hardware threads. The more the better in fact and that is one of the reasons that the new Intel hardware is so important.
Now that is not saying that SL running on a dual core system won't benefit from GCD just that the benefit is not going to be as large as it could be. Part of the problem here is that the common usage patterns have at least one core sporadically active in background, it is not like the cores are always free to process GCD work units. Combine that with the OpenMP like calls that would effectively use as many threads as can be managed by the OS and you can see that a minimal system for SL ought to be built around a processor with at least four cores.
Quote:
Of course one would expect qaud-core or dual-quad core's to be faster than a dual-core alone. That was never in question. I actually look forward to quad-core for the entire line, but passing on disinformation, even if it's just implied, doesn't help.
Nor does your statements here. It implies a misunderstanding of the nature of GCD and what it tries to do. Sure it will work on a two core system, that can not be debated as I'm running SL right this minute on my old MBP. However there is no way in hell that I'm going to buy anything new from Apple with less than four cores, it just doesn't make sense.
I expect to see next year, dual core Arrandale with integrated graphics for the mini and Clarksfield for the iMac with discrete graphics. Hopefully ATI mobile 4800 series class.
Title says it all! The iMac is simply grossly over priced for what you get, it needs to be updated now. Lets face it if Apple can't get Clarksfield into iMac in a couple of months they will have screwed up royally.
There is no doubt in my mind that to truly leverage GCD you need more than two hardware threads. The more the better in fact and that is one of the reasons that the new Intel hardware is so important.
Now that is not saying that SL running on a dual core system won't benefit from GCD just that the benefit is not going to be as large as it could be. Part of the problem here is that the common usage patterns have at least one core sporadically active in background, it is not like the cores are always free to process GCD work units. Combine that with the OpenMP like calls that would effectively use as many threads as can be managed by the OS and you can see that a minimal system for SL ought to be built around a processor with at least four cores.
Nor does your statements here. It implies a misunderstanding of the nature of GCD and what it tries to do. Sure it will work on a two core system, that can not be debated as I'm running SL right this minute on my old MBP. However there is no way in hell that I'm going to buy anything new from Apple with less than four cores, it just doesn't make sense.
Dave
And I will ask again, is there anything, other than your belief (as in links to real benchmarks) showing an improved scalability for quad or dual-quad cores under SL other than increases that one would expect from additional cores?
At the most basic level and assuming GDC is 100% efficient, a dual-core could process a task in half the time that a single core could.
Again at 100% efficiency, a quad core could process the same task in 25% of the time that a single core could.
You're implying that simply adding more cores will somehow make GDC more efficient. The GCD will simply work with what it has, be that 2 cores, 4 cores, or 8 cores, or whatever. It won't suddenly get more efficient with more cores, It will work within that efficiency regardless of the number of cores, and we know that it isn't 100% efficient (it will never be able to dole out all work perfectly across X number of cores in any case).
It seems that several people are concerned about the possibility of a Quad-core chip skipping the new iMac release, and they are comparing the iMac to some randomly beefed-up Windows PCs.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
Quote:
When Apple creates the hardware and software, and, as put in other posts, the "user experience" as well, it creates a harmonious, synergistic machine that is not only accelerated by the speed of your chip.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
Quote:
If anyone's forgotten, Snow Leopard is (as I think I remember reading) something like up to 2.5x faster than Leopard when performing certain tasks, which clues in to the importance of a powerful operating system. It doesn't matter if you had a 5 GHZ Windows machine; if the software bites and the components do not work well together like in a Mac, than your extra specs are virtually worthless.
SL is relatively faster, excluding a few regressions, it however isn't anywhere near 2.5X on average. The problem that you don't seem to grasp is that the software will bite in the future trying to run on a dual core machine. Read up a bit on how GCD works and you will see how silly your position is.
Quote:
Aside, Apple has always sought to shave thickness from all of its products, and so I don't see this as an exception. It would also be nice if an anti-glare option similar to that on the MBP was available in the next iMac for those of us who need accurate color and less eye strain.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
Title says it all! The iMac is simply grossly over priced for what you get, it needs to be updated now. Lets face it if Apple can't get Clarksfield into iMac in a couple of months they will have screwed up royally.
