[P]

Originally Posted by Big Mac 

Uh huh, sure. Components I bought for a total of $1,400 months ago that would give the new Mac Pros runs for their money - I'm sure Apple could easily get 30+% margin out of such a box easily, price it at $1899 and sell a ton to customers who want a midrange tower, not a high-end iMac.

The only reason why I can see it isn't done is because Apple has to protect the iMac line. Forcing people to buy iMacs when iMacs are inappropriate for their needs is part of the sales strategy. And it's the major reason why my brother and I ended up building our own system.

What it would it cost you to but the exact same major components as the low-end MP? Same question for the top iMac. Now compare their prices.

Or put another way: Look at what it would cost to get a computer identical to the top iMacs at another major OEM - big gray box if you like, but the same CPU, RAM, HD, display and equivalent desktop GPU. Now look at what it would cost to get a computer equivalent to the low-end MP - or just use any of the multiple calculations already available in this thread. Compare the price delta.

The iMac doesn't need protecting, it's quite competitive as it is even if you don't like it. There's this other computer that noone would buy, however.

[besson3c]

This is changing the topic a little, but I'm looking forward to the day when these discussions are completely irrelevant.

It seems like the name of the game now is no longer so much about going after faster and faster clock speeds, but about farming jobs out to multiple CPU cores. In high performance computing environments single crazy fast computers are being replaced with clusters, operating systems real or virtualized. The cost benefits are immense this way, and the increase in overall efficiency quite significant.

To me, this is eventually where all high performance computing is going, including eventually for home use for non portable devices. I envision big cases with room to add as many CPUs and as much RAM as you want, and individual workstations going back to being dumb terminals. I know it will take a good while for this to happen and to wean CPU makers off of their current model of selling new single computers every 3+ years, but the first company that can sell these computing appliances and make all of this work will make a killing. There is no reason why you couldn't own an appliance like this for, say, 6 years or even longer, and just upgrade components as needed while remaining cutting edge.

[xmattingly]

Originally Posted by P 

Person Man commented on this (and Spheric, and I once before) but to be certain this is crystal clear: Complete desktop market. Not Apple market, or Mac market.

Yeah, I get it. Believe it or not, I am relatively aware of what has been said in this thread previously. The biggest sticking point is what Spheric continues to reference, which is the iMac being their shining light for desktops: problem being that those are your only choice for a top end consumer machine.

Some of us do not want a built-in monitor, and there are certainly prosumers who would like easy expandability and don't need 8+ terabytes of built in storage. Mac Mini's are Apple's BYOKMM; somewhere either above the iMac as a mini tower or below it as a headless iMac, I think there is certainly room for a different type of Mac.

Originally Posted by P 

Linky for that? I watched those, and the bit usually commented on from there was his comparison of tablets as cars and computers as trucks

I'd have to sit through that video again. I remember the analogy to trucks; somewhere along the line the "hub" bit was mentioned as well.

Originally Posted by P 

All I'm doing is comparing the cost of the components in the current low-end MP with a midrange tower and finding that they're basically the same.

I still think Apple could pull it off, if they really wanted to go after it. How many pundits and experts pieced together component costs before the iPad before it was introduced, and came to the conclusion that it'd be a $1000 product?

Originally Posted by P 

Also, there is a difference between a point of view and a fact. When I'm saying that the low-end MP is overpriced, that is a point of view. When I price together a cheaper one at HP or Dell (both of which I've done in this thread), that's a fact that I bring up in support of my point of view.

Well, pricing vs. what is available in the Mac product line is an apples & oranges comparison. Yes Mac towers are competitive to top end PC's. And - as this has been mentioned before - the price between the top end iMac & low end pro is $500. All I'm saying is there is room in their product line to offer more choice.

Originally Posted by P 

It was an attempt to make something between an iMac and an MP. It let you easily upgrade GPU and HD, two of the most frequently requested features of a midrange iMac on these forums. It had a lot of other things going in its looks etc, but it showed that Apple was thinking about this segment between MP and iMac.

The big issue - and probably the biggest reason for its failure was not for lack of demand, but that Apple priced it out of the market. That one cost MORE than a low end G4 tower, not less, so between that, having no PCI expansion or room for an extra hard drive I would not classify the Cube experiment as a mini tower.

