[P]

Originally Posted by Simon 

But hardly one that withstands closer inspection.

I don't doubt Steve likes the AIO iMac and hates "generic" desktops. But we all know that Steve is all about business. If he knew he could earn a lot off of an xMac he'd do it. Therefore, the question I'm most interested in debating is why Apple thinks such a Mac couldn't make a lot of moeny. Especially considering that its parts are cheaper and it sells to a crowd entirely out of Apple's reach right now.

Apple's prices from Intel have little to do with the prices you see. There are volume discounts, and Intel even makes special models available to Apple. Apple's saving on moving to an Allendale or whatever its 45nm successor is called will be much smaller than what you'd calculate from the official prices. I do think that that might happen, but it will be in the form of the mini moving to Allendale and 3.5" drives.

Originally Posted by Simon 

<margin similar to Mac Pro>
That would only be true if they sold only as many xMacs as MPs. Considering everybody likes to point out how the xMac would kill the iMac I'd say the xMac would sell at least as many units as the iMac which is way more than the MP. So even if margins on the xMac are 30% lower than on the MP, if you sell twice as many units you are easily making more money.

That's not the way to count it. The only way to count is to see how many more customers Apple (in total) will gain by adding such a model, and see how large the margin drop would be for existing customers dropping from an MP (I don't think that the iMac margin is that ginormous). The only way this works is if that number is huge, but they didn't sell that many when Apple tried it, which they did in...

Originally Posted by Simon 

<previous xMac attempts>

Before 1999 that is. And before anybody says Cube let's just remind ourselves that the Cube was again a special design made to be compact and quiet. It was not even close to an xMac. Expensive to manufacture, only one slot, at a reduced length actually, no extra drive bays, no adequate cooling system, not easily serviceable. Seeing that Apple's last true "experiment" with such a Mac is the PM G3 Desktop offered a decade ago, I'd say we have pretty much no adequate data (i.e. not ancient) to compare with.

There is a more recent example - the low-end Powermac G5, sold at the same price as a similarly specced iMac until summer 2005. That was your xMac, and it didn't sell.

Originally Posted by Simon 

<upgrades>
How many people updated their Mac minis? The mini's or iMac's boards can't be upgraded easily so nobody is doing it. The only thing you can do on a Mac mini is upgrade the CPU. But since chipsets and sockets change so quickly, this isn't really a long-term upgrade strategy either. I can't imagine Apple is holding off the xMac because they're worried about the three sales they'd lose due to upgrades on an xMac that wouldn't have been possible on an iMac or Mac mini. How upgradeable a Mac is, is pretty much up to Apple. That's just as true on the xMac as on the iMac.

Now I'm very confused. What would be the point of an xMac if you can't upgrade it? You can buy a mini and an external 3.5" HD and put them in a gray box, if you like.

Originally Posted by Simon 

Sure they prefer to lead. But sometimes they lead the wrong way. And in about half those cases they're smart enough to realize that they made a mistake. They correct it even if it comes at the expense of taking a "step back". Example are DVI instead of ADC, CDRW drives instead of pure DVD readers, USB on the iPod, ...

Neither of those were really a step back. DVI is barely a step back from ADC, and DVI wasn't a very popular standard back when ADC launched. CDRW drives replaced CDs in the low-end, and combos replaced pure DVDs on the path towards superdrives everywhere. USB 2.0 replaced Firewire 400 - USB 2.0 did not exist in silicon back when the iPod launched.

[Simon]

Originally Posted by P 

Apple's saving on moving to an Allendale or whatever its 45nm successor is called will be much smaller than what you'd calculate from the official prices. I do think that that might happen, but it will be in the form of the mini moving to Allendale and 3.5" drives.

I'm not calculating anything. I'm acknowledging the simple business fact that every Dollar Apple saves on parts is one more they make as profit.

That's not the way to count it.

Of course it is. The only thing that counts is how much money Apple makes. And if they win customers they can even with a reduced margin make more money at the end of the day. It's not like Apple would lose Mac mini customers to Dell just because they happen to start also selling an xMac.

There is a more recent example - the low-end Powermac G5, sold at the same price as a similarly specced iMac until summer 2005. That was your xMac, and it didn't sell.

Actually that wasn't an xMac. It was a single-CPU PM. It was still big and expensive to manufacture. And actually, why don't you link to those sales numbers, because I don't recall Apple ever publishing how many of them they sold. That said, that PM is being sold even today. It's the quad-core MP at $2299. For obvious reasons that's not the Mac the OP or us others here have been talking about.

Now I'm very confused. What would be the point of an xMac if you can't upgrade it?

It doesn't exist so we don't know how upgradable it would be. My point was that its upgradability would be at Apple's discretion just like on any other Mac.

You can buy a mini and an external 3.5" HD and put them in a gray box, if you like.

I still wouldn't have a graphics card.
I'd still be limited to 4GB of RAM.
I still wouldn't be able to directly connect any eSATA disk or optical.
I still couldn't use any other PCI card.
And most importantly I'd still have a slow Mac that's too expensive to manufacture because it was designed to be super small and quiet.
So yeah, as obvious as that might have been, it's not quite what we were talking about.

I still haven't heard any good business reasoning why Apple thinks it can't make money on an xMac.

[CharlesS]

Originally Posted by P 

Of course it would grow - it's almost zero - but the market for expansion cards on Windows is also dropping. The old days where you needed a sound card, a network card, a "game card", a graphcis card, a controller card and whatever are gone. Few need any cards, and those that do need one specific one.

With miniaturization getting ever more effective and the desktop share of the market sliding, it makes sense to put even more stuff into the southbridge, further limiting the expansion card market. Pretty soon it's only GPUs, and even discrete GPUs are dropping as a share of the market. I think this type of internal expansion is heading for the grave on all non-specialist platforms.

Again, if something is dropping, that does not mean it is heading for zero - it is more likely that it is heading for a new equilibrium point somewhere. There is always going to be a need for expansion, whether it's to upgrade your GPU, to add some functionality that Apple didn't anticipate you needing (*cough* FireWire *cough*), to add some functionality that didn't exist at the time you bought the machine (USB 3.0), to replace something in your machine that has stopped functioning properly (blown USB, Ethernet, FireWire ports), or to connect specialized hardware you need for whatever niche you work in. It's just a general safety net against the future. Some people are always going to want/need this stuff, especially in situations like a business or a technology lab or something where you have a lot of machines, and the probability that some of them are going to develop some kind of problem goes up significantly, and just buying a $20 USB card to replace a blown-out USB controller turns out to be a lot cheaper (and greener!) than junking the whole machine.

Almost every desktop PC made today includes some kind of PCI or PCI-Express expansion. That hardly indicates to me that the market in general is going to dump them anytime soon.

Originally Posted by Simon 

It's not like Apple would lose Mac mini customers to Dell just because they happen to start also selling an xMac.

Well, I'd guess what would happen would be that the xMac would significantly cannibalize sales of the Mac mini, which of course wouldn't bother anyone as the xMac would almost certainly have far higher margins than the mini.

[P]

That last line sums up this entire debate: There isn't a business reasoning why they CAN'T. They don't WANT to do it, and there is no solid business reasoning why they SHOULD. Every argument comes down to "they would expand their market" and that is nowhere near proven - in fact, the given evidence suggests otherwise.

Actually that wasn't an xMac. It was a single-CPU PM. It was still big and expensive to manufacture. And actually, why don't you link to those sales numbers, because I don't recall Apple ever publishing how many of them they sold. That said, that PM is being sold even today. It's the quad-core MP at $2299. For obvious reasons that's not the Mac the OP or us others here have been talking about.

It was expensive to manufacture? Why? It was a tower. According to you, that's as cheap as possible. If you expected a grey plastic case, then I'm sorry - that's not going to happen (and the cost saving would be tiny). It used what was probably the best price/performance Mac CPU of its day and off-the-shelf parts. And there is a quiet significant difference between $1599 and $2299. What feature of it was not what you wanted? From what I hear, it's only the price. This is where this argument degenerates - people basically want a Dell-priced box that they can stuff full of exactly what they want, and there's no way Apple will ever make that. If that's what you want, build a hackintosh.