Dave
Hardly. It's competitively priced for the hardware in it. You'll find the same prices on any hardware manufacturer's site and an iMac includes a 20" or a 24" monitor. A decent quality one at that.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
SL is relatively faster, excluding a few regressions, it however isn't anywhere near 2.5X on average. The problem that you don't seem to grasp is that the software will bite in the future trying to run on a dual core machine. Read up a bit on how GCD works and you will see how silly your position is.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
dave
Agree. If it's do-able in a laptop, then I don't doubt that apple could pull it off in a much larger iMac case:
I'm genuinely curious where this information is that SL will only really benefit quad core? This is the first I'm reading of it. Could you post a link I'd be interested in reading the technical details. From what I've ready, any multi-core processor will benefit, not just quad-core or an 8-core.
All I can say if you don't believe what is written here read the SL and GCD documentation. Nobody is saying that SL won't work on dual core hardware, all we are saying is that to truly benefit from the tech you need as many cores as is practical. Today that sweet spot is with Quad core hardware.
In a nutshell SL with GCD manages a system wide thread pool. The more threads that can be processed by real cores the faster GCD enhanced software will be. Now a dual core processor can just as easily process threads from that pool but the reality is only two will ever get processed in parallel. A four core machine doubles that and if SMT is available doubles(somewhat) that again.
Now the bad part here is that this implies software that is written to leverage GCD or at a higher level NS"Operation" calls. Some software will likely never be able to leverage those features. A key realization here though is that software that does leverage the new features won't need to know how many cores you have nor make arbitrary decisions about how many threads to use. GCD handles that for the programmer and automatically uses as many threads as is rational for the work taking place. Thus software written to leverage SL & GCD will become faster simply by running on a machine with more cores.
It is important to realize that this doesn't necessarily happen with multithreaded apps written yesterday using older threading models. In the past programmers had to make assumptions about how many cores the threaded app was likely to have or in some cases the ability of a single core to handle a threaded app. For the most part that is gone now.
So I hope you understand it is not an issue of SL not leveraging dual core machines. The issue is that it is a stupid and unnecessary limitation on Apples part. Frankly Im to the point where I have to question Apples ethics to continue to market dual core midrange machines with SL installed. Especially when cost effective quad cores have been around for so long.
As you pointed out, the TDP for the new Clarksfield chips are 45W and 55W. However, note that the power dissipation at the system level will be actually less than a Core 2 system since you don't have a hot northbridge chip with wide memory buses anymore. 45W TDP for Clarksfield includes the memory controller. TDP for the PM55 chip is only 3.5W!
As for the perceived "marketing" problem you mention below, Apple is probably the one company who can deal most effectively with this issue. One simple solution is to add the the peak (Turbo boost) clock rate on the data sheet in addition to the base clock speed. There are some interesting performance comparisons vs a quad-core Core2 system in Anand's test reference above. Also note that Clarksfield clock rate can be made to match or exceed Core 2 clock speeds when only two cores of the four cores are under high load in a Clarksfield system.
Quote:
Originally Posted by solipsism
Sure, they CAN, but at what cost? The iMac already runs pretty hot with a 44W CPU so I don’t see how a 130W CPU would be possible without Apple making it thicker and/or having a major cooling breakthrough. It’s not just the CPU that now needs to be considered, but larger fans, larger power supply, etc. We know Apple doesn’t like to go thicker—even though they CAN if they wanted to—so that option shouldn’t be considered as it’s not part of of their core philosophy. McDonald’s could make a steak burger using Kobe beef if they wanted to, but a $20 burger isn’t part of their business model.
Remember that Apple’s desktop line is growing faster than the industry. I doubt it’s because of the Mac Pro or the Mac Mini so I have to think it’s because of the iMac. Apparently Apple is doing something right. Even their notebooks are underpowered when you look at the larger notebooks from other vendors using cheaper components that run hotter than Apple’s notebooks, yet Apple is still selling more and dominating the high end market.
PS: Check out the last section of the link below regarding Clarksfield. These are the mobile i7 chips that using 45W and 55W. But now there is a marketing problem with these chips as their clock speed is significantly lower than the previous dual-cores. Most cnsumer use clocks peed as a baseline for determining value, they don’t read charts to determine overall performance. Eventually we’ll see quad-cores in everything, but it’s s tricky time for PC vendors.
