My 8-core future - 2.26 or 2.66? That is the question.

[sigmadog]

I'm getting one of the new Mac Pros in a couple weeks. I've looked at the merits of backtracking to 2008 models and decided not to go that route. The new Mac Pros are my future. End of debate.

The question I am batting back and forth is whether to go for the 2.26 Octad or the 2.66 Octad. The basic question I need to answer in the next couple weeks is:

Is a 17% increase in processor speed worth an extra 39% in cost?

That's what it boils down to for me. I'm anxiously awaiting some benchmarks and testing by objective online sources (not that I don't appreciate the stats contributed here - they've been very helpful in narrowing down the field) in order to make a determination whether that 17% will really make a difference to me.

Whichever version I buy, I will order it with the ATI Radeion 4870 and upgrade to 16GB RAM and add Two 1TB Drives from a secondary source (OWC, my favorite). For the record (I've mentioned this a few times, so at the risk of being too repetitive), I'm upgrading from a Dual 2.0 G5 (June 2004) PowerMac.

My initial thoughts are that, considering where I'm coming from, the difference between a 2.26 and a 2.66 will not be all that noticeable. It will most likely be lights out faster than my current setup.

Also, because the type of work I do is not likely to change all that much ( mostly graphic design with the CS4 Design package, an increasingly large amount of digital illustration using Painter X, with some 3D and web work thrown in), I'm wondering if the extra speed provided by the 2.66 will be wasted on me.

Finally, I plan on keeping this computer as my main workhorse for at least 8 years(!). This seems like a long time (I'm used to a new computer every 5 years or so), but I'm increasingly convinced that most of the future advances in computer technology will largely benefit those involved in motion and video production, and that the static arts that I do have reached a level of proficiency that will likely remain stable (edit to add: aside from improvements in software) for some time to come. Given this scenario, I don't think an 8+ year lifetime is a radical proposition given the capabilities of the machine and my projected needs.

Bottom line: I'm leaning towards the 2.26, but don't want to sell myself short when I consider how long I plan on using this beast.

So, while we all wait for BareFeats, and others to provide testing data, anyone else have any thoughts?

 

[Tesselator]

whether to go for the 2.26 Octad or the 2.66 Octad. The basic question I need to answer in the next couple weeks is:

Is a 17% increase in processor speed worth an extra 34% in cost?

the type of work I do is not likely to change all that much ( mostly graphic design with the CS4 Design package, an increasingly large amount of digital illustration using Painter X, with some 3D and web work thrown in), I'm wondering if the extra speed provided by the 2.66 will be wasted on me.

Yeah, while I personally think the Pros are a little over priced right now - if I were in your situation with your goals then I'd go for the 2.26.

Finally, I plan on keeping this computer as my main workhorse for at least 8 years(!).

In just 3 or 4 years you'll be able to upgrade those 2.26 procs to 2.93 ones (or even higher maybe) for less than $800 total. If things go the way they have in the past that is.

 

[jjahshik32]

If its 8 years, definitely go with 2.66GHz or might as well jump to the 2.93GHz.

 

[wheezy]

If you want it to last longer, then the 2.66 is the way to go. 2.93 if you can...

 

[Cliff3]

Expecting 8 years of useful life out of a computer is unreasonable. In 2001 I was running a PII 300MHz laptop that was around 3 years old at the time and I retired it early in 2002. I don't recall what sort of desktop machine(s) I had at the time - probably some flavor of PIII with a 1 GHz clock. Just try to imagine where we'll be 8 years from now.

I figure on a 3 year replacement cycle for my laptops, and that cycle is tied to AppleCare coverage. I just ordered a 2.26 octad - it might go 5 years but that will probably be pushing it. If you need to push the life out to 8 years to justify a 2.66, then I would opt for the 2.26 and a shorter replacement cycle.

 

[VirtualRain]

Is a 17% increase in processor speed worth an extra 39% in cost?

Would you consider a 30% increase in processor speed for a 10% reduction in cost?