And no, the current Mac Pro is not the same as what you'd make as a hackintosh. I'm aware of that, and it is a worse xMac than that PM G5 ever was.

<xmac upgradeability>
It doesn't exist so we don't know how upgradable it would be. My point was that its upgradability would be at Apple's discretion just like on any other Mac.

OK then, what specific feature would you say would differentiate your xMac from an iMac? Reading below, it doesn't have a display, and it has upgradeable graphics. Possibly a PCI slot would be nice. Apple can add upgradeable graphics and limited expansion to the iMac any day they want to. They choose not to.

I still wouldn't have a graphics card.

OK, you want a GPU.

I'd still be limited to 4GB of RAM.

As are many PCs in effect, if not in theory, as they're sold with a 32-bit OS.

I still wouldn't be able to directly connect any eSATA disk or optical.

Stretching. USB works fine, and you can pull off the lid of the Mac mini in that box and connect a big SATA internally if you like.

I still couldn't use any other PCI card.

For which there aren't drivers, anyway, but that again isn't a feature per se. Name a specific thing you'd like a PCI card for that you cannot get over USB.

(I recently looked around a bit for a TV capture card for Windows, and realized that they're mostly USB now. Apparently noone likes to install stuff inside their boxes anymore.)

And most importantly I'd still have a slow Mac that's too expensive to manufacture because it was designed to be super small and quiet.

So price.

Basically, you'd like a cheaper Mac that you can put a powerful GPU in. For price, I guess you'll have to ask yourself if the Apple tax is worth it. Whatever added engineering cost might have gone into a Mac mini is a small part of the price increase. For a GPU, Apple could do that in the iMac tomorrow if they chose to.

Just because I'm curious: If Apple launches a new Mac Pro line at Macworld, using regular DDR3 RAM, that can be configured with a Core i7 920 for $1599 with a decent midrange GPU, would you buy that? If not, why?

[CharlesS]

Originally Posted by P 

For which there aren't drivers, anyway, but that again isn't a feature per se. Name a specific thing you'd like a PCI card for that you cannot get over USB.

FireWire?

eSATA?

USB 3.0?

FW3200?

The next, faster version of SATA?

SCSI (yes, some people still use it as it's very high-performance)?

High-end audio cards?

Replacements for your blown-out USB ports (try that over USB )?

And need I mention, GPUs?

[Big Mac]

Just because I'm curious: If Apple launches a new Mac Pro line at Macworld, using regular DDR3 RAM, that can be configured with a Core i7 920 for $1599 with a decent midrange GPU, would you buy that? If not, why?

I bet that would be a very popular Mac, P.

[CharlesS]

Originally Posted by P 

Just because I'm curious: If Apple launches a new Mac Pro line at Macworld, using regular DDR3 RAM, that can be configured with a Core i7 920 for $1599 with a decent midrange GPU, would you buy that? If not, why?

Well, I can't speak for everyone, because I'm a laptop user myself, but I think part of the problem with the Mac Pro is its size, not just its price. It really is a huge, heavy monster machine, and includes a lot of stuff that makes it kind of overkill for many. For example, most people don't need 4 drive bays. I think if you lose two of the extra drive bays, one of the PCI-Express slots, the second Ethernet port, the FW400 ports (leave FW800 in though, please), then you could have a basic tower with two PCI-Express slots, two hard drive bays, and a normal complement of ports, and if you add in the lower amount of cooling you'd need for a regular desktop processor instead of a Xeon, you could probably fit this into a quite smaller case and have yourself a nice little Mac mid that would probably win over a decent number of PC desktop users. Just my opinion, though.

[Simon]

Originally Posted by P 

That last line sums up this entire debate: There isn't a business reasoning why they CAN'T. They don't WANT to do it, and there is no solid business reasoning why they SHOULD.

Up to the very last part I tend to agree. The last part however is only your assumption and quite frankly I don't buy it. New customers are a good thing. More revenue is a good thing. So before I buy that line you're gonna have to show us how an xMac decreases revenue and makes them lose customers.

Every argument comes down to "they would expand their market"

and could increase revenue becaue of reduced cost.

and that is nowhere near proven - in fact, the given evidence suggests otherwise.

No such "evidence" has been given.

It was expensive to manufacture? Why? It was a tower.

The tower isn't the point. The point is that it's an expensive case, it offers far more expansion than needed in the mid-range area, and it uses a CPU and chipset much more expensive than necessary.

According to you, that's as cheap as possible.

Nowhere did I say that.

The MP is expensive because it offers way to much compared to the system the OP has proposed. The CPUs are twice as expensive as Wolfdale/Yorkfield, the chipset is more expensive, the case is huge and expensive. There's a cost for the MP's massive expansion options. Remove all those fancy extras and you end up with a fairly inexpensive mid-range box. Take the iMac. Decent performance, decent price. The proposed box would offer more performance and yet cost Apple less to make. Keep the price the same and Apple has just made more money. Yes, it is as simple as that.

If you expected a grey plastic case, then I'm sorry - that's not going to happen (and the cost saving would be tiny).

Actually I was thinking of Al box. The cost savings come from the components and the reduced expansion options. It has been detailed several times what those would be. Why are we going over this again and again?

It used what was probably the best price/performance Mac CPU of its day and off-the-shelf parts. And there is a quiet significant difference between $1599 and $2299.

Well $500 alone come from the CPU. The remaining $200 are easily saved with the chipset (~$75) and the case.

This is where this argument degenerates - people basically want a Dell-priced box that they can stuff full of exactly what they want, and there's no way Apple will ever make that. If that's what you want, build a hackintosh.

Nope. The Dell box is $600, plastic, and sucks. We're talking Macs here. People want a Mac with desktop level performance and a swappable graphics card plus maybe an extra slot or bay. Right now the only way to get that is the $2299 MP and that is out of the question. It's a huge overkill. And what the OP and others here have done is to illustrate how Apple could do a Mac that offers exactly these typical mid-range features for significantly less. What you so far have failed to do is illustrate why Apple thinks they can't make money with such a box.

OK then, what specific feature would you say would differentiate your xMac from an iMac?

Once again then:
- desktop CPU (Wolfdale, Yorkfield)
- desktop chipset (Bearlake), >4 GB RAM (not an issue with OS X)
- exchangeable GPU
- exchangeable screen
- extra PCI slot
- extra disk bay
- eSATA

Basically, you'd like a cheaper Mac that you can put a powerful GPU in.

Nope. The iMac's price is fine. What I want is better performance per Dollar, exchangeable graphics, etc. Seriously P, I know what I want and I don't need you to rephrase it for me. The point is that it's technically clear how to build a headless iMac that offers better performance and better margin at the same time. The question that has not been answered is why Apple doesn't want to do it. And all I've heard on this so far is sketchy speculation. So how about we just stop arguing semantics here and continue the debate once somebody has some actual insight as to why Apple would not want to add a new line that drives more revenue. So far none has been offered. And you telling others that Apple knows better what they want than they themselves is a silly debate. We've had all these before. I'd like to see something new for once.

Just because I'm curious: If Apple launches a new Mac Pro line at Macworld, using regular DDR3 RAM, that can be configured with a Core i7 920 for $1599 with a decent midrange GPU, would you buy that? If not, why?

Depends. The price seems reasonable, a bit on the low end actually. The CPU and its chipset would fit what I'm asking for. Can I swap the GPU? How many slots? How many bays? The same case as the current MP or something more compact?

[CharlesS]

Originally Posted by Simon 

The question that has not been answered is why Apple doesn't want to do it.

Answer: it's ideological.

http://www.folklore.org/StoryView.py...rt%20by%20Date

[Simon]

I get that. And at least in that point you and P seem to agree.

But don't any of you other people here find it astonishing that a company so good at making money is ready to pass up such a chance just because Steve doesn't fancy expansion? For all we know that little grudge could be costing the company millions of Dollars every year.

