Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
The next Mac mini "update" will be Cloud OS X - all "your" software, rented to you ($$$) monthly, with the added bonus (for Apple) of selling you a physical box you an croon and coo over, as you stroke it and marvel at the wonders of machined aluminium (unibody silver is a bit tired now, Apple - how about some EXCITING COLOURS, like the Lumia phone range?).

You won't have to "worry" about installing software - it will all be "on da clowd" - and you won't have to consider any likely problems, as the internet never, ever goes down... honest.

You are really bashing Apple for something they didn't do and likely never will do that you came up with yourself?
 
No one in my family has ever had a Windows Machine last for more than a year. I would bet that Apple considered the scenario you presented already and still decided that this mini release was more important given their future plans if any for the mini.

Really? I have a cheap $400 Compaq from 2004 that I custom ordered back in 2003 with focus on processing speed / cores and skimped on the rest so I could enhance it with parts I had available and save money... I've given it a little more memory, a second hard drive, and a secondary graphics card... And, that thing is still going and going strong. It's more relevant today than any of the Mac's I have in my collection (including my Mac Pro).

And, for that matter, my Mac Pro from 2006 is still more relevant today than the ridiculous "update" to the current Mac Mini. Sadly, cause I was actually getting ready to replace my Mac Pro soon, and recognized that the last Mac Mini was almost to a point of being a suitable replacement, and was waiting for the update to make it just right. However, the new Mac Mini took significant steps in reverse, and killed that idea.

I don't need the current Mac Pro versions... more than I need. Too expensive. I also neither need or want an iMac. I have my own monitors already. And, I've already been too many rounds with the iMac's where one small part fails and you're not only out your entire machine, but also your monitor as well. Ridiculous.

The truth is that PC's are relevant longer than any Mac cycle. You know, I can still install Windows 8.1 on my PC from 2003. That old PC still exceeds the minimum requirements by more than 3 times.

Unfortunately, I haven't been able to update my 2006 Mac Pro beyond OS X 10.6.8 due to Apple being ridiculous and coding their OS to prevent me from upgrading. If I want to run Apple's current OS version, I have to resort to some hack that will always be temperamental, or buy a new machine. Apple has specifically written OS X to prevent it from being installed on my older machine without hacking it manually to get around Apple's road block.
 
Really? I have a cheap $400 Compaq from 2004 that I custom ordered back in 2003 with focus on processing speed / cores and skimped on the rest so I could enhance it with parts I had available and save money... I've given it a little more memory, a second hard drive, and a secondary graphics card... And, that thing is still going and going strong. It's more relevant today than any of the Mac's I have in my collection (including my Mac Pro).

And, for that matter, my Mac Pro from 2006 is still more relevant today than the ridiculous "update" to the current Mac Mini. Sadly, cause I was actually getting ready to replace my Mac Pro soon, and recognized that the last Mac Mini was almost to a point of being a suitable replacement, and was waiting for the update to make it just right. However, the new Mac Mini took significant steps in reverse, and killed that idea.

I don't need the current Mac Pro versions... more than I need. Too expensive. I also neither need or want an iMac. I have my own monitors already. And, I've already been too many rounds with the iMac's where one small part fails and you're not only out your entire machine, but also your monitor as well. Ridiculous.

The truth is that PC's are relevant longer than any Mac cycle. You know, I can still install Windows 8.1 on my PC from 2003. That old PC still exceeds the minimum requirements by more than 3 times.

Unfortunately, I haven't been able to update my 2006 Mac Pro beyond OS X 10.6.8 due to Apple being ridiculous and coding their OS to prevent me from upgrading. If I want to run Apple's current OS version, I have to resort to some hack that will always be temperamental, or buy a new machine. Apple has specifically written OS X to prevent it from being installed on my older machine without hacking it manually to get around Apple's road block.
Apple clearly doesnt want your money. Sorry. You can say anything but I have to go with first hand experience.
 
they'd rather choose an expensive ultra low TDP chip for less performance in a DESKTOP machine. Isn't that ridiculous enough? So to answer your question, for $200 I CAN cook a more powerful desktop than the mini.

desktop i3 - 4150T: benchmark 4537, $117
mobile i5 - 4250U:benchmark 3462, $342

To be fair, Apple probapbly get the mobile i5 for less than the desktop i3.

