Was (is) the Mac Pro a misstep for Apple?

[Bwa]

My dual processor (single core) G5 was one of the worst Macs I ever owned. I had a dual core, single proc that was mildly better.

As far as pricing goes, you can construct any argument you want. The PowerMac 8100 was $4,200 in 1994. The 9500/132 was a 6-slot tower and started at $5,300 with no CD-ROM in 1995. That's $7,500 in 2010 dollars.

That said, I'd love to see a $1,500 tower come back. But I suspect the chances of that happening are about as good as having a new 6-slot tower come back :-(. 4 slots isn't enough.

I think what makes the most sense for a case redesign is something that will fit into a 19" rack.

 

[shaun319]

I think the issue that Apple is having with the MacPro is that it just does not sell enough of them to justify the energy, time, people, and money invested in the R&D to do it right. Those resources would be much better used on more consumer oriented devices IMHO.

 

[dansmac]

You miss the point completely. The PowerMac days you could get a top of the range tower for £1,500 now it's £2,000. Value for money is not in the equation.

We’re getting a bit off subject. That’s normal for forum discussions. Looks like two subjects going on now. Was the Mac Pro a fluke and is the current Mac Pro overpriced?

Was the Mac Pro a fluke? Not sure it was a total misstep but I still hold to my original conclusion that the Mac Pro is a “sidebar” business at Apple. I assume the “Pro” in Mac Pro means “Professional”. That said everything else Apple does is consumer based. The motives stated earlier regarding attraction of “anti establishment” culture of artists remains valid and we all hope that Apple continues to maintain a level of support and products for the professional community. Time will tell.

Is the current Mac Pro overpriced? Well that depends. For my case the answer is a decisive no. We own a professional photography studio. Our clients expect a level of quality and speed from our small studio in California. From 2005 to 2011 not much has changed from what our clients demand in product. High quality photography at a price that is competitive and customer focused service experience.

But the technology to get me to their expectations is a whole lot different. It comes down to value and return-on-investment…. remember “pro” = professional … so we treat the investments in computer platforms like any other business expense. What we wanted is a computer system that will be; dependable, fast, and expandable. The Mac Pro series from Apple hit the mark. So is the Mac Pro overpriced? No. The value of a trustworthy product with little to no maintenance with virus and other garbage has a price. The current Mac Pro gets you a lot for the money. And it’s not all just Intel although the move to the new Xeon architecture did a lot to enable other stuff. I get 10X performance over a 2005 variant at basically the same price point and don’t have to muck with these systems like we do with our Windows stuff.

For me it’s all about return-on-investment over the years of use. Options for our business are somewhat limited due to the applications we invest in. So systems like SUN or Linux based knock offs do not play. It’s down to the Wintel suspects (Dell, HP, others) and Apple. In the end the total cost of ownership (original purchase + software applications + maintenance + downtime + enhancements) over a five year life the Apple Mac is a winner hands down and a real bargain for OUR use….may not the same for your use…..

 

[derbothaus]

I think the issue that Apple is having with the MacPro is that it just does not sell enough of them to justify the energy, time, people, and money invested in the R&D to do it right. Those resources would be much better used on more consumer oriented devices IMHO.

They don't sell them to the home user generally.
What R&D? Keeping a tower in a product lineup is one of the easiest things to do. Everything exists already. Case, procs, mobo, memory. It all comes from other sources. They tweak very little. Just add the new board and procs and you are done. The real R&D goes into the smaller form factors as they have to deal with the thermals in a confining way. When you have a case that can distribute so much heat it is not all that hard or time consuming. And I've bought or sold hundreds maybe even thousands. More Mac Pro's/G5's/G4's/G3's than any other Apple product including the Books and iMac. Only since the i7 have I even entertained using these consumer products. And it is always the most top end version.

 

[goMac]

I think the issue that Apple is having with the MacPro is that it just does not sell enough of them to justify the energy, time, people, and money invested in the R&D to do it right.

I don't think they're doing it wrong.

 

[szolr]

If Apple continue upgrading the iMac way more frequently than the Mac Pro, they'll become less distinguishable in performance. At which point the Mac Pro will suddenly be all but obsolete. Yet there's no real reason why that should happen. The technology is available to make the Mac Pro a fair few steps ahead of the iMac at every corner.