[Spheric Harlot]

I'll reference it again: the iMac may be the only mid-range option for Mac users, but it's the entire desktop in the entire INDUSTRY (which includes the entire corporate market) that's showing any growth at all.

People. Just. Don't. Want. Towers.

The high end market is an exception, and that's where the Mac Pro goes.

We can keep repeating ourselves at each other, but the fact is that yes, Apple could produce a midrange tower, but the only empty space available for it is a market that's increasingly devoid of buyers.

It would lose Apple a lot more money than the Cube.

[Spheric Harlot]

Also, you're forgetting that the pundits that predicted a $1000 iPad were predicting spectacular failure and expecting laptop-class hardware running a Mac OS X touch.

[sek929]

The overall market may be declining, but if the iMac is selling like hotcakes then that's the new desktop market right there for Apple.

Are you telling me that noone that buys iMacs would get xMacs if they existed? Apple does sell tons of desktops, the iMac, and I bet that a good number of those iMac buyers would have gladly paid a little extra for an xMac.

[The Final Dakar]

I do think if an agreeable xmac was released the new debate of choice would be over the sizing and pricing of Apple's LCD line.

[besson3c]

Originally Posted by Spheric Harlot 

I'll reference it again: the iMac may be the only mid-range option for Mac users, but it's the entire desktop in the entire INDUSTRY (which includes the entire corporate market) that's showing any growth at all.

People. Just. Don't. Want. Towers.

The high end market is an exception, and that's where the Mac Pro goes.

We can keep repeating ourselves at each other, but the fact is that yes, Apple could produce a midrange tower, but the only empty space available for it is a market that's increasingly devoid of buyers.

It would lose Apple a lot more money than the Cube.

I don't follow how it would lose Apple much money? Making a case for a tower cannot cost them that much in R&D.

[Big Mac]

Originally Posted by P 

What it would it cost you to but the exact same major components as the low-end MP? Same question for the top iMac. Now compare their prices.

But that's just it - I don't care what the low-end MP components cost, aside from perhaps the case. Apple's cases are the only thing I can't buy in the PC component market, and I can't believe they account for any truly substantial percentage of the Apple hardware cost overall. Take the class of components I bought for $1,400 (or $1,250 without the case) stick it in an Apple case and sell it. I don't care how much a low-end MP costs - give me the xMac I'm asking for. I don't see why you (Apple) can't.

The iMac doesn't need protecting

Then why not let it compete against a real mid-range desktop? Apple used to be able to do it. The fact that Apple refuses to do it now tells me the iMac is a sacred cow protected from normal internal desktop competition.

Originally Posted by Spheric Harlot 

I'll reference it again: the iMac may be the only mid-range option for Mac users, but it's the entire desktop in the entire INDUSTRY (which includes the entire corporate market) that's showing any growth at all.

You have yet to demonstrate why a computer built on the components I bought couldn't be sold by Apple with a very healthy profit margin and still come in at a price attractive enough to both steal some portion of iMac sales but also, more importantly, address the mid-range desktop (non-AIO) market that Apple has refused to provide for since the Intel switch.

People. Just. Don't. Want. Towers.

Oh sure they don't at all. . Desktops still sell very well. System building components still sell very well. How do I know? They're priced competitively to sell. If there weren't a market for desktops then the individual component market would be a lot more expensive and less competitive - that's what one would assume if the only customers of those components were OEMs. What you claim is just simply untrue, SH. Laptops may account for larger percentage of sales, but desktops are still very popular, I assure you with every confidence.

The high end market is an exception, and that's where the Mac Pro goes.

The last time I looked MPs weren't exactly flying off the shelves. You've got it back assward, bro. The MPs are a small niche in the desktop market. And iMacs wouldn't be nearly as popular if a mid-range tower line still existed.

We can keep repeating ourselves at each other, but the fact is that yes, Apple could produce a midrange tower, but the only empty space available for it is a market that's increasingly devoid of buyers.

You keep claiming that without offering any proof to back it up, so count me incredulous.

It would lose Apple a lot more money than the Cube.

Surely you jest. A mid-range tower would cost Apple very little R&D and net huge returns from those who need mid-range towers. The Cube was a much more substantial engineering effort that failed because Apple priced it to fail. And it should be pointed out that the Cube wasn't a total failure - I count the mini as its successor.

[Waragainstsleep]

It depends how you look at it, but its more a case of making less money, rather than actually losing it. The Cube probably didn't lose Apple much money. It just didn't make them much because they weren't selling and had a few issues. Ultimately it was a spiritual predecessor to the Mac Mini in a way.

