Not to hijack a thread, but this happens in many industries - some to a degree worse than seen here.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Post your last guess of Mac Pro 2012 spec!!
- Thread starter Amethyst
- Start date
- Sort by reaction score
Not to hijack a thread, but this happens in many industries - some to a degree worse than seen here.
There are, of course, other factors to take into consideration, such as the average salaries in similar jobs at similar levels of education. However, with globalisation, companies can buy/produce their products all over the world to benefit from the cheapest labour and best prices. As a consumer, I feel that I should be able to do the same and buy from where it's the cheapest without being forced to pay customs. Unfortunately that is not the case. That's why this whole globalisation hoopla is a bunch of baloney, in my opinion.
Anyway, now I've gone completely off-topic.
any one see this link?
http://micgadget.com/26895/upcoming-new-features-and-specs-for-the-new-mac-pro/
http://micgadget.com/26895/upcoming-new-features-and-specs-for-the-new-mac-pro/
I think (and hope) we'll see the 1650, instead of the 1620, in the base model Mac Pro. The problem with the 1620 is that the i7-3770 that will probably be the top-end option in the 27"iMac outperforms it (not by much, but the 3770 is a bit faster than the i7-3820 that is a VERY close relative of the Xeon 1620 on most benchmarks). The 2010 quad core Mac Pro is significantly outperformed by the current iMac (especially the 3.4 gHz upgrade model) , but (barely) outperformed the fastest iMac available when it came out (and wasn't well liked by most reviewers because it wasn't enough faster than an iMac). This release is likely to turn the tables with the iMac actually being faster than a base Mac Pro, unless all Mac Pros have at least 6 cores.
The 1650 neatly solves this problem, since a 1650-based 6 core Mac Pro would be approximately 1.5 times as fast as any iMac (it's about the same clock speed, has 6 cores instead of four, and the 5% hit because it's Sandy Bridge instead of Ivy is pretty much made up by the faster quad-channel memory controller), restoring the type of performance gap seen in the better value Mac Pros of 2008. The chip is only about $300 more expensive than the quad-core.
My hoped-for lineup would be a 6 core Xeon 1650 in the base model, then one of the mid-priced dual 6-core 2600 series (there are a bunch of them, perhaps dual E5-2630s) as the lower-end dual-chip machine, with a couple of dual 8-core upgrade options (one of which might be a pair of high-speed, but very expensive, E5-2680 or E5-2690 chips).
Apple should also (hopefully) take care to separate the graphics performance between the iMac and the Mac Pro - either the base graphics card or a reasonably priced ($100) upgrade should clearly outperform any upgraded iMac GPU.
Since the base Mac Pro (without a display) is roughly the same price as the top-end iMac (with a $1000 display included), it needs to offer performance that is compellingly better than the iMac in order to offer any value at all. This is especially true if the iMac is upgraded to a Retina Display, which could be one of the finest displays available anywhere, at any cost (assuming that the color gamut and fidelity match the resolution).
The 1650 neatly solves this problem, since a 1650-based 6 core Mac Pro would be approximately 1.5 times as fast as any iMac (it's about the same clock speed, has 6 cores instead of four, and the 5% hit because it's Sandy Bridge instead of Ivy is pretty much made up by the faster quad-channel memory controller), restoring the type of performance gap seen in the better value Mac Pros of 2008. The chip is only about $300 more expensive than the quad-core.
My hoped-for lineup would be a 6 core Xeon 1650 in the base model, then one of the mid-priced dual 6-core 2600 series (there are a bunch of them, perhaps dual E5-2630s) as the lower-end dual-chip machine, with a couple of dual 8-core upgrade options (one of which might be a pair of high-speed, but very expensive, E5-2680 or E5-2690 chips).
Apple should also (hopefully) take care to separate the graphics performance between the iMac and the Mac Pro - either the base graphics card or a reasonably priced ($100) upgrade should clearly outperform any upgraded iMac GPU.