If that sounds more appealing, then you should get the 2.93 quad which is still a multi-threaded beast.

 

[Daim]

go with 2,26.. in 3 years you can upgrade those to 3,8

 

[sigmadog]

Expecting 8 years of useful life out of a computer is unreasonable.

It depends on what you use it for.

My use is primarily for graphic design and illustration for print. The state of technology at present has created what I consider to be a level of speed and ease that likely will not be improved by faster, more powerful technology - at least for a print designer like me. The key advances in my industry will, I think, come from improvements in software that will take advantage of existing processor technology (multi-processors, for instance) for the foreseeable future.

When desktop computers for publishing first came on the scene in the late 80's and early 90's, they lifted my industry away from T-squares and technical pens into the digital age. However, as neat as my Mac IIfx was, it was still pathetically slow and underpowered for many design/publishing tasks. Since then, there has been a steady progression towards faster, more powerful computers (leaving software aside for the moment) so that today, the computer is essentially the fastest link in the chain. Creating static imagery today is more a matter of memory, storage, and creativity, because the processing power is definitely there.

The next big leap for computer technology has been and will continue to be for video and motion media industries. Us publishing types don't need more speed - we just need software that takes advantage of what we already have.

So, assuming my 2.26 or 2.66 Octad maintains its processing power and does not degrade in ability (i.e. break or slow down), it should easily accomplish the tasks I require for the print/design industry for at least eight years.

Which brings me back to my question: Is a 17% increase in processing power worth 39% increase in price?

Right now, I'm thinking the answer is no. But I anxiously await some comprehensive benchmarks to enlighten me.

 

[sigmadog]

Would you consider a 30% increase in processor speed for a 10% reduction in cost?

If that sounds more appealing, then you should get the 2.93 quad which is still a multi-threaded beast.

Not even gonna consider it, because I want lots of RAM.

Until the quads are confirmed as being able to handle 4GB sticks, I'm shutting the door on that possibility.

 

[Cliff3]

Which brings me back to my question: Is a 17% increase in processing power worth 39% increase in price?

Right now, I'm thinking the answer is no. But I anxiously await some comprehensive benchmarks to enlighten me.

You're really the only person who can answer that question. Most application software has not yet evolved to take advantage of multiple core processors. As new versions of software are released, that should enhance the effective performance of multi-core machines. If you need the clock speed now to do your work effectively, then you need it and the cost difference is secondary to that.

I still stand by my statement about an 8 year lifespan being an unreasonable expectation. Even if your applications have achieved some level of functional stability, the underlying operating system will continue to change in response to new requirements and capabilities. Those changes will spur changes in the applications you use, and likely result in a demand for newer hardware.

 

[jjahshik32]

You're really the only person who can answer that question. Most application software has not yet evolved to take advantage of multiple core processors. As new versions of software are released, that should enhance the effective performance of multi-core machines. If you need the clock speed now to do your work effectively, then you need it and the cost difference is secondary to that.

I still stand by my statement about an 8 year lifespan being an unreasonable expectation. Even if your applications have achieved some level of functional stability, the underlying operating system will continue to change in response to new requirements and capabilities. Those changes will spur changes in the applications you use, and likely result in a demand for newer hardware.

I have to disagree. These days speed increase in the processor speeds are not that dramatic and each year it's just an incremental increase in performance. If you wanted to, you could still chug along just fine on a p4 3.0 ghz desktop.

Also think about the 2.66ghz quad that's 3 years old now, you could still use that machine just fine in the next 3-5 more years, especially just for printing and storage capcity is there as well.

 

[PowerPaw]

If you do the ROI calculations over 5 years for the 2.26, you'll need to keep the 2.66 for 7 years for the same return and the 2.93... well forget about it.

Personally I'd take the 5+ year depreciation route and not 8 years on the 2.66GHz as this represents the best value, especially as you don't state that you need the fastest possible computer to incorporate into your ROI, i.e saved seconds in completing a particular operation = $$$.

One other thing... nobody knowingly depreciates over 8 years for PC hardware as you can't guarantee that spare parts will be available over that period so if the machine breaks you'll have to write down the remaining years as a loss.