Essentially this mythical tower that people want is an iMac with its display swapped for a PCI-E slot or two. I think games are the only reason Apple would ever consider such a tower. Everyone else is catered for already, but asking someone to drop £2k just to play Call of Duty 6 is preposterous. To add a PCI-E slot for a big-ass graphic card to an iMac, you would make it uglier, and you murder its power consumption. Remember the Mini and iMac are essentially laptops in many ways.

USB graphics cards have gotten a lot better in recent years, maybe lightpeak will allow for worthwhile external video from a future Mac Mini. Thats the only way I see it happening. If you distill this discussion right down, it basically becomes "I really wish Apple would make a high spec desktop with no display built in, and no effort to reduce power consumption for a really low price".

Put like that, it may as well be "I wish Apple would give me a top of the line Mac Pro for free". We all wish we could get more for less, but if it bothers you that much that you can't, you can build a hackintosh.

[P]

Originally Posted by Big Mac 

But that's just it - I don't care what the low-end MP components cost, aside from perhaps the case. Apple's cases are the only thing I can't buy in the PC component market, and I can't believe they account for any truly substantial percentage of the Apple hardware cost overall. Take the class of components I bought for $1,400 (or $1,250 without the case) stick it in an Apple case and sell it. I don't care how much a low-end MP costs - give me the xMac I'm asking for. I don't see why you (Apple) can't.

The point is cannibalization. You are asking them to sell basically the same box at two different prices. Why on Earth should they do that? Apple, in their infinite wisdom, has decided that they want to charge us through the nose for the low-end MP. I disagree with that choice, but that doesn't change anything. Why should they add a second computer that is more of the same, except with a lower price? In what way could they possible cripple your favourite xMac so that none - or a tiny segment - of current MP buyers opt for it?

Originally Posted by Big Mac 

Then why not let it compete against a real mid-range desktop? Apple used to be able to do it. The fact that Apple refuses to do it now tells me the iMac is a sacred cow protected from normal internal desktop competition.

Apple very rearely competes with itself. The Cube overlapped a bit, the single CPU 1.6 GHz G5 overlapped, and the last PM G4 (after the G5 release) overlapped, but they were special cases and all quickly killed. iMac G3 was never priced anywhere near the Powermac, and earlier than that Apple was losing money hand over fist. There was some pricing overlap in the iMac G4 days, with the eMac below and the low-end Powermac at the top, but that was the exception brought on by Apple embracing LCDs a little too soon (and Apple didn't very well during that time).

Originally Posted by Big Mac 

You have yet to demonstrate why a computer built on the components I bought couldn't be sold by Apple with a very healthy profit margin and still come in at a price attractive enough to both steal some portion of iMac sales but also, more importantly, address the mid-range desktop (non-AIO) market that Apple has refused to provide for since the Intel switch.

That's good, because I wasn't trying to demonstrate that. I was demonstrating that an xMac like you describe it would cannibalize the low-end MP, thus being bad for Apple's bottom line. Any effect it has on iMac sales would be a wash - it might hurt Apple's volume discounts on display panels or something, but that's it.

[Big Mac]

Then you've just defeated your argument. We all know MPs don't sell well, and I'd be willing to assume the low-end model sells just about worst of all of them. So if Apple needs to kill the low-end MP to bring out an xMac that people actually want to buy, wonderful! You say it won't hurt the iMacs much, so based on that logic I fail to see why Apple refuses to drop the xMac on the market.

[P]

Originally Posted by xmattingly 

Yeah, I get it. Believe it or not, I am relatively aware of what has been said in this thread previously. The biggest sticking point is what Spheric continues to reference, which is the iMac being their shining light for desktops: problem being that those are your only choice for a top end consumer machine.

That's not a problem for them, though. You're talking about the only company that manages to make good money and grow the market on desktops, and you're telling them to stop doing what they're doing and start doing what everyone else is doing.

Originally Posted by xmattingly 

Some of us do not want a built-in monitor, and there are certainly prosumers who would like easy expandability and don't need 8+ terabytes of built in storage. Mac Mini's are Apple's BYOKMM; somewhere either above the iMac as a mini tower or below it as a headless iMac, I think there is certainly room for a different type of Mac.