Apple want to serve their customer interests. While I'm not a pro user I recognise many are. And pro users tend to buy pro machines, which tend to cost more. So Apple make lots of pro users even though there's less of them than standard consumers. Ahem... No the Mac Pro was not a mis step for Apple.

 

[-hh]

They don't sell them to the home user generally.
What R&D? Keeping a tower in a product lineup is one of the easiest things to do. Everything exists already. Case, procs, mobo, memory. It all comes from other sources. They tweak very little. Just add the new board and procs and you are done....

If only it were so easy. Each change in CPU architecture, a new graphics card and so forth will need to undergo QA testing for every OS update, for every Application and so forth...and I'm sure that since the Mac Pro's are the only hardware that Apple's selling that has dual physical CPUs that means that all of those specific QA costs don't get amortized across other hardware division.

The real R&D goes into the smaller form factors as they have to deal with the thermals in a confining way. When you have a case that can distribute so much heat it is not all that hard or time consuming...

A big case makes thermal management less difficult, sure ... but there's now many fans and how many thermal sensors now? And whose team is doing the programming for that subsystem? Once again, its a bunch of work that gets expensed against just the Mac Pro product line, and its increasingly limited production quantities.

FWIW, this is part of the fallacy that comes from just looking at piece parts (CPUs that now cost $300, etc): it doesn't get into the management of all of the product development's fixed costs, and how these ultimately have to get paid for...


-hh

 

[Neodym]

I assume the “Pro” in Mac Pro means “Professional”. That said everything else Apple does is consumer based. [...]

Is the current Mac Pro overpriced? Well that depends. For my case the answer is a decisive no.

It depends - exactly. The problem that many people do have is that Apple does not offer any other machine that can be upgraded and enhanced (hard drives, Ram, graphic card) anywhere close as far and easy as the MacPro (not to mention the fantastic thermal management). Those people don't actually need all of the "Pro" side features of the MacPro, but have to pay for them nevertheless.

With MP prices drifting further upward (which may be okay for a business customer) the gaping hole in Apple's computer lineup (a.k.a. the infamous missing xMac) becomes even bigger. If Apple would be consequent in focusing the consumer market, they would offer a more competitive tower model (or a truly upgradeable iMac with external 2.5"-hard drive slots and user-serviceable graphic card) for the non-business customer - and suddenly no one would bitch and moan anymore about MacPro pricing: Business customers don't really care anyway and non-business customers would have a viable alternative!

But that will probably stay to be wishful thinking... :-/

 

[zephonic]

In 2006 Apple came out with the mac pro. The starting mac pro contained 2x 2.66 dual core processors from intel at this price point, so at the time, the cpus necessary to build a $2500 machine retailed for roughly $1380. It used 2x Xeon 5150s.

Today the machine occupying this price point has been reduced to four slots for ram and a W3530 cpu that retails for $300.

The 6 core should actually be just under the $3k mark. There are so many areas it could use just little improvements on drivers, features, etc, but Apple has been kind of sitting on the line for the past couple cycles, and Intel has been slow as hell with rolling out new Xeons.

This. +1

Not sure it was a total misstep but I still hold to my original conclusion that the Mac Pro is a “sidebar” business at Apple. I assume the “Pro” in Mac Pro means “Professional”. That said everything else Apple does is consumer based. The motives stated earlier regarding attraction of “anti establishment” culture of artists remains valid and we all hope that Apple continues to maintain a level of support and products for the professional community. Time will tell.

While this makes sense considering what we know of SJ's vision, Apple has served the desktop publishing and creative pro market for 25+ years and if it wasn't for them (and Adobe) the original 1984 Macintosh would have been relegated to a mere footnote in computer history.

But the average consumer has always been SJ's #1 target, no doubt.

So is the Mac Pro overpriced? No. The value of a trustworthy product with little to no maintenance with virus and other garbage has a price. The current Mac Pro gets you a lot for the money.

In the end the total cost of ownership (original purchase + software applications + maintenance + downtime + enhancements) over a five year life the Apple Mac is a winner hands down and a real bargain for OUR use….may not the same for your use…..

It has been said MANY times on this here forum, but I'll repeat: the single-CPU MacPro's are not just overpriced, they are absurdly expensive.

What R&D? Keeping a tower in a product lineup is one of the easiest things to do. Everything exists already. Case, procs, mobo, memory. It all comes from other sources. They tweak very little. Just add the new board and procs and you are done.