Since the base Mac Pro (without a display) is roughly the same price as the top-end iMac (with a $1000 display included), it needs to offer performance that is compellingly better than the iMac in order to offer any value at all. This is especially true if the iMac is upgraded to a Retina Display, which could be one of the finest displays available anywhere, at any cost (assuming that the color gamut and fidelity match the resolution).
You're looking at it from a buyer's perspective. It doesn't matter to Apple which system people buy.
I think we'll likely see the top end single processor GPUs, whether NVidia or ATI. Power connectors will get boosted to 8 pin.
16 cores at the top end. USB 3. Thunderbolt.
Possibly drives changed to 2.5" slots. Possible elimination of the optical drives.
Monitors are still up in the air. Not sure I see Apple doing this with monitors that have ethernet jacks on them. Possible we'll see some sort of new Apple pro display?
16 cores at the top end. USB 3. Thunderbolt.
Possibly drives changed to 2.5" slots. Possible elimination of the optical drives.
Monitors are still up in the air. Not sure I see Apple doing this with monitors that have ethernet jacks on them. Possible we'll see some sort of new Apple pro display?
Why is this a problem? Is the some "social caste system" in play here. All the Mac Pro has to be is close. People aren't just buying CPUs. If so then perhaps have a small problem, but the Mac Pro isn't targeted at people who are just looking for the cheapest box wrapped around a CPU.
The vast majority of those folks are going to be buy a Windows PC with a much cheaper infrastructure wrapped around a CPU.
An equally easy solution is replace the HDD in the standard entry config with a SDD (that costs about the same as a 1TB drive, so taking a capacity trade off). If the iMac config has a HDD then there will be difference in most benchmarks that don't just sit in the CPU onboard cache.
The iMac could have a SSD also, but it isn't standard and the difference between the user putting in a second themselves and the user putting it into the Mac Pro is largely all the "value proposition" you really need for the folks in the market the Mac Pro primarily targets. If someone needs a "faster than the entry" Mac Pro and has $300 more dollars there is a the "better" model the next step up.
Even with the $300 more dollars they are stuck with the iMac. There is your "value proposition" difference.
Furthermore, seems likely the high clocked i7-3770 will be a CTO option anyway. It is a
Just wait for the tests. The memory bandwidth of the Mac Pro is likely to be higher on tasks that actually use alot of memory. It also likely will run better during sustained CPU loads.
More expensive. Apple is extremely unlikely to "eat" that increased cost. So what you are proposing that Apple increase the Mac Pro entry costs up to $2800 level. That's going in the wrong direction.
Frankly, the entry Mac Pro would probably compete better if its price were $2,200 or $2,300. That is better accomplished with a 1620. The GPU would be the major differentiators, if just want to fixate on CPU/GPU perspective. The iMac's i7 is a CPU/GPU device. It only seems fair to evaluate the Mac Pro with the same combination.
That's not new. That's the case now.
Myopic. CPU Performance is not the only attribute that has value.
It is somewhat doubtful that iMacs will get Retina Displays. The problem with a "Retina" 27" display is that number of pixels is quite large. The "mobile" class GPUs in an iMac will do worse against mid-range desktop GPUs (like the Mac Pro has at the entry level) with those kinds of workloads once the users has lots of windows open with Apps that use OpenGL/OpenCL to accelerate things.
Retina is more likely going to be confined to sub 20" screens. (Probably below 17" ). Below 17-20" means the iMac is out. It is extremely unlikely Apple is going to roll back the iMac to last decade sizes just for short term "Retina display" gains. They can wait for the technology to catch up at affordable prices.
Somewhat. But is also doesn't matter to Apple which system they sell. Apple could quit the $2-3K Mac Pro market if the systems are "purely equivalent" to Apple.
Mac Pro growth is flat/negative Apple will nuke the product.