What you want isn't terribly relevant. Sorry. Apple is not in the business of pleasing you - they are in the business of making money, and they don't think that there's any money to be made in making a box for you because there aren't enough of you. There are good reasons to believe that they may be right.

I'll ask you the same question: Tell me how you'd design your favourite xMac so it doesn't cannibalize the MP and still does what you want and no iMac does.

Originally Posted by xmattingly 

I'd have to sit through that video again. I remember the analogy to trucks; somewhere along the line the "hub" bit was mentioned as well.

The thing is that the hub was what Apple USED to do, and the truck analogy is almost a reversal of that. Then they had a severe case of TV envy (ie, everyone needs a computer like everyone has a TV) - now they're making devices that do certain pieces of what a computer does.

Originally Posted by xmattingly 

I still think Apple could pull it off, if they really wanted to go after it. How many pundits and experts pieced together component costs before the iPad before it was introduced, and came to the conclusion that it'd be a $1000 product?

Only silly experts predicted $1000 iPads, but anyway, there's a big difference: Jobs said straight up that they were pricing it low to win market share. He's also said before that the computer market share wars are over, and Microsoft won. He probably would like you reminding him of that now, but it's evident in Apple's strategy: The Mac is where Apple makes money - they're not going for share, because it's a mature market. They used to go for share in iPods - now they're money makers, except for the Touch which might still build a market.

Originally Posted by xmattingly 

Well, pricing vs. what is available in the Mac product line is an apples & oranges comparison. Yes Mac towers are competitive to top end PC's. And - as this has been mentioned before - the price between the top end iMac & low end pro is $500. All I'm saying is there is room in their product line to offer more choice.

Depends on what you want to offer, but we'll see what you put together.

[besson3c]

If Apple were to reduce the price of the Mac Pro to something approaching reasonable, would that reduce the need for an xMac? If the Mac Pros aren't selling much, why doesn't Apple start by doing this? The profit margins on those computers are absolutely ridiculous, but surely not something Apple depends on.

[P]

Originally Posted by Big Mac 

Then you've just defeated your argument. We all know MPs don't sell well, and I'd be willing to assume the low-end model sells just about worst of all of them. So if Apple needs to kill the low-end MP to bring out an xMac that people actually want to buy, wonderful! You say it won't hurt the iMacs much, so based on that logic I fail to see why Apple refuses to drop the xMac on the market.

Because they'd make less money on it. It's be easier to just drop the price of the bottom MP. Also, see Waragainstsleep.

[P]

Originally Posted by besson3c 

If Apple were to reduce the price of the Mac Pro to something approaching reasonable, would that reduce the need for an xMac? If the Mac Pros aren't selling much, why doesn't Apple start by doing this? The profit margins on those computers are absolutely ridiculous, but surely not something Apple depends on.

Didn't see that post when I made my last.

I've been hoping for some time that Apple would cut the price of the low-end MP to something like $2000, or maybe even less with a cheaper chip, but that didn't happen. As the iMac moves to Sandy Bridge (Core i7/i5/i3 2x00 series, now), the insanity of that price will only grow. If Apple uses the same CPU bins that they do today, the top i7 iMac will use a 3.4 GHz version of a more modern CPU, whereas you need to max out the single socket MP to get to 3.33 GHz on the older, weaker Westmere architechture - and that isn't even counting Turbo. The Xeon 3x00 and 5x00 series aren't due for a refresh yet, so Westmere is what the MP will be stuck with until mid-to-late next year.

[Big Mac]

Originally Posted by P 

Because they'd make less money on it. It's be easier to just drop the price of the bottom MP. Also, see Waragainstsleep.

If you want to claim they'd make less money than on the MPs (which I don't really understand because you've also claimed the MPs are pretty price competitive for the components used, except perhaps for the entry level), fine. They would more than make up for any reduced margin through volume of sales. I don't see how you can deny that it makes so much sense.

If as you advocate, the actual solution is to drop the low-end MP's price, then do that. I'm not wedded to the notion of the xMac. Make the low-end MP one. But as it stands Apple's losing a lot of loyal customers who have no Mac desktop option suited to them. Maybe the new Apple Inc. doesn't give a flying fark.

[Spheric Harlot]

For one, i don't think they'd make it up in more sales, and even if they do, how long would they?

Because if the market continues to develop as it does, they'd probably need to kill the product line after three or four years.