That leads to another question. When Steve leaves and Tim takes over, can we expect an xMac within a year?

[CharlesS]

One would hope so, although it's a shame that something as basic as this can only be obtained by losing the many other qualities Jobs brings to Apple.

[P]

Originally Posted by CharlesS 

FireWire?

eSATA?

USB 3.0?

FW3200?

The next, faster version of SATA?

Neither of these are features per se either, they are interfaces to connect stuff. Name a usage that is seriously limited by a USB 2.0 interface.

Originally Posted by CharlesS 

SCSI (yes, some people still use it as it's very high-performance)?

High-end audio cards?

High-end stuff that works well in the high-end Mac. Business workstation was the thing.

Originally Posted by CharlesS 

Replacements for your blown-out USB ports (try that over USB )?

They're called "Hubs"... Seriously, repairs.

Originally Posted by CharlesS 

And need I mention, GPUs?

Which is one of the two things that we've agreed upon, the other being price.

My point here is that expansion of this kind is much less useful now than it used to be, due to better external interfaces and a larger amount of built-in features in the southbridge. Sure it's nice with faster connection standard, but it works as it is.

[CharlesS]

Originally Posted by P 

Neither of these are features per se either, they are interfaces to connect stuff. Name a usage that is seriously limited by a USB 2.0 interface.

DV camcorders, audio equipment, hard drives - you name it, it works better and faster with FireWire or eSATA than it does with USB. And some devices aren't available for USB at all.

They're called "Hubs"... Seriously, repairs.

Yeah, a USB hub's really going to help when you plug it into a dead USB port. Seriously, WTF?

There's a bunch of problems with repairs. One, if the USB controller is toast, repairs generally entail replacing the entire logic board, which is a lot more expensive than a $20 USB card. Two, in labs and business settings where you are managing whole buildings full of computers, you're going to see some failures not all that infrequently. In this situation, it's advantageous just to be able to go up to the machine and fix it yourself than to have your staff constantly driving to and from the repair shop (especially if the nearest Apple store happens to be an hour's drive away). Usually, the IT staff has better things to do.

[P]

Originally Posted by Simon 

No such "evidence" has been given.

I'm referring to the fact that desktops are dropping as a share of the market, that said share is tiny in the price segment Apple likes to operate in and that the last "xMac-like" Mac did not sell. I don't see any evidence on your side of the balance.

Originally Posted by Simon 

The tower isn't the point. The point is that it's an expensive case, it offers far more expansion than needed in the mid-range area, and it uses a CPU and chipset much more expensive than necessary.

The current Mac Pro, yes. But I wasn't talking about that. I was talking about the 1.8 GHz G5, which used the same CPU as the then-midrange iMac. All the other stuff is thinsg that are too expensive, so...price again?

Originally Posted by Simon 

The MP is expensive because it offers way to much compared to the system the OP has proposed.

I know that. The MP is no xMac. The old single 1.8 GHz G5 was.

Originally Posted by Simon 

Take the iMac. Decent performance, decent price. The proposed box would offer more performance and yet cost Apple less to make. Keep the price the same and Apple has just made more money. Yes, it is as simple as that.

Which. Is. What. They. Did. And for some reason they stopped selling it after one revision. Wanna take bets on why?

Originally Posted by Simon 

Actually I was thinking of Al box. The cost savings come from the components and the reduced expansion options. It has been detailed several times what those would be. Why are we going over this again and again?

Because you're referring to the current MP, and I'm referring to the old PM G5 that I consider the example of an xMac that didn't sell.

Besides: My home-built HTPC has a brushed aluminium case with 7 3.5" places and 2 5.25" places. The motherboard has 7 slots, 3 PCI and, 3 PCIe x1 plus the GPU slot. I did not look around for that level of possible expansion, it was there is most cases of the type. I really did try to save money (I kept the entire deal under $1000 without the OS). I doubt that there is much to save by cutting down the level of expansion.

What I want is better performance per Dollar, exchangeable graphics, etc. Seriously P, I know what I want and I don't need you to rephrase it for me.

Well, I'm not you. I've been in these debates before when one of xMac requesters decides to completely change what they really wanted in the first place. I want to see what it is that's really missing from the current models.

I think that, with the exception of the 2.5" HD in the mini, you are overestimating the savings you'd get back from the cheaper desktop parts, and you're certainly overestimating any savings from a cheaper case. Apple makes a few models in large series - I doubt they spend more on the iMac or Mac mini case than Dell does on one of theirs (Dell cases aren't gray plastic anymore either).

Originally Posted by Simon 

Depends. The price seems reasonable, a bit on the low end actually. The CPU and its chipset would fit what I'm asking for. Can I swap the GPU? How many slots? How many bays? The same case as the current MP or something more compact?

It's the MP case - go nuts on what sort of graphics you want in it. I may have overspecced it a bit, considering the iMac specs at the moment, but say a midrange quad Penryn instead. I'm thinking about a do-over on that old single-CPU G5. So basically you'd buy that model, if this were 2005?

My point is that that old model was pretty much exactly the xMac, and it didn't sell - the likely reason being that the people begging for an xMac are a very vocal but small group.

[P]

Originally Posted by CharlesS 

DV camcorders, audio equipment, hard drives - you name it, it works better and faster with FireWire or eSATA than it does with USB. And some devices aren't available for USB at all.

Firewire is still there on the iMac - probably will be, as long as those DV camcorders work with Firewire.

Originally Posted by CharlesS 

Yeah, a USB hub's really going to help when you plug it into a dead USB port. Seriously, WTF?

There is usually more than one port to begin with.

Originally Posted by CharlesS 

There's a bunch of problems with repairs. One, if the USB controller is toast, repairs generally entail replacing the entire logic board, which is a lot more expensive than a $20 USB card.

Isn't the USB controller built into the southbridge for all modern boards? If that's toast, then few things will work.

Originally Posted by CharlesS 

Two, in labs and business settings where you are managing whole buildings full of computers, you're going to see some failures not all that infrequently. In this situation, it's advantageous just to be able to go up to the machine and fix it yourself than to have your staff constantly driving to and from the repair shop (especially if the nearest Apple store happens to be an hour's drive away). Usually, the IT staff has better things to do.

I worked part-time as an admin in the computer labs at university for about 4 years. I saw broken displays - a lot of them - broken laser fuses, a bad FPU (that was an old IRIX machine), lots of bad RAM, broken HDs, bad GPUs, any number of broken mice and keyboards but I don't think I ever saw a broken USB port. Or, for that matter, parallel, serial or PS/2 port - the older machines had those.

[P]

Originally Posted by Simon 

That leads to another question. When Steve leaves and Tim takes over, can we expect an xMac within a year?

That IS what happened last time - Sculley launched the Mac II.

I don't think it will happen as long as Ive is involved in Apple, though

[Simon]

I'll pick this line since I find it the most important one.

Originally Posted by P 

I want to see what it is that's really missing from the current models.

Exchangeable GPU.
Yorkfield/Wolfdale and Bearlake.
1 extra PCI slot
1 extra disk bay
No emphasis on ultra-small (mini) or ultra-thin (iMac), instead easily-serviceable, no external PS, etc. etc.
(is it just me or did I actually write down that list for about the third time in this thread?)

So let me ask you
• Do you agree that is something no present Mac offers?
• Do you agree there are computers being sold right now that match pretty exactly that description?
• Do you agree those desktop parts w/o space constraints make for a cheaper manufacturing costs than mobile parts crammed into small enclosures?

So then, can we agree Apple isn't doing this not because
• they don't know how to
• they can't make this box and sell it with 25% margin
but because
• they simply don't like it.

In which case (I'm assuming we agree so far) the question again comes up if Steve has to kick the bucket before we finally see this. And yes, I agree with CharlesS, it's a shame because otherwise (maybe with one or two exceptions) he's not a jackass.