So I hope you understand it is not an issue of SL not leveraging dual core machines. The issue is that it is a stupid and unnecessary limitation on Apples part. Frankly Im to the point where I have to question Apples ethics to continue to market dual core midrange machines with SL installed. Especially when cost effective quad cores have been around for so long.
Dave
You'll get no argument there, although I don't think it was 'stupid' of them to not include quad in an iMac as it isn't intended to compete directly with the Mac Pro.
It sounds like a lot of folks are really wanting something between the iMac and the Mac Pro, or at a minimum, an iMac Pro of sorts. I would certainly be interested in an iMac Pro if one was available, but I'm not interested in paying a premium price as we already see with the Mac Pro.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
Ask a plethora of perfessionals why they are upset about the glossed-over mirror that has replaced the standard matte screen, and you will hear that the colors are totally off. Why do you think apple added the anti-glare option to the upper-end machines to begin with? Have you ever tried working in Illustrator or any pro application on unibody glossy MacBook? It is virtually impossible when working with color, say for the 17" which has exceptional color even through the gloss panel version (but still nowhere near the matte version). The viewing angle is terrible and the colors are washed. For entertainment and iLife and such it is fine.
That aside, have you read some of the earlier posts in this particular forum? Does the subject of a Quad core chip missing the iMac's next release come up often? It does, yes. The reason that I mentioned how Apple has always sought to shave thickness from it's products is because it's true. Also, I wouldn't quite call a Core 2 Duo an antique just yet.
When one buys an Apple, I'd say that form is a key element to the brand, yes. When one pays for an Apple, they are also buying into the brand. Original members of Apple have said so... how can one argue with the makers themselves? I never made a point that Apple should abandon a quad core in the first place and stay behind, so I don't know where that is from. (documentary: Welcome to Macintosh.)
And I will ask again, is there anything, other than your belief (as in links to real benchmarks) showing an improved scalability for quad or dual-quad cores under SL other than increases that one would expect from additional cores?
Phoronix has been benchmarking SL against Linux based machines. Haven't been to the sight in awhile so I don't know the status of the testing which has been on going. However some codes are showing very good behaviour under SL. Unfortunately testing to date has been on a dual core machine.
Quote:
At the most basic level and assuming GDC is 100% efficient, a dual-core could process a task in half the time that a single core could.
Yes a highly parallizable task. This is actually a key concern the app or sections of it need to be amendable to parallel processing in the first place. I can assure you though that the apps that have parallizable code will benefit from the additional cores.
Quote:
Again at 100% efficiency, a quad core could process the same task in 25% of the time that a single core could.
We both know however that 100% seldom happens. In any event why deny the utility of more cores? Nobody that I know of denies that GPU processing works for an assortment of apps rather well, so why deny the value of four cores over one or two?
Quote:
You're implying that simply adding more cores will somehow make GDC more efficient.
Efficient is your word. That isn't what I'm really saying at all. What I'm saying is that there is a thread pool managed by the OS that can make use of all cores available to it. That is the huge concept.
Now combine this with the fact that the programmer doesn't need to know about the hardware anymore to leverage those cores then you have a bigger win. With GCD the system figures out how many cores to run the parallel threads on. This means that additional cores in many cases automatically speed up the programmers code.
Quote:
The GCD will simply work with what it has, be that 2 cores, 4 cores, or 8 cores, or whatever. It won't suddenly get more efficient with more cores,
Exactly, GCD works with the cores available to it. This can speed up an apps highly parallel code significantly as more cores are added. The obvious point here is that for many users four cores will be far better than two.
Quote:
It will work within that efficiency regardless of the number of cores, and we know that it isn't 100% efficient (it will never be able to dole out all work perfectly across X number of cores in any case).
Yes it is true that there is no such thing as 100%. Depending on the code you can come awfully close though. The converse is also true in that some code is very much bound to a single prcessor or a small number of threads.
There are other benchmarks floating around the net that highlight GCD and SL but I don't have links at the moment. You could try looking at www.macresearch.org ?? And see what they have to read up on.
It sounds like a lot of folks are really wanting something between the iMac and the Mac Pro, or at a minimum, an iMac Pro of sorts. I would certainly be interested in an iMac Pro if one was available, but I'm not interested in paying a premium price as we already see with the Mac Pro.
.
Quad core is far from 'pro' domain these days. Quad core cpus can be had for $99 now.