The Apple quality I think that is more important here is that Apple is patient. If there is a future technology gap that will further define the difference they will let that play out. In short, by the Ivy Bridge update later it won't be a problem with this same basic design.
Frankly, it looks like Intel "kneecapped" the 1620 turbo range to nudge folks into buy the 1650 and 1660. When the shrink to Ivy Bridge comes I bet the Turbo range on the 1620 goes up but so does the new base rate on the 1650 1660 put more distance between them and the 4 core "more base GHz" focused model.
Either that or the 1620 goes 6 core on Ivy, but loses substantive base speed. The i7's the iMac uses are stuck at 4 core for probably at least 2 more years (Haswell for sure and likely Broadwell also). Again the 1650 will pick up speed and the 1660 might go 8 core.
I suspect Ivy Bridge E5 1600 and i7 39xx will just clock higher instead of more cores. That sets the stage for them getting some iGPU injections at Haswell.
Last edited:
Ripped off? Ever ship something the size and weight of a Mac pro from china to the states?
It's cheap. Ever ship the same to oz? It ain't.
+ 10% tax. $3000 sounds a bit right.
Apple's are more expensive in England and much of the rest of Europe because there's more Europeans, percentage wise who can afford it. That goes for most products.
Kind of why many icelanders make vacations to NY to pick up new macs. Iceland air on their end is about the difference.
It's cheap. Ever ship the same to oz? It ain't.
+ 10% tax. $3000 sounds a bit right.
Apple's are more expensive in England and much of the rest of Europe because there's more Europeans, percentage wise who can afford it. That goes for most products.
Kind of why many icelanders make vacations to NY to pick up new macs. Iceland air on their end is about the difference.
I'm assuming that a top-end iMac will have a fairly high-powered mobile GPU, just like it does now (ignoring the built-in graphics on the 3770, which will either go unused or be used in non-graphics intensive applications to save power) - non-upgradeable, but equivalent to a bit better than the cheap desktop GPU Apple usually uses as the default on the Mac Pro.
My definition of a top-end iMac is a 27" with both the processor bump and the GPU bump, although I might well ignore the internal SSD option if it's still as overpriced as it was last generation (I'm happy to boot from a Thunderbolt SSD if they're going to charge $600 for a $200 internal). Any Mac Pro that can't keep up with that ~$2400 machine is, at least to my way of thinking, useless (unless it costs, say, $1799).
That was not nearly as true before Thunderbolt, when the fastest way of connecting anything to an iMac was Firewire 800. The Mac Pro had huge advantages in disk capacity and there were numerous applications which required specialized I/O that took advantage of the PCIe slots. Now, though, Thunderbolt provides a very fast interface for disks, fast enough, in fact, that most other fast interfaces can run over Thunderbolt, and eSATA, USB 3.0 and even Fibre Channel adapters exist (the first two are fairly reasonably priced for connecting big storage and Fibre Channel is never reasonably priced, no matter how you connect it). Thunderbolt interfaces for pro video gear are becoming increasingly available, and the cost premium over PCIe is coming down rapidly.
Of course, if you have multiple big RAIDs and the highest-end multilane HD-SDI interfaces, you'll saturate Thunderbolt - but the people running that equipment will be buying 12 or 16 core Mac Pros, not 4 or 6 core models. The entry Mac Pro is for photographers (like me) and small video shops, not James Cameron. I can't think of any application (other than gaming, with its need for big GPUs and relatively modest CPUs) that would require I/O that can't be handled with Thunderbolt, yet suggest a single-chip Mac Pro instead of a dual-chip with 12+ cores. Maybe some folks who already own a bunch of PCIe cards (or a RAID that connects via PCIe) might buy a Mac Pro that was no faster than an iMac...