And going through the whole development, design, materials, prototyping, and advertising scheme ("The Mac everybody else is making") is a considerable investment (regardless of your claims, it is a completely new machine, or would be if Apple built it), and not something they'll do for a brief run of selling a couple to the in-betweeners.

I really, REALLY doubt that there's "a lot" that wouldn't build a hackintosh anyway.

[Apple Pro Underwear]

Speaking as a graphic artist who recently seriously looked into buying a new Mac... I nearly decided on a Mac Mini.

Why? Based on my satisfaction with my 3.5 year old laptop's performance... I knew the even faster modern day Mac Mini could probably do the job.

Also, I want to add that the print industry is dying. I don't do as much print work as I do web work and I can suffer Photoshopping a 100MB file twice a year if need be. What's it gonna take me time-wise? An extra few minutes? I remember in 2001 when as a intern, the pros at the ad agency I worked at used to go to lunch and come back waiting for files to open/save. That doesn't happen anymore.

[Veltliner]

Originally Posted by Apple Pro Underwear 

Speaking as a graphic artist who recently seriously looked into buying a new Mac... I nearly decided on a Mac Mini.

Why? Based on my satisfaction with my 3.5 year old laptop's performance... I knew the even faster modern day Mac Mini could probably do the job.

Also, I want to add that the print industry is dying. I don't do as much print work as I do web work and I can suffer Photoshopping a 100MB file twice a year if need be. What's it gonna take me time-wise? An extra few minutes? I remember in 2001 when as a intern, the pros at the ad agency I worked at used to go to lunch and come back waiting for files to open/save. That doesn't happen anymore.

Photoshop file sizes of only 100 Mb?

That's rare.

An unprocessed TIFF conversion from my camera's RAW files is 120 Mb. And this is without any editing. I always go over 500 Mb, and up to 800 Mb.

The thought of having to work on a mini causes me severe discomfort.

In regard to open/save time: on my 24" white iMac I have open times up to 2 minutes, save times between 1 and 3 minutes. My record file save time was 12 minutes.

I also want a good graphics card. The mini's graphics card - does this still qualify as a graphics card? Or isn't it rather a graphics appendix?

This is why my next machine will be a workstation. One painful investment, and years of good working.

[P]

Originally Posted by Big Mac 

If you want to claim they'd make less money than on the MPs

Yes. Thank you.

Originally Posted by Big Mac 

(which I don't really understand because you've also claimed the MPs are pretty price competitive for the components used, except perhaps for the entry level), fine.

2 socket MP uses expensive Xeon 5500/5600 CPUs, and are quite competitive. 1 socket MP uses cheap Xeon 3500/3600 processors, and is nowhere near competitive.

Originally Posted by Big Mac 

They would more than make up for any reduced margin through volume of sales. I don't see how you can deny that it makes so much sense.

Because I don't believe that the sales volume would increase enough to make up for it. Simple as that. That is where we always end up in these discussions.

Originally Posted by Big Mac 

If as you advocate, the actual solution is to drop the low-end MP's price, then do that. I'm not wedded to the notion of the xMac. Make the low-end MP one. But as it stands Apple's losing a lot of loyal customers who have no Mac desktop option suited to them. Maybe the new Apple Inc. doesn't give a flying fark.

They probably don't think that they lose very many customers. I happen to think they're right.

[Spheric Harlot]

Originally Posted by P 

YBecause I don't believe that the sales volume would increase enough to make up for it. Simple as that.

Especially since most of that volume would come at the expense of iMac sales (machines which probably have excellent margins due to sheer quantity sold) and low-end Mac Pros (which have excellent margins due to excellent - for Apple - pricing).

[besson3c]

Originally Posted by P 

2 socket MP uses expensive Xeon 5500/5600 CPUs, and are quite competitive. 1 socket MP uses cheap Xeon 3500/3600 processors, and is nowhere near competitive.

The 5620 Westmere quad core Xeons are $400 each, so two of them = $800. 1 gig of DDR3 1066 ECC sticks are $30, 6 of them are $180. The ATI Radeon 5870 is about $370 or so. 1 TB hard drive is cheap. That equals $1500 or so. Motherboard, case other stuff will cost more, but how on Earth is $3500 competitive in any way?

$3500 is highway robbery. ****ing insane.

[OreoCookie]

Originally Posted by Veltliner 

I also want a good graphics card. The mini's graphics card - does this still qualify as a graphics card? Or isn't it rather a graphics appendix?