[P]

Originally Posted by CharlesS 

Well, I can't speak for everyone, because I'm a laptop user myself, but I think part of the problem with the Mac Pro is its size, not just its price. It really is a huge, heavy monster machine, and includes a lot of stuff that makes it kind of overkill for many. For example, most people don't need 4 drive bays. I think if you lose two of the extra drive bays, one of the PCI-Express slots, the second Ethernet port, the FW400 ports (leave FW800 in though, please), then you could have a basic tower with two PCI-Express slots, two hard drive bays, and a normal complement of ports, and if you add in the lower amount of cooling you'd need for a regular desktop processor instead of a Xeon, you could probably fit this into a quite smaller case and have yourself a nice little Mac mid that would probably win over a decent number of PC desktop users. Just my opinion, though.

The idea was to base something on the current case. Developing a new case design isn't something Apple would do in a hurry, and there aren't any real savings in doing so. Besides, tower designs are limited by the width of the optical drive - unless you put that vertically, which I understand many don't like due to the decreased speed, that tower is going to be that wide.

Also: Apple uses Xeons from the 5400-series. They have roughly the same TDPs as similarly clocked desktop processors. A midrange Penryn Quad like the 9550 comes in at 95Ws - the three Xeon models Apple uses are rated at 80, 120 W and 150W TDP, respectively, the difference made up by the FSB speed. Of course, moving to a single-socket model would simplify things.

[Simon]

Something else bothers me about Steve here. When he came to Apple his restructuring was necessary. The four product matrix was a good thing. It helped make Apple profitable. It helped Apple regain focus. It helped Apple become big again.

But now is a completely different time. Apple is in a strong position. They have critical mass. They have tons of cash. And they have the chance to innovate out of a recession. So why the heck don't they just deliberately try to do new products? Just let the balloon fly. See what happens. What sells, what doesn't. Heck, they did it with the AppleTV. A box nobody really asked for. And it flopped. They tried improving it. It's still flopping. But they're still giving it CPR like they just can't call it. And Steve goes on about hobby? What a colossal load of baloney. If they're going to do hobbies, why not with Macs? Not just the xMacs, but other toy ideas like tablet Mac, netbook Mac, gamer Mac, whatever Mac. Instead the Mac still gets treated like it's 1999. What's the deal with that? Why is it that they're curious and daring about anything else (take the iPhone gamble for example) but get all conservative and cautious when it comes to the Mac?

[CharlesS]

Originally Posted by P 

Firewire is still there on the iMac - probably will be, as long as those DV camcorders work with Firewire.

You probably would have said that about the MacBook a few months ago. I wouldn't hold my breath.

And eSATA, of course, is not available on any Mac models at all.

There is usually more than one port to begin with.

If the controller is toast, then those might not work either.

And you generally don't want to have your USB 1.1 and USB 2.0 devices on the same port anyway.

Isn't the USB controller built into the southbridge for all modern boards? If that's toast, then few things will work.

I dunno, then substitute Ethernet or FireWire for USB - you're just as screwed if those blow out.

I worked part-time as an admin in the computer labs at university for about 4 years. I saw broken displays - a lot of them - broken laser fuses, a bad FPU (that was an old IRIX machine), lots of bad RAM, broken HDs, bad GPUs, any number of broken mice and keyboards but I don't think I ever saw a broken USB port. Or, for that matter, parallel, serial or PS/2 port - the older machines had those.

When I worked in the computer lab at university, I saw my share of blown FireWire ports (all of which involved both FireWire ports blowing out, not just one). We also had a machine with a USB PCI card in it, IIRC because its built-in USB ports were dead. And I sure as hell saw a lot of hard drives die - which are technically replaceable on the iMac, but it's definitely a lot more of a PITA than on any mini-tower machine. There's also the issue of the Ethernet port - that can die, and you don't even have a second one of those to fall back on most of the time.

It's just less of a maintenance hassle to be able to replace those.

[P]

Originally Posted by Simon 

I'll pick this line since I find it the most important one.

It wasn't meant to be. I'd rather talk about the single-CPU G5.

Originally Posted by Simon 

Exchangeable GPU.

OK, fair enough.

Originally Posted by Simon 

Yorkfield/Wolfdale and Bearlake.

Why? As in, which feature of those CPUs do you want that makes them better than a Penryn?

Originally Posted by Simon 

1 extra PCI slot

Just in case?

Originally Posted by Simon 

1 extra disk bay

That you don't really need. Bus-powered HDs are quiet, readily available, work well and can be hidden from sight if you don't like the ugly big box.

Originally Posted by Simon 

No emphasis on ultra-small (mini) or ultra-thin (iMac), instead easily-serviceable, no external PS, etc. etc.

The only one of those that make sense in itself is "easily serviceable".

Originally Posted by Simon 

(is it just me or did I actually write down that list for about the third time in this thread?)

Yes, you really ought to stop doing that...

Originally Posted by Simon 

So let me ask you
• Do you agree that is something no present Mac offers?

Precisely what you wrote? Of course. If you want the expandability you need to go to the MP, which is a large step up pricewise. Maybe you can get an iMac, leave the glass and the display off and put it in a gray box?

Originally Posted by Simon 

• Do you agree there are computers being sold right now that match pretty exactly that description?

As written in this post? Yes. But then you also wrote this, not too long ago:

Nope. The Dell box is $600, plastic, and sucks. We're talking Macs here.

So if you want all of the above with the elegance of a Mac, then I'm sorry - the answer is no, I don't know of a machine like that. I worked really hard to get my HTPC quiet enough for me to be happy, and I had to step down to a midrange GPU (4670) to make it do that.

• Do you agree those desktop parts w/o space constraints make for a cheaper manufacturing costs than mobile parts crammed into small enclosures?

Not significantly, no - except for a 3.5" HD, as already mentioned, and the iMacs have that. Intel's advertised prices for CPUs are widely different, but that doesn't mean that Apple would get a desktop CPU that much cheaper. The big difference is also on the highest clocked CPUs, the ones that would make a godawful noise in any box smaller than an MP. RAM prices seem to be the same for DIMMs and SO-DIMMs at the moment, so there's no reason to believe that production costs for the two lines are any different before packaging. Maybe you'd save $100 or so.

So then, can we agree Apple isn't doing this not because
• they don't know how to
• they can't make this box and sell it with 25% margin
but because
• they simply don't like it.

That has been my argument all along, so yes.

In which case (I'm assuming we agree so far) the question again comes up if Steve has to kick the bucket before we finally see this.

At this rate, the business desktop will kick the bucket before Steve does. Barring that, well, if the desktop market picks up for some reason, then I'm sure we'll see the awesomest sleek tower you could ever imagine.

[Simon]

Originally Posted by P 

At this rate, the business desktop will kick the bucket before Steve does.

In which case Steve would be the one to kiss the iMac goodbye? That should make for an interesting sight.

Barring that, well, if the desktop market picks up for some reason, then I'm sure we'll see the awesomest sleek tower you could ever imagine.

Too bad. I'm really not that much a fan of towers. I'd prefer a case like the PM6100 actually.

[CharlesS]

Originally Posted by P 

At this rate, the business desktop will kick the bucket before Steve does.

What makes you think the business desktop would ever "kick the bucket"? Consumers may be moving toward laptops, but how does that make sense for businesses and labs?

[P]

Originally Posted by CharlesS 

<Firewire still on iMac>
You probably would have said that about the MacBook a few months ago. I wouldn't hold my breath.

Actually, no. I was more surprised when they added Firewire to the iBook back in the day.

Originally Posted by CharlesS 

I dunno, then substitute Ethernet or FireWire for USB - you're just as screwed if those blow out.

Ethernet's also in the southbridge. Firewire is not, but I'm not sure that that makes much of a difference - I'd want to replace the motherboard anyway.

Originally Posted by CharlesS 

When I worked in the computer lab at university, I saw my share of blown FireWire ports (all of which involved both FireWire ports blowing out, not just one).

That's more common. Apple supplied less current over Firewire than certain devices required. I'm not sure who was right anymore.