That i3-4150T has older HD4400 Graphics, which would cause complaints in itself. It is a 35W part - the i7-4578U is 28W, so they can give over that 7W of thermal budget to something else. They have experience and economies of scale with the 4250U in other products. Most importantly, it uses a different socket, which leads to the whole "needs two different motherboards" issue that Apple would probably prefer to avoid.

And if Apple's concern was that almost all of the lower-cost quad-core mobile chips have the slower HD 4600 graphics, that shouldn't be a problem - Intel has been known to create custom chips for Apple, before. I'm sure they'd be willing to throw together a low-end HD 5000 or Iris quad-core chip that has higher power draw than the standard 'mobile' chip, but still in the same form factor, for ease of integration.
Apple can probably influence Intel in some areas, but probably less so for something that wouldn't be a high volume item. The other question is how much thermal budget Apple could devote to the Haswell after everything else is accounted for. I'd suspect those 47W i7-4770HQ quad-cores people are likely too "hot" without more room/fans, etc.

At the end of the day, Apple is probably satisfied to offer the highest performance top-end mini option possible without unreasonable effort given a form factor they like for the low and mid-end, rather than starting with a high-end target and having that drive the decisions. For Haswell, that eliminates the quad-core for now. They may not be entirely unhappy about a certain amount of business moving up the product line (and must be willing to lose a few), but I'd suspect it is a side-effect, rather than the primary reason for what's being offered.
 
I prefer a low-power SoC but useful for my daily tasks (browsing, mail, watching streams and YouTube content, multimedia and some low-power games).

And like i said in other threads, i would go to base version without extras. Its good for my needs, OSX is efficient, and its a silent computer with fresh thermal dissipation and low-power consumption, which i don't need to power-off all time.

For who needs more from CPU, i seriously suggest waiting some several time for new Intel chips or go to another solutions, with or without OSX.
Its sad not having quad-cores and more said the no-friendly RAM upgrade.

Obsessing about power consumption of a computer is ridiculous compared to other life events and daily tasks that increase your electric bill.

Here's an example... I used to turn off all my computers when I was done with them, and turn them on when I wanted to use them. Saving every bit of idle electricity waste I could.

My electric bill was $180 per month.

Wanna know what made a huge difference in my electrical bill??? Getting rid of a wife... The absolute only change in the house, booting her to the curb cut my electric bill by more than half.

Now, I run with all my computers on full time. And, I've added tons of other devices to the house that draw electricity, and never turn them off. Every electronic device is always on and waiting for me to tell it to do something. And, my average electric bill now??? $60 a month.

I've gone from being ultra conservative on electrical consumption to leaving all the computers on 100% of the time, and adding more electrical appliances, and I'm still saving tons of money. What do I attribute this to??? Simple, no hair dryers running, no curling irons, straightening irons, and other such items that only serve to alter appearances in hopes of appearing more attractive.

I wouldn't worry so much about 15 or 30 watts maximum power draw on a CPU. I'd worry more about a 1500 to 2000 watt hair dryer.
 
The truth is that PC's are relevant longer than any Mac cycle. You know, I can still install Windows 8.1 on my PC from 2003. That old PC still exceeds the minimum requirements by more than 3 times.

Interesting.

My experience has been almost the exact opposite.

Also, Ive found the Windows "minimum requirements" to be a joke. Running Windows at the bare min config is usually a recipe for disappointment. Conversely, Ive found that Macs actually seem to run at the lower end much more effectively...
 
??
I said there's no difference using mail on 2,4,or6 core.. yes, mail.app (and safari) is multi threaded.. but neither have any CPU intensive tasks therefore it's pretty irrelevant (or certainly bad examples to be using when trying to make a case for buying more cores)
. a singlecore computer will run them equally well as a 12core.

Again missing the point...

The point is, that most applications are designed to utilize multiple cores. Therefore, most applications will gain benefit from multiple cores.

He chose e-mail as a simple program that is so ridiculously simple that if any program wouldn't utilize multiple cores, it would be that program. By using such a menial program for illustrative purposes, it shows that practically anything you do could benefit from multiple cores to spread the work load.

Yes, e-mail will function on a single core. But, as with other programs, it's effect on the system is less noticeable if it's workload is dispersed across multiple cores.

You're being thick on purpose. Give it up.

----------

Interesting.

My experience has been almost the exact opposite.