Photoshop and Lightroom don't benefit from a faster graphics card, they don't use it for things other than zooming, panning and rotating the workspace. You won't benefit from spending extra money for an upgraded graphics card.

Other factors are much more substantial (slower 2.5" hard drive, less RAM, slower cpu) in determining how fast you can process photoshop files. But most of them are alleviated when you consider an iMac (with the obvious exception of storage).

[P]

Originally Posted by besson3c 

The 5620 Westmere quad core Xeons are $400 each, so two of them = $800. 1 gig of DDR3 1066 ECC sticks are $30, 6 of them are $180. The ATI Radeon 5870 is about $370 or so. 1 TB hard drive is cheap. That equals $1500 or so. Motherboard, case other stuff will cost more, but how on Earth is $3500 competitive in any way?

$3500 is highway robbery. ****ing insane.

It's competitive because the competitors charge roughly the same. The cost comes at least partially from a very complicated motherboard and the testing of it.

[Waragainstsleep]

Actually even Apple charges significantly less for a Mac Pro logic board than it did for the PowerMac G5s (which were priced lower than Mac Pros). Its still two or three times what you'd pay for a high end PC mobo but there you are.

I do feel Apple is getting to a point where they are walking a fine line. They have never, ever shown any inclination to become a "pile 'em high, sell 'em cheap" kind of operation like Dell, but one suspects they might eventually reach a point where the only way to sell more is to sell cheaper. BMW started to cross that line and eventually decided to limit their supplies to preserve their exclusivity. I wonder which way Apple will go.

[Spheric Harlot]

Originally Posted by Waragainstsleep 

Actually even Apple charges significantly less for a Mac Pro logic board than it did for the PowerMac G5s (which were priced lower than Mac Pros). Its still two or three times what you'd pay for a high end PC mobo but there you are.

I do feel Apple is getting to a point where they are walking a fine line. They have never, ever shown any inclination to become a "pile 'em high, sell 'em cheap" kind of operation like Dell, but one suspects they might eventually reach a point where the only way to sell more is to sell cheaper. BMW started to cross that line and eventually decided to limit their supplies to preserve their exclusivity. I wonder which way Apple will go.

Apple has made it absolutely clear which way they're going in the Mac market.

[Big Mac]

In the Mac desktop market, the company shown where it's' going. In the Mac laptop market at least Apple's trying to be more competitive. You can still find much cheaper hardware deals on the other side, but at least the difference isn't so gigantic that it slaps the customer in the face.

[besson3c]

Originally Posted by P 

It's competitive because the competitors charge roughly the same. The cost comes at least partially from a very complicated motherboard and the testing of it.

I'm not buying that... At least not yet. Can you show me an example of a competing machine that is similarly equipped and priced?

[Veltliner]

Originally Posted by besson3c 

The 5620 Westmere quad core Xeons are $400 each, so two of them = $800. 1 gig of DDR3 1066 ECC sticks are $30, 6 of them are $180. The ATI Radeon 5870 is about $370 or so. 1 TB hard drive is cheap. That equals $1500 or so. Motherboard, case other stuff will cost more, but how on Earth is $3500 competitive in any way?

$3500 is highway robbery. ****ing insane.

P noted that the two socket MPs use a different CPU, so, in this case, just doubling the prices of the one socket MPs to get to the cost of a two socket MP wouldn't work.

[besson3c]

Originally Posted by Veltliner 

P noted that the two socket MPs use a different CPU, so, in this case, just doubling the prices of the one socket MPs to get to the cost of a two socket MP wouldn't work.

The price I quoted was for two quad core Xeons... What do you mean by "doubling the price of the one socket Mac Pros"?

Regardless, unless this makes up a difference of around $2000 I think my point still remains...

[P]

Originally Posted by besson3c 

I'm not buying that... At least not yet. Can you show me an example of a competing machine that is similarly equipped and priced?

Someone already tried earlier in this thread. It is tricky because you can't pick the exact models of GPUs that Apple supplies - you have to pick a FirePro GL, so it becomes more expensive - but an HP Z800 with 2 E5620, a 5770 equivalent, 12 gig ECC RAM and 1 TB HD comes to $4263.

[OreoCookie]

Originally Posted by besson3c 

I'm not buying that... At least not yet. Can you show me an example of a competing machine that is similarly equipped and priced?