No. First of all, Apple uses bespoke logic boards for the MacPro. Second, if you compare a current Bloomfield/Westmere with the 2008 (or earlier) models, you'll see they are very different in design.

I guess Apple COULD just throw them together like you say, but I think it is a testament to their design philosophy and quality assurance that they don't.

 

[nanofrog]

Each change in CPU architecture, a new graphics card and so forth will need to undergo QA testing for every OS update, for every Application and so forth...and I'm sure that since the Mac Pro's are the only hardware that Apple's selling that has dual physical CPUs that means that all of those specific QA costs don't get amortized across other hardware division.

Exactly.

This is the part of R&D that seems to be overlooked, as QA costs are a large portion of the total R&D, and can exceed what's spent on PCB design (not as high for Apple, as they don't offer the hardware options other vendors do, but as you mention, it's only attributable to the MP which has a small sales volume vs. the consumer products = higher cost per system comparatively speaking).

 

[Neodym]

Exactly.

This is the part of R&D that seems to be overlooked, as QA costs are a large portion of the total R&D, and can exceed what's spent on PCB design

<Bitching mode>
Apple recently seems to save on QA anyway, if you look at the (subjectively felt) increasing quality issues in their products...
</Bitching mode>

 

[BornAgainMac]

The Mac Pro needs to be mainstream. It needs to replace the iMac. The iMac really makes more sense with limited components such as a ARM processor and priced as a sub $1000 machine.

 

[Neodym]

The Mac Pro needs to be mainstream. It needs to replace the iMac. The iMac really makes more sense with limited components such as a ARM processor and priced as a sub $1000 machine.

It's quite possible that the MacPro goes ARM earlier than the iMac. Instead of putting two expensive (Xeon) CPU's in there (with up to 12-16 cores), they could as well implement a board with 8 inexpensive ARM quadcores. As they already purchase/produce ARM's (the Ax chips are based on ARM architecture) in huge quantities for the iOS devices, they could probably get really good prices for an even higher volume in order to support MacPro's (and probably in the long run also other machines like Mac mini, iMac, MacBooks etc.).

I'm not sure whether the performance of the ARM architecture is already there yet, but i'm convinced that this is being researched in the Apple labs (there also have been rumors in that direction some time ago).

Precedence also exists - supercomputers have used inexpensive PC CPU's in the past and continue to do so today (only recently AMD Opterons have been announced to be used in upcoming CRAY supercomputers).

Performance is only a question of quantity and depends on the price points of individual components. Balance of scale however demands a good price point in the quantities Apple uses for ARM chips, while the system is already developed actively towards massively multi-CPU/multi-Core environments (think e.g. of Grand Central Dispatch).

And if 8 ARM quadcores are not enough to provide the planned performance - simply add another batch of 2-4 and eventually you will be there (while costs are probably still in the same ballpark as with the "standard" server/desktop CPU approach, if not lower).

Apple developed the x86-OSX in parallel to the PPC version for years - i see no reason they could not do this again and currently have an ARM-based OSX developed in parallel to the x86 version officially in the market today!

 

[dansmac]

It's quite possible that the MacPro goes ARM earlier than the iMac. Instead of putting two expensive (Xeon) CPU's in there (with up to 12-16 cores), they could as well implement a board with 8 inexpensive ARM quadcores. As they already purchase/produce ARM's (the Ax chips are based on ARM architecture) in huge quantities for the iOS devices, they could probably get really good prices for an even higher volume in order to support MacPro's (and probably in the long run also other machines like Mac mini, iMac, MacBooks etc.).

Cool! Hey, I could ask BMW to install a couple Ford Fiesta engines in my BMW M5 E60. That would save some money? Get real!

 

[Neodym]

Cool! Hey, I could ask BMW to install a couple Ford Fiesta engines in my BMW M5 E60. That would save some money? Get real!

Unfortunately mechanics get in the way there - even if the basic idea would be applicable as well!

In the IT world it looks a little different (which is why car comparisons often don't fit - as here). If you want car comparisons, think of replacing a huge (and heavy) central explosion motor (like in the M5) with 4 lightweight electro motors (one per wheel), which together have more power (-> torque and weight/performance ratio) and cost less than the M5 motor due to mass production and much lower complexity.