Also, Ive found the Windows "minimum requirements" to be a joke. Running Windows at the bare min config is usually a recipe for disappointment. Conversely, Ive found that Macs actually seem to run at the lower end much more effectively...

If you reread my post, you'll see that I point out that my PC / Windows system from 2003 exceeds the minimum requirements of Windows 8.1 by a factor of at least 3.

In other words, it is 3 times faster than the minimum requirements. It also has more processor cores than required. It has more memory than is required. It exceeds the graphics capabilities with ease.

However, my Mac Pro from 2006 can't run the current or any recent versions of OS X simply because Apple says no. What a way to extend the useful life of a computer. Apple flops on it's face in that aspect.
 
Now, I run with all my computers on full time. And, I've added tons of other devices to the house that draw electricity, and never turn them off. Every electronic device is always on and waiting for me to tell it to do something. And, my average electric bill now??? $60 a month.

I bought these USB-heated slippers. They are only a slight fire hazard (also tripping and electrocution), but I was able to lower my thermostat and decrease my energy use.

I'd recommend not plugging straight into your iMac... get a good hub, and don't spill nothing on them.
 
I bought these USB-heated slippers. They are only a slight fire hazard (also tripping and electrocution), but I was able to lower my thermostat and decrease my energy use.

I'd recommend not plugging straight into your iMac... get a good hub, and don't spill nothing on them.

Great idea... buys a pair and sends them to ex for Christmas with rewritten directions that recommend charging while wearing for best results :D
 
What are the odds of that? At best, 50/50, I think. By my recollection:

Core Duo: Compatible.
Core 2 Duo: Same socket, but not compatible with early Core Duo chipsets.
Nehalem: New socket.
Westmere: Compatible.
Sandy Bridge: New socket.
Ivy Bridge: Compatible.
Haswell: New socket.
Haswell refresh: Compatible.
Broadwell: Same socket, but not compatible with early Haswell chipsets.

So basically, they're changing the socket with every non-minor-speed-bump release, and with some of the minor speed-bump releases. This translates to about a two-year incompatibility cycle. Most people don't upgrade their CPUs more often than that, so in practice, motherboards are now disposable even if you build a machine yourself. The only things you can usefully reuse are the case, the power supply, and mass storage devices.

IMO, the entire industry might as well switch to soldered-on CPUs. There's no real advantage to socketed CPUs at this point, given the relatively short upgrade window. Ditching sockets would halve the number of SKUs (no need to stock the CPU and motherboard separately), would reduce effort for system builders (both in choosing parts and in installation), and would improve reliability.

There's a downside, of course. Most Haswell machines on the market draw 30W idle consumption or higher. At $0.38 per kWh (what I pay for additional consumption), assuming the machine runs 24 x 7, that translates to $67 per year more in your power bill, or a couple hundred bucks over a three-year product lifetime. Heck, the Alienware I looked at idles at 46W, or $120 per year more in power costs. That adds up rather quickly.

I'd *love* to find a PC that draws 10W or less when idle, with a desktop-class Haswell Core i7. In theory, I'm told that it is possible to build systems that are roughly in that ballpark, but most of the systems out there don't get close.

There's a major downside to your soldered motherboard/CPU argument: you wouldn't be able to pair your desired CPU with your desired motherboard. It would certainly work for manufacturers like Dell, however. But you're right on the non-upgradability of most PC parts. I wanted to upgrade mine recently, and found that I basically had to build a new computer. I went in the opposite direction of you, however, and built a machine that consumes quite a few watts with top-notch performance.
 
Unfortunately, I haven't been able to update my 2006 Mac Pro beyond OS X 10.6.8 due to Apple being ridiculous and coding their OS to prevent me from upgrading. If I want to run Apple's current OS version, I have to resort to some hack that will always be temperamental, or buy a new machine. Apple has specifically written OS X to prevent it from being installed on my older machine without hacking it manually to get around Apple's road block.
Can you not install 10.7 Lion on your 2006 Mac Pro? I have a Late 2006 iMac which I believe has the same issue as you - the boot EFI system is only 32-bit and won't boot the 64-bit kernel (although you can run 64-bit programs with the 32-bit kernel). 10.8 Mountain Lion and later requires booting the 64-bit kernel, so those 2006 iMacs and Mac Pro models can't run it, because they can't boot a 64-bit kernel. But 10.7 Lion should still work (I've booted it on that old iMac, although I kept it on 10.6.8).