I've just configured a Dell T3500 64 bit workstation with 2 E5630 Xeons (they were the closest ones I could find), 12 GB RAM (comes with 2 (!) GB stock), 1 TB harddrive (which cost a whopping $345 extra ) and an nVidia Quadro FX-580 (the stock graphics card was a `professional 2D graphics card,' whatever that is ). Dell wants $3,354 for this machine. That's in the same ball park as the Mac Pro. They also feature the T5500 and T7500 lines (same config costs around $3,700) which are obviously more expensive.

If you compare Apple's Mac Pro to its competition, i. e. Intel-based workstations from other main manufacturers, you find that they are comparably priced if not a bit cheaper.

A workstation is more than just specs, lift one up and you'll be able to tell the difference (the Mac Pro weighs 18~19 kg according to Apple's spec site, . Even if you want to build one yourself, Intel's own workstation cases (without motherboard) cost around ~350 € alone.

[P]

Yes, that's what I mean by confusing GPU choices. That GPU (Quadro FX-580) is essentially a Geforce 9500. That's even weaker than what Apple tends to put in their MPs - in fact, it's even weaker than the 4670 in the bottom iMac.

[OreoCookie]

What's weird is that I was able to configure the T7500 with the E5620 while the cheaper T3500 was only configurable with the E5630. Actually, I did not want to spend a ton of time on this, it just comes to show that Dell charges approximately the same amount, if not more.

Plus, they really are charging extra: by default, these machines come with a measly 2 GB RAM, 250 GB hard drive and a very slow graphics card (the one I've included is already an upgrade!). Paying $345 for a 1 TB drive, for instance, is quite a lot, I can't imagine the whining if Apple dared to do the same.

[besson3c]

Okay then, shame on Dell and HP as well then for ripping off the consumer. There is no need to charge that much when it is just not that hard to build your own PC with the same specs for a fraction of that price.

[OreoCookie]

@besson
A workstation is not a pc and it's not targeted at consumers: it's not a question of performance or price, it's a question of reliability. That starts with the case and the cooling system, but goes much further and includes much more extensive reliability testing. The other part is actually service -- which is very important in certain business sectors.

It's not a rip off if that is exactly what you want. Buying cheaper pc parts is not a substitute: nowadays, manufacturers save fractions of pennies by going for cheaper and less reliable cap capacitors and such.

If you think those workstations are expensive, they're not. When I was a student, Dec and Compaq AlphaStations were in broad use. Once, one of the big AlphaStations had a problem. The physics department was very happy to learn it was still under warranty, because 2 cpus needed to be replaced (2 x $2,500) and the RAM modules were replaced as well ($2,000). The workstation cost roughly $20k new. On the plus side, the machine was up and running again within 6 or 8 hours (don't remember anymore).

In any case, you can make a valid argument that most people don't need all these bells and whistles -- and you'd be right, hence this thread.

[besson3c]

I don't really get the reliability testing argument.

For starters, a common point of failure with all computers is the HDD. These workstations (or whatever you want to call them) use SATA drives just like any of those you can get from NewEgg, nothing special there. As far as cooling, I just don't think it is that difficult to figure out how to cool your system at the noise level and efficiency you want. If you were to get a barebones computer this would be taken care of for you...

As far as the reliability factor goes, you'd be better off just buying spare parts which would still end up being cheaper.

I get the argument that buying a computer like this is a little easier and quicker for some companies - having one warranty that covers all the parts, a prefit/preassembled machine that you know will work and all that, but for us as home consumers or even some small businesses, if $2000 is not chump change for you I would strongly, strongly recommend building your own PC, cause it's just not that hard to do.

I still have a hard time stomaching the idea that a machine like this is good value, cause it's not. At all.

[P]

Value is not absolute. It's probably not good value to you. To someone for whom downtime is very expensive, it makes excellent sense.

[Veltliner]

Originally Posted by besson3c 

The price I quoted was for two quad core Xeons... What do you mean by "doubling the price of the one socket Mac Pros"?

Regardless, unless this makes up a difference of around $2000 I think my point still remains...

What I meant is, that the CPUs in the single Quad cores are not of the same quality level than the Quad Core CPUs in the dual Quad Cores.

At least that's what I got from P's post.