For some reason Mac users seem to have been beaten senseless and made to feel inadequate for a quad core machine unless they work in FCP all day.
Comments
I think they should use the latter desktop grade CPUs. The motherboard will have to be redesigned from scratch anyway. With the past AL iMacs, the Core2 boards were tweaked and modified to accommodate different, but similar components. Now is a good time to reinvent the iMac.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
That's a good point. But if Apple moves the iMac and mini (or an incarnation of the mini) to AIO type CPUs then they can still leverage the cost savings. I think the MacBook and MBP together can get guarantee the same discount the currently enjoy. Plus I'm assuming a huge cost difference between mobile versions and AIO versions that they can make up any price differences via improved margins on the mini and iMac (provided they don't lower the prices dramatically).
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
They could do that if they wanted to, but that is not a feasible solution. Liquid cooling adds cost and a higher potential for problems. There was a patent they put out awhile back for drawing heat away from the components but there has been plenty of time that their engineers may have simply found a better solution of component placement to allow for a a higher wattage CPU. There is also the shrinking of certain chips, a more refined MoBo and maybe even a smaller power supply that would allow for additional space in the system for another fan. Even that may be enough; we really don’t know, but I think we can deduce that Apple’s focus with the iMac is not to make it the fastest machine possible (they have no AIO competition) but to make it appealing to their target market, which isn’t me and likely isn’t many on this forum.
Is there anything preventing them from simply installing a different thermal solution, like water cooled? I've never looked at the guts of an iMac so I don't know how much clearance there is between the CPU and the back cover, but it seems like there are solutions out there that could be workable.
There's not enough space for that type of solution under the current case constraints and Apple's had so many issues with liquid cooling (G5 tower) that I don't see them going down that road again.
There's not enough space for that type of solution under the current case constraints and Apple's had so many issues with liquid cooling (G5 tower) that I don't see them going down that road again.
Sorry about that. G5 is before my time. I'm only a recent convert. I didn't know they had already taken a whack at water cooled solutions.
They can get more performance for less money on Intel?s per 1000 chip price and the increase in wattage seems like a feasible from an engineering standpoint over the current setup. My only concerns, trying to look at it from Apple?s perspective, is: Will these desktop-grade CPUs for AIOs get the same future-forward attention from Intel the way their mobile-grade CPUs get? Would it be more economical to buy mobile CPUs over these AIO desktop-grade CPUs since we?re buying so many for our notebooks? Does a little extra performance for all that thermal reengineering really make a difference to people interested in buying an AIO computer?
I'm not convinced that the difference in performance are all that little. So given that what many of us want Apple to do is to address the performance gap between the Mac Pro and the iMac. It doesn't have to be on every iMac but there needs to be a high end model that truly delivers on the performance problem.
Admittedly this is not really what the iMac is designed for and an XMac would be a better product to fill that gap. The problem is many of us are resigned to never seeing an XMac or any thing close to that. A good compromise here would be a high end Mini, which would require a taller cased Mini and a more robust power supply. Even so I can see Apple refactoring the Mini into a power house box, all they need to do is reconsider a few design elements.
Dave
That's not the point. He's implying that SL and Grand Central are really only beneficial for quad-core or higher processors and that it's wasted on the current dual-core implementation all of the Mac's except of course for the quad and dual-quad Mac Pro's. I can only assume this argument is to try to reinforce the idea that iMac needs a quad-core to properly utilize SL, which is untrue. Any multi-core processor will benefit from SL's Grand Central, including a dual-core processor.
There is no doubt in my mind that to truly leverage GCD you need more than two hardware threads. The more the better in fact and that is one of the reasons that the new Intel hardware is so important.
Now that is not saying that SL running on a dual core system won't benefit from GCD just that the benefit is not going to be as large as it could be. Part of the problem here is that the common usage patterns have at least one core sporadically active in background, it is not like the cores are always free to process GCD work units. Combine that with the OpenMP like calls that would effectively use as many threads as can be managed by the OS and you can see that a minimal system for SL ought to be built around a processor with at least four cores.
Of course one would expect qaud-core or dual-quad core's to be faster than a dual-core alone. That was never in question. I actually look forward to quad-core for the entire line, but passing on disinformation, even if it's just implied, doesn't help.