I think, however, that a reasonably priced 6-core E5-1650 (or i7-3930, but given the similar pricing, Apple will pick the Xeon variant) Mac Pro would sell relatively well. It would be a very appealing machine to people like me - photographers and other creative pros who'd like to pick their own display, need a lot of computer power, but don't have Pixar budgets. If the $2500 Mac Pro is a laughably bad deal with the performance of an iMac (or a $900 Dell XPS 8500), it won't appeal to that relatively savvy group, but if it is a 6-core (which are always expensive - Dell's cheapest 6-core is a $2000 Alienware), it will sell a lot of units to photographers, indie filmmakers, etc...
I certainly don't need the $500 case and 1000 watt power supply the low-end Mac Pro has long had to support the needs of the 12 core, and I can't imagine most folks like me do, either. We'd like something significantly faster than any iMac, with easier drive upgrades and the ability to pick a display, but we don't need a supersize tower, and don't want to pay for case and power supply and end up with a relatively low-end CPU...
My definition of a top-end iMac is a 27" with both the processor bump and the GPU bump, although I might well ignore the internal SSD option if it's still as overpriced as it was last generation (I'm happy to boot from a Thunderbolt SSD if they're going to charge $600 for a $200 internal). Any Mac Pro that can't keep up with that ~$2400 machine is, at least to my way of thinking, useless (unless it costs, say, $1799).
That was not nearly as true before Thunderbolt, when the fastest way of connecting anything to an iMac was Firewire 800. The Mac Pro had huge advantages in disk capacity and there were numerous applications which required specialized I/O that took advantage of the PCIe slots. Now, though, Thunderbolt provides a very fast interface for disks, fast enough, in fact, that most other fast interfaces can run over Thunderbolt, and eSATA, USB 3.0 and even Fibre Channel adapters exist (the first two are fairly reasonably priced for connecting big storage and Fibre Channel is never reasonably priced, no matter how you connect it). Thunderbolt interfaces for pro video gear are becoming increasingly available, and the cost premium over PCIe is coming down rapidly.
Of course, if you have multiple big RAIDs and the highest-end multilane HD-SDI interfaces, you'll saturate Thunderbolt - but the people running that equipment will be buying 12 or 16 core Mac Pros, not 4 or 6 core models. The entry Mac Pro is for photographers (like me) and small video shops, not James Cameron. I can't think of any application (other than gaming, with its need for big GPUs and relatively modest CPUs) that would require I/O that can't be handled with Thunderbolt, yet suggest a single-chip Mac Pro instead of a dual-chip with 12+ cores. Maybe some folks who already own a bunch of PCIe cards (or a RAID that connects via PCIe) might buy a Mac Pro that was no faster than an iMac...
I think, however, that a reasonably priced 6-core E5-1650 (or i7-3930, but given the similar pricing, Apple will pick the Xeon variant) Mac Pro would sell relatively well. It would be a very appealing machine to people like me - photographers and other creative pros who'd like to pick their own display, need a lot of computer power, but don't have Pixar budgets. If the $2500 Mac Pro is a laughably bad deal with the performance of an iMac (or a $900 Dell XPS 8500), it won't appeal to that relatively savvy group, but if it is a 6-core (which are always expensive - Dell's cheapest 6-core is a $2000 Alienware), it will sell a lot of units to photographers, indie filmmakers, etc...
I certainly don't need the $500 case and 1000 watt power supply the low-end Mac Pro has long had to support the needs of the 12 core, and I can't imagine most folks like me do, either. We'd like something significantly faster than any iMac, with easier drive upgrades and the ability to pick a display, but we don't need a supersize tower, and don't want to pay for case and power supply and end up with a relatively low-end CPU...
That wasn't true during most of the early reign of the Mac Pro 2010 model. The iMac 2010 era GPU topped out at with a 5750M with 1GB VRAM. The desktop 5770 outclasses that card.
It is true that the 6790M is a bit better than the 5770. But the 6790M came out in January, 2011. The 5770 came out October 2009. Sure give the mobile GPU almost a 2 year jump on the Mac Pro card ... yeah it is going to fall behind. It fell behind just like the Xeon 3500 fell behind the i7 3770. They are from different eras.