 

[prowlmedia]

So here we are. The Crazy One’s – the Outcasts – but unfortunately the very few. From a pure company financial point-of-view continuing with the Mac Pro makes no sense. But Apple is different by design. Let’s just hope the new leaders at Apple in the post Jobs era see the value of maintaining a close association with the artists and crazies and a continued support for a professional product line.

My 2 cents!

If they kill off the pro... who and WHAT are are going to create all this great iPhone and iPad content. Do you think Apple are going to let the content creators drift off to Windoze....?! OK, apple makes way more money from iOS and laptops... but that's all relative... they are selling more of everything! I bet they are still selling more Mac Pros now than they did 5 years ago - iOS is driving all mac sales... Hell I am one of them. won't touch windoze again. Apple are just very good at wringing every possible buck from a product iteration as they can

The server enterprise market is different - Apple didn't rely on that to drive anything.

But these financial analysts people don't understand the need for Big desktops. Thinking Photoshop is the be-all and end all of computer graphics.

Graphics, video and content creation apps require very high end kit. Apple work with RED and and their 5K video requires massive amounts of Ram and processing power... iMac is not going to cut it in the if ever... cos there will always be a need for a more powerful machine.

They need to house External cards like the red rocket blackmagic, high end sound cards and other breakout boxes.

Also given the benchmarks..
Mac pro (2010) is still almost twice as fast as the imac (2011)
My Mac pro (2008) is just about as fast. Probably fast as it's go 24GB Rams and SSD's. and a Nvida 285 running Cuda.

let alone the Screens - now I like the imac screens - but even apple know they don't cut it for Video colour correcting.

oh and the memory - Upto 64gb in the Pro. 16 max in the iMac.

Macrumors - Apple's Processor Options for Early 2012 Mac Pro Begin to Firm Up

Do you know what this is... apple analysts talking rubbish until they can start making stuff up about iPhone 5

It's this simple...We'll have 16 Cores (32 threads) coming soon.... then 32 or whatever intel come up with next in a few years.

----------

No. First of all, Apple uses bespoke logic boards for the MacPro. Second, if you compare a current Bloomfield/Westmere with the 2008 (or earlier) models, you'll see they are very different in design.

Yes they do... but the board as I understand it is developed with intel.

The case and other parts are the simplest things to design. Must be Dell and HP redesign their rubbish about 4 times a year...

But desktop design is way simpler than notebook phone and Tablet design. It really is mostly selecting the right componant and making them all work well together. Something windoze has never ever been able to do.

 

[damir00]

If they kill off the pro... who and WHAT are are going to create all this great iPhone and iPad content. ...Apple work with RED and and their 5K video requires massive amounts of Ram and processing power..

There are 10,000 people cranking out HTML5 and Wordpress for every person who actuallys "works with RED".

Fewer and fewer people need high end kit for "creation".

 

[goMac]

Exactly.

This is the part of R&D that seems to be overlooked, as QA costs are a large portion of the total R&D, and can exceed what's spent on PCB design (not as high for Apple, as they don't offer the hardware options other vendors do, but as you mention, it's only attributable to the MP which has a small sales volume vs. the consumer products = higher cost per system comparatively speaking).

Q/A isn't a significant amount of the revenue the Mac Pro generates though. Even if Apple only sold 10k Mac Pros every year (which sounds very low), that's likely 10 million in profit. Is Apple spending 10 million on Mac Pro Q/A?

I'd bet good money there are only about a dozen people on the entire Mac Pro team.

If there was an issue, I would say it's a lack of caring on Apple's end, not profits. Or hardware quality.

There are 10,000 people cranking out HTML5 and Wordpress for every person who actuallys "works with RED".

Fewer and fewer people need high end kit for "creation".

Two points:
• For people who work with RED, what are they going to use if not a Mac Pro? Windows is also abandoning the workstation market it seems (see: Windows 8), and Linux doesn't have the software. That's why you see a lot of people freaking out about the Mac Pro. Apple leaving the Pro market really is looking like it could mean the end of the entire pro market across all platforms. Microsoft seems to think that Apples going to continue with the pro market, and Apple thinks Microsoft is going to continue with the pro market, leading neither one to continue with the pro market.
• I don't work with RED footage and I find my Mac Pro essential. If I was working with 720p I'd find it essential. You don't need RED footage to appreciate the faster rendering times. And on jobs that depend on render/compile times, that extra time spent churning on an iMac is money.