[Veltliner]

Originally Posted by OreoCookie 

I've just configured a Dell T3500 64 bit workstation with 2 E5630 Xeons (they were the closest ones I could find), 12 GB RAM (comes with 2 (!) GB stock), 1 TB harddrive (which cost a whopping $345 extra ) and an nVidia Quadro FX-580 (the stock graphics card was a `professional 2D graphics card,' whatever that is ). Dell wants $3,354 for this machine. That's in the same ball park as the Mac Pro. They also feature the T5500 and T7500 lines (same config costs around $3,700) which are obviously more expensive.

If you compare Apple's Mac Pro to its competition, i. e. Intel-based workstations from other main manufacturers, you find that they are comparably priced if not a bit cheaper.

A workstation is more than just specs, lift one up and you'll be able to tell the difference (the Mac Pro weighs 18~19 kg according to Apple's spec site, . Even if you want to build one yourself, Intel's own workstation cases (without motherboard) cost around ~350 € alone.

Looks like PCs are only cheaper at the lower end of the market - when mass production kicks in. The more you go towards high end, the same pricing as for an Apple applies.

[SierraDragon]

Originally Posted by Spheric Harlot 

People. Just. Don't. Want. Towers.

Except me...

[mduell]

Originally Posted by Veltliner 

What I meant is, that the CPUs in the single Quad cores are not of the same quality level than the Quad Core CPUs in the dual Quad Cores.

what? There's no quality difference. It's the same core with 1 QPI instead of 2. Heck the die probably has 2 with one disabled.

[Veltliner]

Originally Posted by mduell 

what? There's no quality difference. It's the same core with 1 QPI instead of 2. Heck the die probably has 2 with one disabled.

I was referring to this post:

Originally Posted by P 

2 socket MP uses expensive Xeon 5500/5600 CPUs, and are quite competitive. 1 socket MP uses cheap Xeon 3500/3600 processors, and is nowhere near competitive.

[OreoCookie]

Originally Posted by Veltliner 

Looks like PCs are only cheaper at the lower end of the market - when mass production kicks in. The more you go towards high end, the same pricing as for an Apple applies.

Well, yes and no: Apple is simply using higher-end parts for their Mac Pros which cost more than standard pcs -- the same parts that are used in workstations by other manufacturers. Thus, they are similarly priced.

It's not necessarily about mass production, it's also about the quality of the individual parts and the whole package.

@besson
I think you're still missing the point: reliability has less to do with reliability of individual components, but with how these components work together. It has to do with the amount of headroom you have in terms of cooling and expansion. How easily components can be exchanged and for how long components are available. With professional servers, the difference can be even more pronounced, they have many reliability features you cannot get with a home-grown machine. But I digress.

Again, nobody is arguing you can save money by building your own PC, for sure you can. But in many, many business settings, the cost of hardware is very small compared to the cost of software and the cost of manpower. And Apple is not competing with do-it-yourself people who are capable of building their own pcs, they're competing with the companies we've mentioned here, HP, Dell and such -- and they are competitive. For consumers, the calculation is very different -- but the Mac Pro is not a machine that's aimed at consumers.

[P]

Yeah, lol lol. The difference between Xeon 3500/3600 and Xeon 5500/5600 is indeed the extra QPI link. That is one example of "an inherent feature", the original definition of quality - if we're being picky.

[Veltliner]

Originally Posted by P 

Yeah, lol lol. The difference between Xeon 3500/3600 and Xeon 5500/5600 is indeed the extra QPI link. That is one example of "an inherent feature", the original definition of quality - if we're being picky.

So they ARE basically the same processor? As you don't need that additional QPI link in a single processor - there's no second processor to connect to.

[Veltliner]

Originally Posted by mduell 

the die probably has 2 with one disabled.

Sounds reasonable.

[P]

Originally Posted by Veltliner 

So they ARE basically the same processor? As you don't need that additional QPI link in a single processor - there's no second processor to connect to.

It's MOSTLY that. They are also configured slightly differently - the Xeon 5500 series has a faster Uncore and officially has a much higher RAM ceiling. It seems that only the 5500 series officially supports 3 DIMMs per channel, and according to ark the 3500 doesn't support 8 gig DIMMs.

The same chip with features disabled? It seems likely, they're not very different. OTOH the Nehalem generation had different codenames (Bloomfield and Gainestown/Nehalem-EP), which tends to indicate different masks. The Westmere generation has only one codename (Westmere-EP).