There is zero reason why Apple would put "last year" cards in the Mac Pro on this update. Even if the iMac upgrades to Nvidia/AMD latest mobile they likely won't catch the mid-range desktop card Apple selects for the Mac Pro as the entry point.
The only way the Mac Pro would come in lower GPU is if Apple embedded the Mac Pro's GPU on the Mac Pro motherboard. But then you'd have a CTO box that was cheaper than the entry price point and an empty slot could fill with a faster card. That's quite competitive.
First, it is unlikely the Mac Pro will ever be priced below an iMac. ( at least this full sized version. If they came out with a third Mac Pro offering in a small case with an E3 processor perhaps there would be some minor overlap but I doubt it).
Second, this is one of the problems now with Mac Pro pricing. It is not that the Mac Pro is "gapped" from the top end standard configuration but gapped higher than the top end standard with the "deluxe" CTO CPU/GPU options.
The CTO top end MBP 15" is just as fast as the entry MBP 17". That has nothing to do with the entry, "Good" version, level pricing of both models.
Mac Pro still has advantages. Are more folks going to settle for fully pumped up iMacs yes. That's nothing new. The Mac Pro just need to find a few new users with workloads from machines larger than it that it can now reach. For the bulk of the market that the market that the Mac Pro is targeted at many of those workloads will increase roughly in step with the Mac Pro. So the iMac isn't a long term option.
If Thunderbolt completely solves the user's problem those users probably didn't need a Mac Pro in the first place.
By the time you pay $300-900 for that external Thunderbolt RAID box with 4 slots you could have paid $300 extra for the 6 core Mac Pro.
There likely are some using the 6 core models. This all started with you saying the 1650 was "enough" to put a bigger gap between the Mac Pro and the iMac. Well it is. It just isn't necessary that it be at the bottom.
Now you are drifting off into a different "path" where the Mac Pro single package model completely disappears. Not likely to happen. The single package and dual package models are likely going to be coupled for a long while.
Hence the Mac Pro entry level priced exactly where it is. Either you find value in picking your own display or you don't. If you do then you'll pick the Mac Pro over the iMac. The relatively minor speed match between the fully customized iMac and a entry Mac Pro shouldn't be an issue if place any significant value premium on choosing your own monitor.
Apple sells the iMac for those in the $900 Dell box folk. That's it competitor. If the built-in monitor doesn't work for them Apple is OK with that. As long as enough people buy the iMac that is what counts.
The Mac Pro isn't aimed at $900 Dell tower boxes any more than the $2,000 Dell workstation boxes are aimed at the $900 Dell towers.
Last edited:
An iMac drives an external display perfectly capably, so (at least for me, and I suspect for a lot of folks), I'd rather have the built-in monitor as a (very capable) secondary display than have a 60 lb supertower case, given the choice of two comparably performing machines at comparable prices (if the iMac had two thunderbolt sockets, which it does). If the iMac gets a display with upgraded resolution and gamut, the internal might serve as a primary display, and I'd hook up a less expensive secondary. Of course, this also gives the ability to move the machine around (occasionally, a 27" iMac is no laptop) without lugging close to 100 lbs of computer and display.
What would swing these marginal use cases back towards the Mac Pro is an entry level Mac Pro with considerable performance advantages over an iMac. The 2008 Mac Pro ( the last one where I wouldn't consider the entry model overpriced) had a 2.8 gHz quad core entry model for $2199 ( the price of the 24" iMac before any upgrades). That machine was almost twice as fast as any iMac of its time. There was a $2799 octo-core that many consider the best Mac Pro value ever that was about FOUR TIMES as fast as any iMac, and only carried a $300 -$400 price premium over the best CTO iMac. Today's iMac-priced Mac Pro offers iMac performance instead of nearly double that, and the Mac Pro that is 4 times as fast as an iMac costs something like $6299 instead of $2799. This was also before Thunderbolt, and before iMacs accepted 32 gb of RAM, so the iMac was much more crippled than it is today.