 

[TennisandMusic]

Two points:
• For people who work with RED, what are they going to use if not a Mac Pro? Windows is also abandoning the workstation market it seems (see: Windows 8), and Linux doesn't have the software. That's why you see a lot of people freaking out about the Mac Pro. Apple leaving the Pro market really is looking like it could mean the end of the entire pro market across all platforms. Microsoft seems to think that Apples going to continue with the pro market, and Apple thinks Microsoft is going to continue with the pro market, leading neither one to continue with the pro market.
• I don't work with RED footage and I find my Mac Pro essential. If I was working with 720p I'd find it essential. You don't need RED footage to appreciate the faster rendering times. And on jobs that depend on render/compile times, that extra time spent churning on an iMac is money.

Not sure where you get this interpretation from, but it is not correct at all. MS cares quite a bit about the workstation market, and W8 is not changing that in the slightest. Just because it has a new interface does not mean capability is being removed. They are adding capability. I am personally far more worried about Apple abandoning "serious computing".

 

[janstett]

I think you're floating in the right currents, but you're forgetting a few things here.

• Yes, the original Mac was tiny ("never trust a computer you can't lift").
• The NeXT, while sleek, was still a desktop computer. The Apple II was unnecessarily large (I think up to the IIc).
• Undeniable fact of technology, the best stuff (processors, video cards) are desktop sized. If you want high power, cutting edge tech you need a desktop form factor.

However, especially in lieu of the last bullet point, Apple is no longer about THE BEST technology (from a horsepower/bleeding edge point) -- e.g. no USB3/Blu-Ray/eSATA, etc.

----------

Again, I don't think it has anything to do with financials. If you look at what's going on, the entire Mac line, excluding the Air, seems to be being re-examined. The Macbook Pro, which sells extremely well, also seems like it could be on the cutting block with the 15" Air rumors.

Look at Apple's executive team. Steve was one of the last Mac people. The remnants are all iOS engineers. I think you're more looking at a team that doesn't really care about the Mac.

Agreed -- and, in my opinion, the future of the Mac is to become a giant iPad with a 27" screen, keyboard, running some descendant of the current ARM processors. 3-5 years out.

The Mac is going to get caught up in iDevice's wake (and Apple's need to control everything, including the processor). See the battle by the people behind the original iPod and how they lost out to the iOS folks.

 

[janstett]

Cool! Hey, I could ask BMW to install a couple Ford Fiesta engines in my BMW M5 E60. That would save some money? Get real!

Automakers have grafted together two smaller engines to make one big one for a long time.

For example, Aston Martin's V12 is two Ford V6s joined together, essentially.

If I'm not mistaken, BMW joined two inline 6s at a 60 degree angle to create the V12 they used in the 8 series.

Doing this saves money on engineering, time spent in development, and allows reuse of parts.

The reverse also happens (e.g. the Porsche 944's engine was 1/2 of the 928's V8).

Be more careful with car analogies if you don't know much about cars.

 

[goMac]

Just because it has a new interface does not mean capability is being removed. They are adding capability. I am personally far more worried about Apple abandoning "serious computing".

The desktop UI on Windows is being completely abandon in favor of a tablet UI, far surpassing Lion's new UI features.

Can you imagine being a pro and having to work with a full screen start menu? Unacceptable...

The real shame is that if Apple keeps going with the Mac Pro, they could own the Pro market in light of Windows 8.

See the battle by the people behind the original iPod and how they lost out to the iOS folks.

This.

 

[nanofrog]

<Bitching mode>
Apple recently seems to save on QA anyway, if you look at the (subjectively felt) increasing quality issues in their products...
</Bitching mode>

Unfortunately, there's truth to this (and not just Apple - it's become commonplace for consumer electronics as more and more devices are actually computers, even if their functionality is aimed at a specialized task).

In a number of cases, component selection is another area that suffers due to accounting depts in my experience (i.e. base selection on cost only, not performance).

Another part of the problem as I see it, is the back-end; poor QC on the assembly line (no random component testing before a lot is used in assembly, and no random full scale testing on final products off the line <just get a Go/No-Go form of testing and place a "QC Passed" sticker on it>).

It's quite possible that the MacPro goes ARM earlier than the iMac. Instead of putting two expensive (Xeon) CPU's in there (with up to 12-16 cores), they could as well implement a board with 8 inexpensive ARM quadcores. As they already purchase/produce ARM's (the Ax chips are based on ARM architecture) in huge quantities for the iOS devices, they could probably get really good prices for an even higher volume in order to support MacPro's (and probably in the long run also other machines like Mac mini, iMac, MacBooks etc.).

