Apple recently performed yet another in their long series of unorthodox decisions. There is now a new model of the iMac. It is $1099 and it can be described quickly by saying it is a base model MacBook Air but in a desktop chassis with a 21.5-inch display.
As always, the people who know just enough to be dangerous have lots of Opinions™ on the matter, most of which are colored by false notions about what Apple has actually done. But before I get ahead of myself, let's list out the specs of what you get for $1099, and what a buyer can customize.
- 1.4GHz Core i5-4260U processor (up to 2.7GHz turbo-boost)
- 8GB memory
- 500GB spinning hard disk (customization options include 1TB Fusion drive, and 256GB SSD.)
- 802.11AC wireless networking
- Choice of wired or wireless keyboard, and pointing device (can get two wireless pointing devices for extra)
- Mac OS X 10.9, iWork, and iLife software is included with the machine
- 1920x1080, IPS, LED backlit display, well known to be one of the most beautiful displays out of the box
Some of those points are well known, but it's important to note what you're getting for your $1099 when compared to other all-in-one machines for half the price, or when putting your focus on achieving computing performance, and putting your money toward a small tower build with the best CPU and video card you can get.
I think what commenters who suggest things are often forgetting is that if you're in an Apple store shopping for iMacs, you've kind of already given up on building out a high end gaming computer or spending under $600 on your all-in-one computer. At that point, your main choices are the budget conscious machine, with its relatively fixed configuration, the small machine, or the very big iMac, and from Apple's perspective, it really seems to be about how much money you're willing to throw at it. In the Mac world, money is almost always directly related to performance and capacity. Apple will happily tell you how much money is needed to attain a certain level of performance, or how much performance you can get for a certain amount of money.
Contrary to most of the comments, Apple is not stupid and knows fully well what they're doing. To that end, it's actually interesting to note that the $899 MacBook Air and the $1099 iMac are each using a higher performing CPU than the 2.3GHz CPU in the Mac mini for $599 and performs nearly identically to the 2.5GHz processor in the $1099 MacBook Pro. Apple also solders on the maximum choice of 8 gigabytes of memory with the new iMac. For $799, there is a quad-core Mac mini, but that choice implies a focus on performance, at which point, any of the quad-core iMacs are going to be a better choice.
Before I move on, let's talk about how impressive it is that a CPU whose base clock is 1.4GHz can outperform chips of just one generation prior whose base clocks are almost twice as high. To my knowledge, this is only in single-threaded benchmarks, and in multi-threaded benchmarks, the higher base clock of the older chips is going to take over. This is all essentially a result of just how weird Intel's Turbo-Boost technology is. Though, we could also see really different performance numbers in the iMac, because it's going to have much better cooling capabilities than the MacBook Air.
The related important note is that most mobile and home workloads spend most of their time idling or pegging a single core, so the solid state storage in the MacBook Air (or configuring this option in any iMac) is going to result in a bigger performance difference for a user than a CPU with more cores.
The other other important note is that a lot of people wanted Apple to put in a lower power desktop processor. There are a few challenges with this. The lowest-end iMac today is using an i5-4570R, which has the Intel IRIS Pro 5200 graphics built in. There are no other desktop processors from Intel with HD 5000, 5100, or 5200 graphics, and Apple isn't going to renege on the level of graphics performance that is available today for a cheap model, or for the power savings. If this model exists purely on the requirement from some large customer to build a low power system, then another desktop chip would also be unsuitable, because the lowest power desktop chips Intel has are 35 watts. It's a good savings over a 65W or 84W chip in the existing iMacs, but it's not as good as Apple can do. Even if Apple did choose to drop a 35 watt Celeron, Pentium, or i3 CPU into an iMac, they'd need to install a discrete GPU in order to get reasonable graphics performance, bringing them back to approximately the same power usage and overall cost as the i5-4570R iMac.
The other notion I think is really important that that desktop computing as a whole as whole started to really visibly separate into tasks that are limited by available computing horsepower, and tasks that are limited by the human driving the computer. Most home computing tasks fall into the latter group, which is one of the big reasons knowledgeable users tend to think it's reasonable to keep their machines for so long (1). I think the problem is that people really want their tasks to fall into the latter group. The more somebody knows (or thinks they know) about computers, the more likely they are to want their tasks to need as much horsepower, which is I think the source of the people who talk about the $1,299 iMac as though it's a universally better deal, even though it's unlikely for many people to really require 8 gigs of memory or a quad-core processor for the foreseeable future.
This comes down to priorities again, but the other thing people forget about Macs is that it has been quite a long time since Macintoshes were meant to be bought by people who are going to use the same computer for more than about four or five years before recognizing that everything around the computer itself has changed enough that the computer is now holding back whatever they were doing. The thing about "upgradeable" computers, in this respect, is that often by the time it's necessary to upgrade something like the CPU, nothing that fits the socket you have (unless you cheaped out pretty hard to begin with) will provide enough upgrade to merit keeping the machine. Of course, other factors (such as the performance of the storage subsystem, and being able to move from a dual-core CPU (not that any good duals exist for desktops anymore) to a quad-core CPU and being able to replace the graphics card are potential advantages, but those are also tasks that most iMac users do not want to be able to do.
I'm not going to buy one for myself, but I do believe that the new iMac is a good step for Apple. It may represent a permanent decrease in the dollar price on buying an iMac (which has hovered around $1,299 for sixteen years now) and I think it represents the way that many iMac owners, especially those purchasing at the low end of the spectrum anyway, actually use their machines. Of course, I think Apple missed a great opportunity to reduce their cost to manufacture the system even further by removing the hard disk choices, and using the 256GB flash module as the default storage in the system. In that configuration, it would likely feel faster than the 2.7GHz quad-core iMac to most buyers, and it makes sense for Apple to be at the front end of the push for solid state disks.
The other thing that's not yet clear, but likely will be when iFixit (the kings of political statements about computer reparability) gets their hands on the machine, is whether or not there's a fan in it. I think that using such an ultra-low voltage CPU in a system like the iMac has the potential to improve overall performance, as the iMac will have better thermal management. Apple could also take the opportunity to build a totally silent desktop, and I bet either would receive its due commendations.
It'll be interesting to see if any sites review this model, and if so, what professional reviewers say about it. Professional reviewers often have better notions™ about computing than Internet commenters, but they also suffer from sometimes misunderstanding a vendor's motives. (Which, can be difficult then they tell nobody what those motives are.)
1: I actually think it can be reasonable to keep a system for "so long," but it's also important to be aware of changing needs that might prompt the replacement or upgrade of a system, and either keep an old OS off the network, or stay abreast of OS versions.