Originally Posted by CharlesS 

We also had a machine with a USB PCI card in it, IIRC because its built-in USB ports were dead. And I sure as hell saw a lot of hard drives die - which are technically replaceable on the iMac, but it's definitely a lot more of a PITA than on any mini-tower machine.

Less PITA than replacing a laser fuse because some idiot put the wrong kind of transparencies in it.

The current iMac is more of a PITA than previous models, though, or so I understand, but for an HD crash you'd still have to reclone the entire shebang. It's still not a quick operation.

Originally Posted by CharlesS 

There's also the issue of the Ethernet port - that can die, and you don't even have a second one of those to fall back on most of the time.

Also a thing I haven't seen - well, other than the occasional bent pin that you can straighten out.

Sidenote: After graduating, I occasionally helped my uncle with supporting the computers at the school he was the principal of. My favorite error was a brand new computer with two errors listed: Internet does not work. Printer does not work. The reason? Someone had put the USB connector from the printer into the Ethernet port. They do fit, if you push hard enough.

OK, I see your point: a tower-style Mac would be easier to service, and that would be useful in a lab. So would an iMac with a modicum of accessibility, like the first iMac G5s. Not so sure that the PCI slots would be that useful, though, other than in special cases.

[Simon]

Originally Posted by P 

Why? As in, which feature of those CPUs do you want that makes them better than a Penryn?

Penryn costs more and offers less performance. It's a bad choice for desktops. Why? Because it was never meant to be in desktops.

Just in case?

Yep that extra slot comes in handy every now and then. Especially in transition phases. Example: not a single MB buyer would be whining about the omission of FW if the MB had an EC/34. The same is true for any other computer. Just in case = less suffering. In my book that's a good thing.

That you don't really need. Bus-powered HDs are quiet, readily available, work well and can be hidden from sight if you don't like the ugly big box.

They also require bridge chips, extra PSs, etc. All the ugly things that are totally un-Apple. And again, they're more expensive than getting any bare 3.5" HDD and putting it into a ready carrier. If what you said were an acceptable option, the MP wouldn't have it's four bays. On the mini it's omission is part of the mini design. On notebooks it's a necessary trade-off to reach slim and light enclosures. But on a desktop it's exclusion is unacceptable.

The only one of those that make sense in itself is "easily serviceable".

No, you see, by going super-mini and super-thin Apple has brought notebook limitations onto the desktop. That's a shame because many desktop users rather have options than notebook compromises. This is Apple. The user is important. Compromise used to be a PC thing.

Not significantly, no - except for a 3.5" HD, as already mentioned, and the iMacs have that. Intel's advertised prices for CPUs are widely different, but that doesn't mean that Apple would get a desktop CPU that much cheaper.

Let's stop this nonsense, shall we. Even at volume discounts Apple is paying $400 more for a MP CPU than a similarly clocked dual-core or slightly lower-clocked quad-core desktop CPU. They're paying extra for the Xeon chipset. They're paying for 4 PCIe slots. They're paying for 32 GB RAM support. They're paying for the extra Gigabit. But actually, it doesn't really matter how much they pay extra. Fact is every Dollar they save in production is one more they make in revenue. Steve's little grudge is costing money. And as a stockholder that ticks me off.

[P]

Originally Posted by CharlesS 

What makes you think the business desktop would ever "kick the bucket"? Consumers may be moving toward laptops, but how does that make sense for businesses and labs?

I'm just extrapolating the current trend towards infinity...

My theory goes like this: chips get smaller, and they cannot get any hotter - therefore they will have to use less power. For each shrink, it makes even less sense to put them in a big ATX box - the failure of alternatives like BTX and DTX make that worse (I know that some OEMs still use BTX, but I haven't heard a word about future development of it). The price difference between a desktop and a laptop will shrink to the cost of the display, and at that point, there is money in standardizing on a laptop form factor - especially if SSDs conquer the world, as I suspect.

[P]

Originally Posted by Simon 

In which case Steve would be the one to kiss the iMac goodbye? That should make for an interesting sight.

It's a desktop replacement-replacement!

Originally Posted by Simon 

Too bad. I'm really not that much a fan of towers. I'd prefer a case like the PM6100 actually.

I liked my old LC III. I kept that Mac around for a long time, thinking I'd build something into it, but I eventually realized that I'd have to put the PSU externally. The plan died at the point...

[P]

Originally Posted by Simon 

Yep that extra slot comes in handy every now and then. Especially in transition phases. Example: not a single MB buyer would be whining about the omission of FW if the MB had an EC/34. The same is true for any other computer. Just in case = less suffering. In my book that's a good thing.

Slots actually make less sens for a desktop where you have a desk to put extra boxes on. And Apple could put an EC slot on the iMac - they used to have a PCCard one, after all.

Originally Posted by Simon 

They also require bridge chips, extra PSs, etc. All the ugly things that are totally un-Apple. And again, they're more expensive than getting any bare 3.5" HDD and putting it into a ready carrier.

A price to pay for the accessibility of not having to pull the machine apart to get it to work.

Originally Posted by Simon 

If what you said were an acceptable option, the MP wouldn't have it's four bays.

They're there for RAID, and because at the size the case you have to be to fit the Xeon cooling and slots, you might as well.

Originally Posted by Simon 

On the mini it's omission is part of the mini design. On notebooks it's a necessary trade-off to reach slim and light enclosures. But on a desktop it's exclusion is unacceptable.

And yet people buy them.

Originally Posted by Simon 

No, you see, by going super-mini and super-thin Apple has brought notebook limitations onto the desktop. That's a shame because many desktop users rather have options than notebook compromises. This is Apple. The user is important. Compromise used to be a PC thing.

A large case and noise are not compromises?

Originally Posted by Simon 

Let's stop this nonsense, shall we. Even at volume discounts Apple is paying $400 more for a MP CPU than a similarly clocked dual-core or slightly lower-clocked quad-core desktop CPU.

No. Don't bring it back to the MP. The question I answered was if Apple would save money by moving from laptop parts to desktop parts. They would certainly save money by moving from low-volume server parts to desktop parts. I never said that the current MP makes a good xMac.

[CharlesS]

Originally Posted by P 

Actually, no. I was more surprised when they added Firewire to the iBook back in the day.

The iBook was an "iMac to go" - it was pretty much supposed to have the features of an iMac, in a laptop package.

The original iBook didn't have FireWire because the original iMac didn't.

Anyway, I'm not 100% confident that the next iMac is going to have FireWire in it, now that Apple seems to be in the process of phasing it out.

Ethernet's also in the southbridge. Firewire is not, but I'm not sure that that makes much of a difference - I'd want to replace the motherboard anyway.

If it was your personal machine, perhaps. If it was one of a number of lab computers or a computer belonging to a prof who only uses it for e-mail and MS Word, a replacement Ethernet card in a PCI slot would work just as well.

Less PITA than replacing a laser fuse because some idiot put the wrong kind of transparencies in it.

It's also less PITA than replacing the whole lab because a massive earthquake struck and pulled the entire building into the bowels of the earth, never to be seen again. Just because you can imagine a bigger PITA doesn't make this not a PITA.

The current iMac is more of a PITA than previous models, though, or so I understand, but for an HD crash you'd still have to reclone the entire shebang. It's still not a quick operation.

So what? You can do other things while you wait for the clone to finish.

Also a thing I haven't seen - well, other than the occasional bent pin that you can straighten out.

Pretty sure I've seen that once or twice, although since I can't recall a specific example I don't want to say for sure. I'm fairly certain it happens, though.

Originally Posted by P 

And yet people buy them.

In smaller proportions than the percentage of desktops on the PC side. And apparently in small enough numbers that the rumor mill is pretty much always predicting the mini being discontinued.

[Simon]

Originally Posted by P 

Slots actually make less sens for a desktop where you have a desk to put extra boxes on.

GPUs don't come in boxes. HDDs that come in boxes and can be attached to a Mac have bridges. Boxes have PSs. All of that sucks.

And Apple could put an EC slot on the iMac

I would definitely like to see that happen. I guess Steve's mood and the phase of the moon at the iMac's inception were against that though.