I'm not sure whether the performance of the ARM architecture is already there yet, but i'm convinced that this is being researched in the Apple labs (there also have been rumors in that direction some time ago).

Generally speaking, ARM is being used like this for servers (HP has announced such a system - Redstone Project, but nothing is currently available yet).

It's also being examined for workstation use with GPU's for a GPGPU based solution (increases performance to workstation requirements). However, this requires that the software support this type of usage, which is a limiting factor ATM.

Q/A isn't a significant amount of the revenue the Mac Pro generates though. Even if Apple only sold 10k Mac Pros every year (which sounds very low), that's likely 10 million in profit. Is Apple spending 10 million on Mac Pro Q/A?

I'd bet good money there are only about a dozen people on the entire Mac Pro team.

If there was an issue, I would say it's a lack of caring on Apple's end, not profits. Or hardware quality.

For me, it's not the QA costs themselves, but fact the MP's sales volume is shrinking (negative growth), and how that reflects on the ROI.

This doesn't mean that the MP is currently unprofitable, but it will end up in a cycle of increasing MSRP and reduced sales until it reaches $0.00 profit, or worse, a loss if they continue (I don't see Apple continuing with the MP as a loss-leader product).

So it comes down to whether or not they want to play this game.

Given the ROI on the consumer products, I'm thinking they won't play it for long.

As per the number of developers, I'm not sure it even has full time personnel due to the Audio bug in 2009 (took nearly a year before the first fix release, and a couple of revisions before they finally solved it). This left me with the impression that they were pulled off of other projects during a brief break in those cycles. Assuming this is the approach they use for the MP, it could explain why the QA isn't that wonderful as those doing the development work aren't all that familiar with the product. Just a thought anyway...

 

[deconstruct60]

The desktop UI on Windows is being completely abandon in favor of a tablet UI, far surpassing Lion's new UI features.

FUD. Not it isn't. Mac OS X has had Dashboard for several iterations and it doesn't diminish the apps. Similarly,

http://blogs.msdn.com/b/b8/archive/2011/08/31/designing-for-metro-style-and-the-desktop.aspx

And exactly why are they spending time on a task manager for 64+ cores if it is is all "dumbed down" tablet stuff ????

http://blogs.msdn.com/b/b8/archive/2011/10/27/using-task-manager-with-64-logical-processors.aspx




The "classic" UI will be there for folks with stationary boxes with mice and Wacom tablets attached to them. All the user has to do is minimize usage of metro widgets and remove just about every tile from the "start" screen.

Just because they are going to support more than one mode doesn't mean the other mode is going away. There is a cmd prompt if you want one.

Can you imagine being a pro and having to work with a full screen start menu? Unacceptable...

Rubbish. One, it is only there before you started you apps. Once started there is no necessity to look at the "start" screen any more there was a necessity to look at the desktop and start menu before. This is akin to moaning how the boot animation somehow hinders "Pros".

Two, there is nothing that preclude metro widgets being created that are useful for "Pro's". ( a calendar summary , summary of what is up in your linkedin social circle , industry news that fits your filter , etc. ). Customization is a feature:

"...A critical part of creating a meaningful dashboard and launcher is enabling you to customize it to be yours ... "
http://blogs.msdn.com/b/b8/archive/2011/10/04/designing-the-start-screen.aspx

If you look down in the comments the windows team has tile that puts bug tracking info on the start screen. If a user can't make the screen useful for themselves that says as much about them as it does for MS design.

Lots of business people use dashboards every day.

 

[damir00]

You don't need RED footage to appreciate the faster rendering times.

We are rapidly coming to a point where, for folks to whom that actually matters, the answer won't lie on the desktop.

We're not quite there yet - but it's coming. At my current startup, we have immense computational needs, and are just wrapping up an engineering test where we moved virtually all of it up to EC2. The results have been spectacular - much faster run times, at quite a bit lower cost, than maintaining our own mini-farm of desktop/rack Big Iron. The results weren't even close - iMacs + Grid in the Sky >>>> Mac Pro on every desk.

It's not my field, so I haven't looked, but I'd bet dollars to donuts there is at least one video rendering offering already running as an EC2 "appliance".