Using the top iMac as a benchmark in both price and performance to show how fast the rest of Apple's line has been evolving, the price of the entry Mac Pro has remained relatively constant (between 1 and 1.25 iMac units), and the performance of the top model has also remained relatively constant (between 3 and 4 iMac units). However, the relative performance of the entry model has dropped from a high of nearly 2 iMac units at the release of the 2008 Mac Pros to less than 1 (albeit because of old Mac Pros right now) at present. If the entry 2012 model features the E5-1620, it will probably be the first Mac Pro ever released with a performance of less than 1 iMac. The relative performance of the top model has remained fairly constant, but its price has climbed from less than 2 iMac units to about 3 ( and the 2008 octo-core had a model that offered over 3 iMacs of performance for about 1.3 iMacs of price, then offered diminishing returns with slightly faster processors for a lot more money)
For Mac Pros to make sense again, outside of the extremely expensive 12 (and maybe soon 16) core machines used by a few ultra-high end video pros, they need to offer a far better price-performance ratio.
----------
P.s. I never owned a 2008 - I wish I had ( i might well still have it if i had)I owned a 2006 Mac Pro until 2010, and when it came time to upgrade, the Mac Pros just didn't make sense - there was an entry model that offered no performance advantages, and screamers I couldn't afford.
What would swing these marginal use cases back towards the Mac Pro is an entry level Mac Pro with considerable performance advantages over an iMac. The 2008 Mac Pro ( the last one where I wouldn't consider the entry model overpriced) had a 2.8 gHz quad core entry model for $2199 ( the price of the 24" iMac before any upgrades). That machine was almost twice as fast as any iMac of its time. There was a $2799 octo-core that many consider the best Mac Pro value ever that was about FOUR TIMES as fast as any iMac, and only carried a $300 -$400 price premium over the best CTO iMac. Today's iMac-priced Mac Pro offers iMac performance instead of nearly double that, and the Mac Pro that is 4 times as fast as an iMac costs something like $6299 instead of $2799. This was also before Thunderbolt, and before iMacs accepted 32 gb of RAM, so the iMac was much more crippled than it is today.
Using the top iMac as a benchmark in both price and performance to show how fast the rest of Apple's line has been evolving, the price of the entry Mac Pro has remained relatively constant (between 1 and 1.25 iMac units), and the performance of the top model has also remained relatively constant (between 3 and 4 iMac units). However, the relative performance of the entry model has dropped from a high of nearly 2 iMac units at the release of the 2008 Mac Pros to less than 1 (albeit because of old Mac Pros right now) at present. If the entry 2012 model features the E5-1620, it will probably be the first Mac Pro ever released with a performance of less than 1 iMac. The relative performance of the top model has remained fairly constant, but its price has climbed from less than 2 iMac units to about 3 ( and the 2008 octo-core had a model that offered over 3 iMacs of performance for about 1.3 iMacs of price, then offered diminishing returns with slightly faster processors for a lot more money)
For Mac Pros to make sense again, outside of the extremely expensive 12 (and maybe soon 16) core machines used by a few ultra-high end video pros, they need to offer a far better price-performance ratio.
----------
P.s. I never owned a 2008 - I wish I had ( i might well still have it if i had)I owned a 2006 Mac Pro until 2010, and when it came time to upgrade, the Mac Pros just didn't make sense - there was an entry model that offered no performance advantages, and screamers I couldn't afford.
Last edited:
Since an introduction of Unibody iMac. iMac top is away slightly faster than entry Mac Pro. And people still love it (20000+ on Facebook want a new mac pro page). Why Apple have to change that?
Mac Pro is something more than just spec. guys!