Apple hasn't done anything specific to spite you; what they haven't done is go back and write new 64-bit boot EFI firmware for those machines (new ones came with 64-bit EFI). A bummer, but to be fair it would probably have meant not only creating the EFI, but 64-bit device drivers for the old video cards and whatnot in those machines. They may just have said "too much low-level work" and set those machines adrift, or there may be some technical limitation that makes it infeasible to update that generation hardware to a 64-bit EFI.
 
Last edited:
Again missing the point...
what point? there are a whole lotta points being strewn about this thread.

The point is, that most applications are designed to utilize multiple cores. Therefore, most applications will gain benefit from multiple cores.
if that's the point then the point is wrong.. srry

He chose e-mail as a simple program that is so ridiculously simple that if any program wouldn't utilize multiple cores, it would be that program. By using such a menial program for illustrative purposes, it shows that practically anything you do could benefit from multiple cores to spread the work load.

i think you might not understand parallel processing.. it has nothing to do with how 'ridiculously simple' a program might be... for example, here's a ridiculously simple equation--- 6 -2 +4 ÷8 = 1
if you try to process it in parallel, it will be
thread1: 6-2=4
thread2: 4÷8=.5
combine threads-- 4.5

so the multithreaded version is giving the wrong result.. most processes are linear and can't be broken down in little pieces then add the results at the end.. or often, an algorithm which needs to run requires input from a previous calculation so again, the process needs to occur in a linear sequence in order to A) have numbers to work with and B) arrive at the correct solution.

9 women can't make a baby in a month.. it takes one woman 9 months..

Yes, e-mail will function on a single core. But, as with other programs, it's effect on the system is less noticeable if it's workload is dispersed across multiple cores.
i had something to say about this but i'd rather quit talking about mail.app and multicore in the same sentence.. i think it's adding confusion.


You're being thick on purpose. Give it up

:rolleyes:
 
Apple hasn't done anything specific to spite you; what they haven't done is go back and write new 64-bit boot EFI firmware for those machines (new ones came with 64-bit EFI). A bummer, but to be fair it would probably have meant not only creating the EFI, but 64-bit device drivers for the old video cards and whatnot in those machines. They may just have said too low-level work and set those machines adrift, or there may be some technical limitation that makes it infeasible to update that generation hardware to a 64-bit EFI.
Not quite. At least one beta version of 10.8, and maybe all of them, did work on the 2006 Mac Pro. Yes, booting and video included. That's why it has been relatively straightforward for hackers to assemble kernels which do allow booting and running the 2006 Mac Pro with post 10.7 OS versions. Search the web, and you'll find plenty of evidence of this.

So Apple already had all the 32 bit EFI code and video driver code running just fine. But at the last minute, Apple realized that their need for more money via planned obsolescence was greater than the needs of the many people who had purchased the original Mac Pro.

Other models suffered the same fate. There were many, many 64 bit Intel Macs sold which had the same 32 bit boot architecture, and the solution for the Mac Pro would've worked for them as well. But Apple wanted more short term cash to the detriment of both its long term health and the interests of its customers.

And yes, I have a 2006 Mac Pro. I won't be burned again.
 
what point? there are a whole lotta points being strewn about this thread.


if that's the point then the point is wrong.. srry



i think you might not understand parallel processing.. it has nothing to do with how 'ridiculously simple' a program might be... for example, here's a ridiculously simple equation--- 6 -2 +4 ÷8 = 1
if you try to process it in parallel, it will be
thread1: 6-2=4
thread2: 4÷8=.5
combine threads-- 4.5

so the multithreaded version is giving the wrong result.. most processes are linear and can't be broken down in little pieces then add the results at the end.. or often, an algorithm which needs to run requires input from a previous calculation so again, the process needs to occur in a linear sequence in order to A) have numbers to work with and B) arrive at the correct solution.

9 women can't make a baby in a month.. it takes one woman 9 months..


i had something to say about this but i'd rather quit talking about mail.app and multicore in the same sentence.. i think it's adding confusion.




:rolleyes:

You are the determined one aren't you. Thank you for proving that multicore / multithreading is impossible...

Love your anatomical example of electronic processing... however, don't think you'll get 9 women to agree to test your theory... But, if you can get one to develop a leg, another to develop the other leg, and another to develop an arm, etc... then we'd be back on path to multi core processing... just hope the smart one develops the brains, and the central hub (doctor) is good with stitches.