Nor does your statements here. It implies a misunderstanding of the nature of GCD and what it tries to do. Sure it will work on a two core system, that can not be debated as I'm running SL right this minute on my old MBP. However there is no way in hell that I'm going to buy anything new from Apple with less than four cores, it just doesn't make sense.
Dave
Agreed.
I expect to see next year, dual core Arrandale with integrated graphics for the mini and Clarksfield for the iMac with discrete graphics. Hopefully ATI mobile 4800 series class.
Title says it all! The iMac is simply grossly over priced for what you get, it needs to be updated now. Lets face it if Apple can't get Clarksfield into iMac in a couple of months they will have screwed up royally.
Dave
There is no doubt in my mind that to truly leverage GCD you need more than two hardware threads. The more the better in fact and that is one of the reasons that the new Intel hardware is so important.
Now that is not saying that SL running on a dual core system won't benefit from GCD just that the benefit is not going to be as large as it could be. Part of the problem here is that the common usage patterns have at least one core sporadically active in background, it is not like the cores are always free to process GCD work units. Combine that with the OpenMP like calls that would effectively use as many threads as can be managed by the OS and you can see that a minimal system for SL ought to be built around a processor with at least four cores.
Nor does your statements here. It implies a misunderstanding of the nature of GCD and what it tries to do. Sure it will work on a two core system, that can not be debated as I'm running SL right this minute on my old MBP. However there is no way in hell that I'm going to buy anything new from Apple with less than four cores, it just doesn't make sense.
Dave
And I will ask again, is there anything, other than your belief (as in links to real benchmarks) showing an improved scalability for quad or dual-quad cores under SL other than increases that one would expect from additional cores?
At the most basic level and assuming GDC is 100% efficient, a dual-core could process a task in half the time that a single core could.
Again at 100% efficiency, a quad core could process the same task in 25% of the time that a single core could.
You're implying that simply adding more cores will somehow make GDC more efficient. The GCD will simply work with what it has, be that 2 cores, 4 cores, or 8 cores, or whatever. It won't suddenly get more efficient with more cores, It will work within that efficiency regardless of the number of cores, and we know that it isn't 100% efficient (it will never be able to dole out all work perfectly across X number of cores in any case).
It seems that several people are concerned about the possibility of a Quad-core chip skipping the new iMac release, and they are comparing the iMac to some randomly beefed-up Windows PCs.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
When Apple creates the hardware and software, and, as put in other posts, the "user experience" as well, it creates a harmonious, synergistic machine that is not only accelerated by the speed of your chip.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
If anyone's forgotten, Snow Leopard is (as I think I remember reading) something like up to 2.5x faster than Leopard when performing certain tasks, which clues in to the importance of a powerful operating system. It doesn't matter if you had a 5 GHZ Windows machine; if the software bites and the components do not work well together like in a Mac, than your extra specs are virtually worthless.
SL is relatively faster, excluding a few regressions, it however isn't anywhere near 2.5X on average. The problem that you don't seem to grasp is that the software will bite in the future trying to run on a dual core machine. Read up a bit on how GCD works and you will see how silly your position is.
Aside, Apple has always sought to shave thickness from all of its products, and so I don't see this as an exception. It would also be nice if an anti-glare option similar to that on the MBP was available in the next iMac for those of us who need accurate color and less eye strain.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
dave
Title says it all! The iMac is simply grossly over priced for what you get, it needs to be updated now. Lets face it if Apple can't get Clarksfield into iMac in a couple of months they will have screwed up royally.
Dave
Hardly. It's competitively priced for the hardware in it. You'll find the same prices on any hardware manufacturer's site and an iMac includes a 20" or a 24" monitor. A decent quality one at that.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
SL is relatively faster, excluding a few regressions, it however isn't anywhere near 2.5X on average. The problem that you don't seem to grasp is that the software will bite in the future trying to run on a dual core machine. Read up a bit on how GCD works and you will see how silly your position is.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
dave
Agree. If it's do-able in a laptop, then I don't doubt that apple could pull it off in a much larger iMac case:
http://www.notebooks.com/2009/09/16/...et-in-october/
................
I'm genuinely curious where this information is that SL will only really benefit quad core? This is the first I'm reading of it. Could you post a link I'd be interested in reading the technical details. From what I've ready, any multi-core processor will benefit, not just quad-core or an 8-core.