And yet people buy them.

Just as others don't. In part for the reasons I have pointed out here. Nevertheless, according to some people here iMac customer numbers will be approaching zero soon anyway because the notebook will take over the world. Of course eventually (as happy current dekstop owners scramble to find a notebook that meets their needs) Apple will fight the same battle there. The cheapest PC notebook comes with EC/34. Something Apple refuses to put on their MB. Also Apple ties screen size to performance. Another freedom PC manufacturers give to their customers and Apple doesn't. You can rationalize all day long, but Steve's ego will always be a problem.

A large case and noise are not compromises?

It's called options. iMac for those that want a small cases. xMac for those that want expansion. Instead of making everybody happy, Apple prefers to promote Jobs' AIO obsession. In the process they do one thing well but tell all the other people to not give OS X a try.

No. Don't bring it back to the MP. The question I answered was if Apple would save money by moving from laptop parts to desktop parts. They would certainly save money by moving from low-volume server parts to desktop parts. I never said that the current MP makes a good xMac.

So there we go. The current MP makes a bad xMac, the mini and iMac with their lack of expansion and non-exchangeable GPUs too. So we are back where we started: Apple has a gaping hole in their lineup. And contrary to all the pseudo-technical rationalizing I've read here, in the end it's all for one simple reason: Steve likes AIOs. Steve's a smart guy and has done very well for Apple. But on this one he's just wrong. Plain and simple.

[P]

Originally Posted by Simon 

GPUs don't come in boxes. HDDs that come in boxes and can be attached to a Mac have bridges. Boxes have PSs. All of that sucks.

GPUs don't (generally) come in boxes, but they do come in replaceable packages that fit inside laptops. The iMac uses one such package - a variant of MXM. The lack of upgrades is simply because Apple doesn't want to make any, and noone else has bothered figuring out how to - possibly because it would be hard to get to them. GPUs could be made replaceable without moving to a minitower.

Bridgechips don't matter if the drive is only an archive, and the external PS only a box.

Originally Posted by Simon 

<EC slot on a Mac>
I would definitely like to see that happen. I guess Steve's mood and the phase of the moon at the iMac's inception were against that though.

He doesn't like to add expansion options. He likes to sell new boxes. I'm just making the point that such expansion is not predicated on the box being a tower.

Originally Posted by Simon 

It's called options. iMac for those that want a small cases. xMac for those that want expansion. Instead of making everybody happy, Apple prefers to promote Jobs' AIO obsession. In the process they do one thing well but tell all the other people to not give OS X a try.

They will certainly not make everyone happy, because then people will want XXtreme gamer laptop or something. It all comes down to the number of people who'd like to try a Mac, but won't do it unless there is exactly the model they like. Based on past performance and current trends, I think that that number is too small to design a new case for.

Originally Posted by Simon 

So there we go. The current MP makes a bad xMac, the mini and iMac with their lack of expansion and non-exchangeable GPUs too. So we are back where we started: Apple has a gaping hole in their lineup. And contrary to all the pseudo-technical rationalizing I've read here, in the end it's all for one simple reason: Steve likes AIOs. Steve's a smart guy and has done very well for Apple. But on this one he's just wrong. Plain and simple.

As long he's growing the market and making the money, he's doing something right. Sure he likes AIOs, but then it is a logical endpoint for the integration of computers. AIOs doesn't preclude limited expansion, however - in fact, my old 1st gen iMac G5 is easier to work with than most towers, it has a CPU with a TPD similar to a Wolfdale and it has a PCcard slot. Put the GPU on an MXM card and make the CPU socketed you have a pretty moddable machine. Apple could make that without violating the AIO ethos, but they don't.

[Simon]

I agree they could make the iMac much more expandable, but they won't. For the very same reason they're not doing an xMac.

And even if they'd add socketed CPUs, slotted GPUs (just because they use MXM as a physical connector does not mean I can exchange with another MXM card BTW), an extra disk bay, eSATA, etc. we'd still have a Mac that puts thin design before performance. On the desktop I find that half-decent at best. IMHO it's a shame that only one Mac seems not to be a thermal/space compromise.

[SierraDragon]

Note that the OP defined business workstation as

Originally Posted by olePigeon 

By business, I mean a general workstation that can be configured as a relatively cheap workstation, or upgraded to a decent CAD or video workstation.

IMO "decent CAD or video workstation" describes the existing MP tower. We have that, and it is a superb design. The argument then gets to what about "...a general workstation that can be configured as a relatively cheap workstation..."?

Personally I agree with Apple's existing approach of skim-the-cream design/marketing; first develop the upper (profitable) end, and then let prices fall as the market evolves. Or stay only at the high/mid end if tech evolution and competition continually make the low end a waste of resources. That seems to be more and more the case, especially as laptops gain market share.

Originally Posted by Simon 

... If they're going to do hobbies, why not with Macs? ...Why is it that they're curious and daring about anything else (take the iPhone gamble for example) but get all conservative and cautious when it comes to the Mac?

Like the iPhone, the Apple TV is an experiment in an emerging market space with potentially HUGE upside potential and little downside potential. A cheapened Mac tower ("xMac") would address a price-tight stagnant (as opposed to emerging) market segment meaning low upside potential yet with the large downside risk of cheapening the image of the Mac brand as well as cannibalizing sales of other more profitable Macs.

Originally Posted by P 

My theory goes like this: chips get smaller, and they cannot get any hotter - therefore they will have to use less power. For each shrink, it makes even less sense to put them in a big ATX box... The price difference between a desktop and a laptop will shrink to the cost of the display, and at that point, there is money in standardizing on a laptop form factor - especially if SSDs conquer the world, as I suspect.

Certainly enough chip efficiency improvement might at some point make a smaller tower size logical; that is simply an opportunity for case redesign when the time comes, not an argument that Apple should build an xMac for the low end. However note that after 20 years it has not happened. We have continuously added expanding graphics needs to general computer usage such that today's Mac Pro is a beast for good reason.

Today there are more and more still and video digicams proliferating exponentially increasing GBs of image data that need to be handled. We see in these forums novice users say things like "...my needs are light, just surfing, office, iPhoto and YouTube." Even olePigeon included "video workstation" when he defined workstation needs.

Once graphics are in the mix "light" needs are no longer light, and the existing MP tower design is excellent. Hopefully Apple will just keep down the price points, no xMac needed.

-Allen Wicks

[Simon]

Originally Posted by SierraDragon 

A cheapened Mac tower ("xMac") would address a price-tight stagnant (as opposed to emerging) market segment meaning low upside potential yet with the large downside risk of cheapening the image of the Mac brand as well as cannibalizing sales of other more profitable Macs.

I disagree with this for several reasons. But most of all because that market isn't stagnant. It's non-existant in Apple world. It might be stagnant for Dell, but we have no idea how it would do for Apple until they enter it in the first place. Right now there is nothing.

[P]

Originally Posted by Simon 

I agree they could make the iMac much more expandable, but they won't. For the very same reason they're not doing an xMac.

Emphasis mine. I don't think it is for the very same reason. They won't make an xMac because it's ugly and they (and I) think that the market for it is small. Their reasoning about expandability are different - I don't really think that they're negative to it (anymore), but that it is a secondary concern. White non-iSight iMacs were eminently accessible. The latest laptops have improved accessibility, by making the battery and HD easily accessible. iMac CPUs were socketed on the first (white) Intel iMacs, but replacing the CPU is always tricky because you need to replace the heat paste and push the cooler down just right. There isn't really a market in MXM upgrades even on Windows, so there is little sense in making such upgrade terribly easy.

[Simon]

Originally Posted by P 

They won't make an xMac because it's ugly and they (and I) think that the market for it is small.

I'm sure you think that. But in this thread we've established they do not. We actually established the likely reason they're not doing the xMac is because Steve simply doesn't like it. We know it could be done technically. We know it would fit in between the MP and Mm and still drive margins. We know there's a potential market for it. We know it could be beautiful - think about a PM6100 made from Al. We know all the rational reasoning. What remains is personal preference.