You cannot use a pregnant woman or a single mathematical equation to claim the multicore / multithreaded operations are impossible. Also remember that computers do not think like people... There is no number 8 inside a computer chip's brain. Sorry.

It is all translated into our language so that we can easily interact with it. The computer does not process information in any way that resembles what we see and read and think.
 
You cannot use a pregnant woman or a single mathematical equation to claim the multicore / multithreaded operations are impossible.

i never said they were impossible.. what i said was that:
"
The point is, that most applications are designed to utilize multiple cores. Therefore, most applications will gain benefit from multiple cores."


.. is wrong.
most applications aren't designed to utilize multiple cores.. (and further, most processes can't be parallelized.. not with the hardware/algorithms being used up to today in computerworld)

don't take my word for it.. this is googleable.
 
Funny. Dude doesn't understand that 4.5 was the correct answer because of Order Of Operations (OOO). And dude also doesn't seem to understand the concept of divided-and-conquer. Too much.
 
Not quite. At least one beta version of 10.8, and maybe all of them, did work on the 2006 Mac Pro. Yes, booting and video included. That's why it has been relatively straightforward for hackers to assemble kernels which do allow booting and running the 2006 Mac Pro with post 10.7 OS versions. Search the web, and you'll find plenty of evidence of this.

So Apple already had all the 32 bit EFI code and video driver code running just fine. But at the last minute, Apple realized that their need for more money via planned obsolescence was greater than the needs of the many people who had purchased the original Mac Pro.

Other models suffered the same fate. There were many, many 64 bit Intel Macs sold which had the same 32 bit boot architecture, and the solution for the Mac Pro would've worked for them as well. But Apple wanted more short term cash to the detriment of both its long term health and the interests of its customers.

And yes, I have a 2006 Mac Pro. I won't be burned again.

I'm not an expert, but as an interested kibitzer I suspect that any early test release of 10.8 that booted on those Mac Pro's was merely not enforcing the 64-bit kernel requirement, and actually running the 32-bit kernel as Lion did. I think this stopped with Developer Preview 2 - after that there would have been no more 32 bit kernel included in the builds. as dropping support for that would have been an explicit goal, just like gradually removing Rosetta PPC support was. EFI is not part of the OS, it is part of the machine (it is the successor to the BIOS), and if Apple had a 64-bit EFI they would have had to distribute it as a firmware update. I couldn't find an authoritative reference, but it seems there is a chance the boot EFI was not flashable and they couldn't update it anyway.

My understanding is that people running 10.8+ on these old 32-bit EFI Macs are in effect doing a "reverse-hackintosh". They bypass the EFI and boot in Legacy BIOS mode (I think using a different partition), and run a loader which boots a Mac partition in 64-bit mode. And even this doesn't deal with drivers - I've seen references to old nVidia cards having to be replaced, issues with audio etc. This is nothing Apple would have any interest whatsoever in dealing with.

Apple certainly does make certain decisions that tend to increase margins, reduce flexibility, etc. But there isn't some moustache-twirling executive going "let's build a check into Mountain Lion just to exclude people from 2006 and make them buy new hardware"... :)
 
Last edited:
Funny. Dude doesn't understand that 4.5 was the correct answer because of Order Of Operations (OOO). And dude also doesn't seem to understand the concept of divided-and-conquer. Too much.

He also apparently was short sighted in his example and needed a 3rd core, cause he didn't have anywhere to put the additional number 4 in his equation.

lol, order of operations... so 2nd grade... yet lost on someone who supposedly understands multithreaded operations and why they cannot be used to do what we do with multithreaded operations every day.

Sends him over to apply for job with Cyrix... Hears they need the help of a guy like him to launch the revival of a has been CPU line. Thinks the timing could be telling... didn't they disappear before the whole multicore thing? If so, this guy could help bring them back... He understands better than anyone that multiple cores aren't really useful. He's just the man to bring back the defunct Cyrix, and market it's single core power to the world.

Hopefully, no one will throw any double negatives, or x and y variables at him, his brain might explode.

Now, before Flat Five gets into a circle of history about Cyrix processors, yes, I am aware and extremely familiar with their partnerships, acquisitions, and elimination. :rolleyes: Don't bother trying to twist a wiki into the matter, or you'll likely end up trying to convince me that a double positive equals a negative if a double negative equals a positive :rolleyes:
And, I don't want to cause Flat Five to suffer an Exponential Problem from his and / or operand
 
like i said.. don't take my word for it.. this info is readily available elsewhere on the web..
don't go educate yourself though.. this little tweedle dee tweddle dum routine you two have going is pretty funny to watch.
 