All I can say if you don't believe what is written here read the SL and GCD documentation. Nobody is saying that SL won't work on dual core hardware, all we are saying is that to truly benefit from the tech you need as many cores as is practical. Today that sweet spot is with Quad core hardware.
In a nutshell SL with GCD manages a system wide thread pool. The more threads that can be processed by real cores the faster GCD enhanced software will be. Now a dual core processor can just as easily process threads from that pool but the reality is only two will ever get processed in parallel. A four core machine doubles that and if SMT is available doubles(somewhat) that again.
Now the bad part here is that this implies software that is written to leverage GCD or at a higher level NS"Operation" calls. Some software will likely never be able to leverage those features. A key realization here though is that software that does leverage the new features won't need to know how many cores you have nor make arbitrary decisions about how many threads to use. GCD handles that for the programmer and automatically uses as many threads as is rational for the work taking place. Thus software written to leverage SL & GCD will become faster simply by running on a machine with more cores.
It is important to realize that this doesn't necessarily happen with multithreaded apps written yesterday using older threading models. In the past programmers had to make assumptions about how many cores the threaded app was likely to have or in some cases the ability of a single core to handle a threaded app. For the most part that is gone now.
So I hope you understand it is not an issue of SL not leveraging dual core machines. The issue is that it is a stupid and unnecessary limitation on Apples part. Frankly Im to the point where I have to question Apples ethics to continue to market dual core midrange machines with SL installed. Especially when cost effective quad cores have been around for so long.
Dave
If you look Anand's power measurements for a Clarksfield system, it certainly looks promising in terms of what can be made to fit inside an iMac.
As for the perceived "marketing" problem you mention below, Apple is probably the one company who can deal most effectively with this issue. One simple solution is to add the the peak (Turbo boost) clock rate on the data sheet in addition to the base clock speed. There are some interesting performance comparisons vs a quad-core Core2 system in Anand's test reference above. Also note that Clarksfield clock rate can be made to match or exceed Core 2 clock speeds when only two cores of the four cores are under high load in a Clarksfield system.
Sure, they CAN, but at what cost? The iMac already runs pretty hot with a 44W CPU so I don’t see how a 130W CPU would be possible without Apple making it thicker and/or having a major cooling breakthrough. It’s not just the CPU that now needs to be considered, but larger fans, larger power supply, etc. We know Apple doesn’t like to go thicker—even though they CAN if they wanted to—so that option shouldn’t be considered as it’s not part of of their core philosophy. McDonald’s could make a steak burger using Kobe beef if they wanted to, but a $20 burger isn’t part of their business model.
Remember that Apple’s desktop line is growing faster than the industry. I doubt it’s because of the Mac Pro or the Mac Mini so I have to think it’s because of the iMac. Apparently Apple is doing something right. Even their notebooks are underpowered when you look at the larger notebooks from other vendors using cheaper components that run hotter than Apple’s notebooks, yet Apple is still selling more and dominating the high end market.
PS: Check out the last section of the link below regarding Clarksfield. These are the mobile i7 chips that using 45W and 55W. But now there is a marketing problem with these chips as their clock speed is significantly lower than the previous dual-cores. Most cnsumer use clocks peed as a baseline for determining value, they don’t read charts to determine overall performance. Eventually we’ll see quad-cores in everything, but it’s s tricky time for PC vendors.
So I hope you understand it is not an issue of SL not leveraging dual core machines. The issue is that it is a stupid and unnecessary limitation on Apples part. Frankly Im to the point where I have to question Apples ethics to continue to market dual core midrange machines with SL installed. Especially when cost effective quad cores have been around for so long.
Dave
You'll get no argument there, although I don't think it was 'stupid' of them to not include quad in an iMac as it isn't intended to compete directly with the Mac Pro.
It sounds like a lot of folks are really wanting something between the iMac and the Mac Pro, or at a minimum, an iMac Pro of sorts. I would certainly be interested in an iMac Pro if one was available, but I'm not interested in paying a premium price as we already see with the Mac Pro.
Hopefully they can give us something in between.
No we are concerned about being able to really leverage the new tech in SL. I really don't care at all about Windows, it isn't even part of this discussion other than the fact that in the PC world four cores are close to standard now.