They're not doing it for no serious reason other than Steve hates headless desktops.

There isn't really a market in MXM upgrades even on Windows, so there is little sense in making such upgrade terribly easy.

Was that a joke? Of course there's no such market because in the Windows world the GPU update market is based around PCIe cards of which there are literally hundreds to chose from. If the iMac had a standard PCIe slot for a GPU this would be moot. But Steve doesn't want that because it would make the iMac a 1/2" thicker.

No desktop user in the Windows world needs to worry about laptop constraints such as MXM. Because only Apple Steve thinks it's wise to push laptop constrained enclosures on the desktop. Shuttle users get a PCIe card. Pizza box users get a PCIe card. Mini tower users get PCIe cards. A huge selection of cards for various purposes and at various performance/price/heat levels. A huge market we are locked out of because Apple thinks a single standard PCIe expansion card is a "Pro" feature. Every owner of a $500 plastic Dell will roll over laughing when he hears that.

[SierraDragon]

Originally Posted by Simon 

We actually established the likely reason they're not doing the xMac is because Steve simply doesn't like it...

Who established? I could not disagree more. Yes SJ runs Apple, but implying that sound business analysis is just his whim is flat wrong. I have closely observed Apple since the II, and IMO an xMac would (a) cheapen the brand and (b) cannibalize sales of more profitable Macs just to play into a very low-margin market segment (very low-margin potential market segment if you prefer).

The cash and engineering resources that would go to investing in a low margin xMac line are better spent looking for future iPhone-type market successes. Some folks have been whining for an xMac for a long time; one could even argue that the cash and engineering resources that could have gone into investing in a low margin xMac were better spent creating the iPhone.

Of course I could be wrong (and Apple/SJ could be wrong) and investing in creating an xMac could in reality be a good use of cash and engineering resources for Apple. But the argument is based on market analysis, not "because Steve simply doesn't like it."

Note: I have never had any relationship of any kind with Apple or with any Apple employee.

-Allen Wicks

[Northeastern292]

Originally Posted by Simon 

Come to think of it, this is an excellent point. I can see the Mac described here as a very viable solution at work, but I realize that it's probably also my skewed picture. Academia, science, and engineering have significantly different needs from the vast "corporate" computing world.

So although this box would suit many of my colleagues perfectly, if it were to actually go on sale for let's say round $1500 I'm pretty sure Apple would sell most of them as home computers.

Half of the public school in New York State use Dell Optiplexes, so Apple could learn something from Dell in that department.

[Simon]

Originally Posted by SierraDragon 

Who established? I could not disagree more. Yes SJ runs Apple, but implying that sound business analysis is just his whim is flat wrong.

Well you see, that's what I always thought. I knew Jobs was good at business and Apple has been doing very well. The last couple of years have been all about revenue. I was certain there must be a good business reason not to do the xMac and I was determined to find out what it was.

But when it comes to the topic of the headless midrange Mac it just didn't work that way. This is not the first thread on the topic. And every single time we went up and down all the arguments and no matter which way you turn it, bottom line is you can think of an xMac that seems very attractive and could still be priced in such a way that it would drive margins even better than the iMac (which by Apple's standards isn't that huge btw). So you ask people, what are we missing, which part of this doesn't work? The answer is silence.

Here's the thing, if you can see these concepts on a board like this one you can be assured Apple has contemplated it too. And yet they won't budge. And I still haven't heard any compelling reasoning for that. So far the conclusion has been (see this thread, see all the others) it's simply that Steve loves the AIO iMac and hates the headless desktop. I was hoping for something better, but so far this is about the best this (or any other) board has come up with.

[Big Mac]

Well, as has been discussed on other forums, Apple's going to have to really seriously think about the xMac in the near future. There's going to be a much larger disparity between Apple's midrange desktop options and those of the PC makers given that Core i7 won't be able to go into the iMac any time soon.

[P]

Originally Posted by Northeastern292 

Half of the public school in New York State use Dell Optiplexes, so Apple could learn something from Dell in that department.

Because Dell sells so many AIOs otherwise? The AIOs they sell include TV tuners, which is a very pointless extra cost in a school.

[P]

Originally Posted by Big Mac 

Well, as has been discussed on other forums, Apple's going to have to really seriously think about the xMac in the near future. There's going to be a much larger disparity between Apple's midrange desktop options and those of the PC makers given that Core i7 won't be able to go into the iMac any time soon.

Possibly, but then the Core i7 fills the top-of-the-line segment that only represents a small fraction of desktop sales. I'm sure OEMs will put it in their gaming boxes, but the "average business desktop" will not see a Nehalem-style CPU for a long time. Note that the regular desktop CPUs will launch at the same time as the laptop ones. The enthusiast versions launching now are expensive and also require expensive motherboards. The current launch is mainly for server-type workloads.

[P]

Originally Posted by Simon 

But when it comes to the topic of the headless midrange Mac it just didn't work that way. This is not the first thread on the topic. And every single time we went up and down all the arguments and no matter which way you turn it, bottom line is you can think of an xMac that seems very attractive and could still be priced in such a way that it would drive margins even better than the iMac (which by Apple's standards isn't that huge btw). So you ask people, what are we missing, which part of this doesn't work? The answer is silence.

Are we starting the debate all over again? The margin should not be compared to the iMac - it's fine if your xMac leeches the iMac. It should be compared to the MP, which there is a significant risk that it cannibalizes.

Originally Posted by Simon 

Here's the thing, if you can see these concepts on a board like this one you can be assured Apple has contemplated it too. And yet they won't budge.

Except they did, with a single CPU 1.8 GHz G5 Powermac. That didn't sell.

Originally Posted by Simon 

And I still haven't heard any compelling reasoning for that. So far the conclusion has been (see this thread, see all the others) it's simply that Steve loves the AIO iMac and hates the headless desktop. I was hoping for something better, but so far this is about the best this (or any other) board has come up with.

Things aren't black and white. Jobs by all accounts loved the Cube and was crushed when they had to cut it, but cut it Apple did. If you watch his presentations, you will notice that there are features in every OS release that he'd really rather skip - watch the introduction of Spaces in particular. There was even a two-button mouse eventually.

Jobs doesn't like minitowers and there is no compelling reason for him to permit them anyway. Add up the rather minor component savings compared to an iMac and the potential increased revenue and subtract development and support costs along with leeching sales from the higher-margin MP (I don't suppose you expect the xMac to have MP-level margins?) and the end result is basically a wash.

My bet is that Jobs was convinced to permit the single-CPU Mac Pro as an xMac experiment, the same way the Mighty Mouse was permitted, but that its sales remained unconvincing. There's your reasoning.

[Simon]

Sorry, as I already told you that reasoning simply doesn't cut the mustard.

• The 1.8 GHz single-CPU G5 wasn't an xMac. It was an expensive to produce castrated PM. The xMac is the opposite of that. It's a inexpensive to produce version of the midrange Mac.

• The xMac can drive the same margin as the MP.

• The xMac can drive more revenue than the iMac or MP actually. The MP's easy to beat due to its low sales numbers. My guess is the iMac could be beat because of the added flexibility and better price/performance ratio.

• The Cube was very Jobsian: it put form before function. With it's sexy looks and tiny size it was truly a Jobs product. The xMac isn't. And that's what I just said the last few posts.

I know I'm repeating myself since we have already talked about these things several times in this thread. I'd love to hear something better than "Jobs simply hates the xMac", but you're going to have to come up with something a lot better than what you just wrote if you want to finish it with "There's your reasoning." Just my 2¢.

[Northeastern292]

Originally Posted by P 

Because Dell sells so many AIOs otherwise? The AIOs they sell include TV tuners, which is a very pointless extra cost in a school.

Somewhat of a good point, but until my chemistry/physics teacher got a Smartboard, he relied on the Dell's graphics card to have TV-Out, onto the television in the class for PowerPoint slides.

Dell also seems to heavily lowball in that area. Or maybe it's that I live in a very tech-illiterate part of the state, where we are hours from an Apple Store (in addition, the North Country is more Myspace than Facebook, to my disdain).