Yeah, nice try, but I was a CS major and I started multi-threaded programming over 22 years ago. So, maybe you should google something.
 
Yeah, nice try, but I was a CS major and I started multi-threaded programming over 22 years ago. So, maybe you should google something.
You're just a youngster. Multithreading has been around since at least s far back as the 1960s with the Control Data 6000 series of mainframes where each "core" took up a couple of floor-to-ceiling racks. Yes, I wrote system code for such a machine.
 
like i said.. don't take my word for it.. this info is readily available elsewhere on the web..
don't go educate yourself though.. this little tweedle dee tweddle dum routine you two have going is pretty funny to watch.

I'm sure there are a lot of pages out there written from people with perspectives like yours. Not worth the time to go and find them.

I could also search the Internet and find pages that argue that the moon is made out of cheese. Does that make it true? No.. Is it worth my time to go and look them up and cite them? No...

I have several machines which I use... I have single core, dual core, and some quad cores... I have dual cores available to me which are newer technology (only 3 months old) than the quad cores I use (8+ years old) and of same clock speed. I can tell you that in everyday use, doing the same tasks, that the dual cores are noticeably slower than the quad cores. And, the single core machine operating at a speed faster than any of the other machines feels like a sluggish dog coming out of anesthesia by comparison to both the dual and quad cores.

This is on same tasks, same programs, etc. And, if you think that it's not noticeable on simple programs like web browsers, think again... My kid enjoys some of the online children's sites, and complains about long page drawing delays, spinning wheels, etc. when using the single core machine (despite it's much faster clock speed). Using the same web site on the dual core machine is better, using the same web site on my quad core machine is smooth (despite the quad core being the oldest and slowest in the above examples).

----------

You're just a youngster. Multithreading has been around since at least s far back as the 1960s with the Control Data 6000 series of mainframes where each "core" took up a couple of floor-to-ceiling racks. Yes, I wrote system code for such a machine.

Shhh.... quiet, be careful there... you don't want to accidentally give the impression that even back in the 1960's that it was understood that multithreading and multiple cores known to be able to perform tasks more efficiently...

Sheesh... Like NASA knows anything... I bet we'd have gotten to the moon faster if they relied on only one core and one thread at a time.

You know it's funny, was in a conversation earlier on a totally different and unrelated topic about how computers used to fill rooms. Would have loved to have been there. I bet it must be weird to go into a room that used to be wall to wall hardware and see a little PC sitting on a table :D
 
You're just a youngster. Multithreading has been around since at least s far back as the 1960s with the Control Data 6000 series of mainframes where each "core" took up a couple of floor-to-ceiling racks. Yes, I wrote system code for such a machine.

Ouch. You've been around since cores were actually cores. If I was an engineering major when I first started college I would have done punch cards for a few years, but I managed to avoid that.
 
My understanding is that people running 10.8+ on these old 32-bit EFI Macs are in effect doing a "reverse-hackintosh". They bypass the EFI and boot in Legacy BIOS mode (I think using a different partition), and run a loader which boots a Mac partition in 64-bit mode. And even this doesn't deal with drivers - I've seen references to old nVidia cards having to be replaced, issues with audio etc. This is nothing Apple would have any interest whatsoever in dealing with.

The point is that third party hackers got 10.8 and later running with no inside help from Apple. If Apple had had a proper interest, then that interest would've been to support the purchasers of the original Mac Pro -- these purchasers also being the ones providing the money for future Mac Pro models and future OS/X development.

The video is another issue which Apple again failed its earlier customers. By not providing updated drivers which would have allowed forward compatibility, Apple doomed early Mac Pro buyers, prohibiting them from getting nearly any video card upgrade. If you've tried to find a moderately priced video card upgrade for the 2006 Mac Pro, then you'll know what I mean.

Apple's planned obsolescence: The curse which keeps on cursing.
 
In the meantime, the 2012 refurb mini dropped in price. The i7 2.3 mini is now $589, down from $679 just a month ago. Thank you apple! My 2012 refurb mini is on the way.

I have not found any at that price. Where did you get yours from?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.