More gibberish and an acknowledgment that you have no idea what we are talking about here. In simple terms people want modern hardware in their iMacs and today that means a minimal a quad core. Further the advent of SL highlights the ability for those cores to actually be leveraged by the OS.
It is foolish to stress form over function to the extent that you ship a crippled product. If Apple does something like this then they are likely to hear about it. iMac needs more not less innovation and capability.
As to accurate colors, explain how that happens on a matte screen?
Ask a plethora of perfessionals why they are upset about the glossed-over mirror that has replaced the standard matte screen, and you will hear that the colors are totally off. Why do you think apple added the anti-glare option to the upper-end machines to begin with? Have you ever tried working in Illustrator or any pro application on unibody glossy MacBook? It is virtually impossible when working with color, say for the 17" which has exceptional color even through the gloss panel version (but still nowhere near the matte version). The viewing angle is terrible and the colors are washed. For entertainment and iLife and such it is fine.
That aside, have you read some of the earlier posts in this particular forum? Does the subject of a Quad core chip missing the iMac's next release come up often? It does, yes. The reason that I mentioned how Apple has always sought to shave thickness from it's products is because it's true. Also, I wouldn't quite call a Core 2 Duo an antique just yet.
When one buys an Apple, I'd say that form is a key element to the brand, yes. When one pays for an Apple, they are also buying into the brand. Original members of Apple have said so... how can one argue with the makers themselves? I never made a point that Apple should abandon a quad core in the first place and stay behind, so I don't know where that is from. (documentary: Welcome to Macintosh.)
And I will ask again, is there anything, other than your belief (as in links to real benchmarks) showing an improved scalability for quad or dual-quad cores under SL other than increases that one would expect from additional cores?
Phoronix has been benchmarking SL against Linux based machines. Haven't been to the sight in awhile so I don't know the status of the testing which has been on going. However some codes are showing very good behaviour under SL. Unfortunately testing to date has been on a dual core machine.
At the most basic level and assuming GDC is 100% efficient, a dual-core could process a task in half the time that a single core could.
Yes a highly parallizable task. This is actually a key concern the app or sections of it need to be amendable to parallel processing in the first place. I can assure you though that the apps that have parallizable code will benefit from the additional cores.
Again at 100% efficiency, a quad core could process the same task in 25% of the time that a single core could.
We both know however that 100% seldom happens. In any event why deny the utility of more cores? Nobody that I know of denies that GPU processing works for an assortment of apps rather well, so why deny the value of four cores over one or two?
You're implying that simply adding more cores will somehow make GDC more efficient.
Efficient is your word. That isn't what I'm really saying at all. What I'm saying is that there is a thread pool managed by the OS that can make use of all cores available to it. That is the huge concept.
Now combine this with the fact that the programmer doesn't need to know about the hardware anymore to leverage those cores then you have a bigger win. With GCD the system figures out how many cores to run the parallel threads on. This means that additional cores in many cases automatically speed up the programmers code.
The GCD will simply work with what it has, be that 2 cores, 4 cores, or 8 cores, or whatever. It won't suddenly get more efficient with more cores,
Exactly, GCD works with the cores available to it. This can speed up an apps highly parallel code significantly as more cores are added. The obvious point here is that for many users four cores will be far better than two.
It will work within that efficiency regardless of the number of cores, and we know that it isn't 100% efficient (it will never be able to dole out all work perfectly across X number of cores in any case).
Yes it is true that there is no such thing as 100%. Depending on the code you can come awfully close though. The converse is also true in that some code is very much bound to a single prcessor or a small number of threads.
There are other benchmarks floating around the net that highlight GCD and SL but I don't have links at the moment. You could try looking at www.macresearch.org ?? And see what they have to read up on.
Dave
It sounds like a lot of folks are really wanting something between the iMac and the Mac Pro, or at a minimum, an iMac Pro of sorts. I would certainly be interested in an iMac Pro if one was available, but I'm not interested in paying a premium price as we already see with the Mac Pro.
.
Quad core is far from 'pro' domain these days. Quad core cpus can be had for $99 now.
For some reason Mac users seem to have been beaten senseless and made to feel inadequate for a quad core machine unless they work in FCP all day.
Repeat after me; Its ok to want quad core.
Repeat after me; Its ok to want quad core.
It's ok to want quad core.