[Northeastern292]

Originally Posted by P 

Slots actually make less sens for a desktop where you have a desk to put extra boxes on. And Apple could put an EC slot on the iMac - they used to have a PCCard one, after all.



A price to pay for the accessibility of not having to pull the machine apart to get it to work.



They're there for RAID, and because at the size the case you have to be to fit the Xeon cooling and slots, you might as well.



And yet people buy them.



A large case and noise are not compromises?


No. Don't bring it back to the MP. The question I answered was if Apple would save money by moving from laptop parts to desktop parts. They would certainly save money by moving from low-volume server parts to desktop parts. I never said that the current MP makes a good xMac.

And what is with Apple building desktops out of laptop components. As sweet as the current iMacs and Mac Minis are, they suffer from laptop parts.

People, you must understand that I am one of Steve Jobs's biggest critics when it comes to the Mac lineup. I love the current Mac lineup, but no Mac suits any particular need. There's no affordable gaming Mac (unless you want to shell out $2500 for a Mac Pro, there is no nettop Mac (the Linux arena has that covered well, and the Apple TV is perfect for a potential nettop Mac), there is no netbook Mac (and I bet a netbook Mac would sell like hotcakes!)

[P]

Originally Posted by Northeastern292 

And what is with Apple building desktops out of laptop components. As sweet as the current iMacs and Mac Minis are, they suffer from laptop parts.

Two reasons:

1) Because they clock down further when load is low, so the average fan speed can be kept lower.

2) To save on development costs and time. The first Intel iMacs were essentially MBPs in a desktop shell.

With Apple's transition to nVidia chipsets in the iMac, item 2) disappears. As a result, we might actually see a transition to Wolfdale-class chips in the current iMac case, although I doubt it.

Originally Posted by Northeastern292 

Somewhat of a good point, but until my chemistry/physics teacher got a Smartboard, he relied on the Dell's graphics card to have TV-Out, onto the television in the class for PowerPoint slides.

I don't mean TV-out (all iMacs since the G4s or thereabouts have had TV-out through adapters), but a full-featured TV capture card.

[P]

Originally Posted by Simon 

Sorry, as I already told you that reasoning simply doesn't cut the mustard.

• The 1.8 GHz single-CPU G5 wasn't an xMac. It was an expensive to produce castrated PM. The xMac is the opposite of that. It's a inexpensive to produce version of the midrange Mac.

Why? The motherboard was the cut-down version very similar to the iMac, and not the full-size PM mobo (it had half the DIMM slots, for one). It ought to have cost less than the iMac, as Apple did not include a display. Once again: Not a current MP, the equation would be different for that.

Originally Posted by Simon 

• The xMac can drive the same margin as the MP.'

• The xMac can drive more revenue than the iMac or MP actually. The MP's easy to beat due to its low sales numbers. My guess is the iMac could be beat because of the added flexibility and better price/performance ratio.

So, an xMac would have MP margins and be better price/performance than the iMac. This means that it would have to cost quite a bit less than the iMacs, since the display would not be included and the performance would be similar. The 20" iMac display is a midrange model, so let's call it $200 to get an equivalent. The 24" display is better, so call that $350 or maybe even $400 for a reasonable replacement (using new egg prices). Fair?

Note that Dell wants $290 for its 20" widescreen in my pricing exercise below.

That's $999 for the bare bones xMac, 2.4 GHz and a lower midrange GPU - yes, the 2400XT is less than that now, but that model is old. Since price/performance should be better, let's say $899.

I priced up something similar at Dell - a "Studio Slim Desktop". It's more or less what I think you want - it has a tray-loading optical on the side, and Jobs would have to die before Apple releases that, but it's more or less OK. I couldn't get a corresponding CPU, but had to settle for an E7300 - 2.53 GHz but a smaller cache - and I put a Radeon 3450 in it. Getting an exact match for the OS and software bundles Apple includes is hard, but I picked Vista Ultimate, Photoshop Elements, MS Works and Premiere Elements. Know what this ended up costing?

$899.

I really did this as I was writing it, so the $899 price was not picked in advance. I almost laughed out loud when I saw it. Note that there are no speakers or webcam included here, as I expect that to be in whatever display you pick - and note that if you pick the display from Dell, you miss that target.

So for this to work out, Apple would have the same margin on the MP as Dell has on its cut-price desktops. Really? I very much doubt that.

Originally Posted by Simon 

• The Cube was very Jobsian: it put form before function. With it's sexy looks and tiny size it was truly a Jobs product. The xMac isn't. And that's what I just said the last few posts.

I didn't include it as an xMac example, I included it as an example of Jobs going against his likes and dislikes when the numbers are convincing enough.

Originally Posted by Simon 

I know I'm repeating myself since we have already talked about these things several times in this thread. I'd love to hear something better than "Jobs simply hates the xMac", but you're going to have to come up with something a lot better than what you just wrote if you want to finish it with "There's your reasoning." Just my 2¢.

If your really think that an xMac would have MP margins while having better price/performance than the iMac, then yes, I can understand why you think the xMac is such a done deal. I think I've shown with this post if not before why I don't think that that is the case.

[Simon]

Let' assume Apple makes roughly a $1k for every MP sold. If the xMac makes $600 a pop and you sell two for every lost MP sale you have already made more money off of it. The better price/performance ratio BTW is part of what makes that possible. The fact that Apple can get components that perform better than the mobile components in the iMac for half the price allows them to build a box that performs as good as the iMac for a whole lot less -> better margins. Best of all, since it's modular it will allow users to put emphasis where they deem it important. Not everybody who wants a fast CPU and big disk needs a beefy GPU. Not everybody who wants a 24" screen needs an extra fast processor.

By the way, did you notice that if you slap $500 extra profit onto those $899 you came up with we have a $1399 xMac? Do you have any idea how many people would love a 2.4 GHz headless xMac for $1399? And every one of those xMacs will have made Apple more profit than they make on the 2.4 GHz iMac today - even if they lose the screen sale. You see, Apple's not competing with the Dell on price (not even close), but they would be offering a mid-range box at a decent price that gives people a lot more flexibility than anything else Apple's line-up offers below $2200.

The bottom line is that $2200 for a midrange home or office computer is outrageous by today's standards. Apple's answer to that has so far been a limited selection of mobile computers (some in non-portable cases). And a lot of people (especially those that don't worship Apple out of principle) simply would never consider such a Mac. The xMac is Apple's chance at entering the area around $1500 with a product a lot of people can relate to. Like it or not, the product has to convince the competitor's customers if you want to expand. And Apple could be expanding faster - just like they innovated themselves out of the last recession. New and smart products are the key.

[Simon]

Originally Posted by P 

1) Because they clock down further when load is low, so the average fan speed can be kept lower.

I think this argument reverses cause and effect. If you make a case ultra slim or small you need tiny fans. Tiny fans need large RPMs and tend to be loud. I you losen the case size restraints you can use larger fans. They're very efficient and very quiet. The 120 mm fan I have in my new RedHat workstation at work is silent even under heavy load. And it removes ~100 W. Had Apple not insisted on ultra-small enclosures, we wouldn't need all this downclocking (which incidentally also makes parts expensive).

2) To save on development costs and time. The first Intel iMacs were essentially MBPs in a desktop shell.

The iMac has always been about size reduction,. Apple didn't chose a mobile chipset and CPU just because they were already using it. This choice was dictated by the design. Remember the first iMac G5? Apple was determined to not increase the size of the Intel iMac (and today it's smaller yet!) but they didn't want to hear about heat or noise. Therefore mobile components were a must.

If Apple had just a single midrange product where an anorexic case design was not a prerequisite people would be seeing better performance and a lower manufacturing cost. But since anorexic cases are the number one design constraint, the amount of choice we see in Apple's lineup is severely limited as soon as you don't want to settle for a portable. In a sense their success in the portable area (where these super-slim cases are important) is preventing them from doing really well in the non-portable area.