 

[Shownarou]

Remember that Adobe just recently released multi-core enhancement for the Windows side of things. I can only imagine that they will do the same for OSX soon enough. The processor is going to become more important in the next few years even for things such as indesign and illustrator.

That said, I do believe that the 2.26 will handle it nicely. I have a 400mhz powermac g4 that was just last year retired from a big newspaper company.

While 8 years may seem like a lot, you don't need to be on the cutting edge to do your work well. It may feel nicer, but it's definitely not necessary. I say go with the 2.26.

 

[RebootD]

While 8 years may seem like a lot, you don't need to be on the cutting edge to do your work well. It may feel nicer, but it's definitely not necessary. I say go with the 2.26.

My co-worker was still using a G4 with 2GB of ram and trust me it was far from a happy experience. Plus what is the point of buying an expensive machine so you can work faster, only to keep it so long it no longer serves that purpose?

If you are buying a MacPro for speed, 3-4 years tops then it's off to ebay and buy a new one. Now if I can just make some more money so I can buy one I'd be happy.

 

[Cliff3]

I have to disagree. These days speed increase in the processor speeds are not that dramatic and each year it's just an incremental increase in performance. If you wanted to, you could still chug along just fine on a p4 3.0 ghz desktop.

Also think about the 2.66ghz quad that's 3 years old now, you could still use that machine just fine in the next 3-5 more years, especially just for printing and storage capcity is there as well.

Actually, an 8 year old state of the art machine would be a 1.7 GHz P4 and I would rather not be using one of those today. A 3 GHz P4 didn't arrive until the end of 2002.

In terms of the tax code and depreciation schedules, computer hardware is assumed to have a 5 year life. That's probably about right for this sort of machine, and a bit optimistic for a laptop. That doesn't mean the machine has no intrinsic value at the end of 5 years, but in all likelihood its ability to solve the problem for which it was originally purchased will be limited.

 

[jjahshik32]

Actually, an 8 year old state of the art machine would be a 1.7 GHz P4 and I would rather not be using one of those today. A 3 GHz P4 didn't arrive until the end of 2002.

In terms of the tax code and depreciation schedules, computer hardware is assumed to have a 5 year life. That's probably about right for this sort of machine, and a bit optimistic for a laptop. That doesn't mean the machine has no intrinsic value at the end of 5 years, but in all likelihood its ability to solve the problem for which it was originally purchased will be limited.

End of 2002 so it makes it around a little over 6 years already and I'm sure you can still use it just fine for the next year or two if you really have to.

I know many professionals still working with dual 2.x powermac g5 machines.
I think people can get by just fine for 7-8 years but its mostly the wanting the newer faster machines that feels like you need to upgrade.

Keep in mind that compared to p3 or p4 from 1999-2002 to the machines now with 4 cores and 8, 16 will last you MUCH longer than back then. Think about it for the past 3 years multithread hasn't even been really fully utilized until just recently.

 

[Cliff3]

End of 2002 so it makes it around a little over 6 years already and I'm sure you can still use it just fine for the next year or two if you really have to.

I know many professionals still working with dual 2.x powermac g5 machines.
I think people can get by just fine for 7-8 years but its mostly the wanting the newer faster machines that feels like you need to upgrade.

Keep in mind that compared to p3 or p4 from 1999-2002 to the machines now with 4 cores and 8, 16 will last you MUCH longer than back then. Think about it for the past 3 years multithread hasn't even been really fully utilized until just recently.

My buddy just bought his first Intel Mac, an Imac for his wife. His dual G5 isn't looking so good any more and I suspect he's going to want to get a good hard look at the machine I've got coming.

As far as longevity goes, I've been in IT for 25 years. The one thing you can count on is constant change. You buy hardware to stay ahead of the curve, but the curve always catches up. Usually about a year before you get budget for the inevitable upgrade.

 

[jjahshik32]

My buddy just bought his first Intel Mac, an Imac for his wife. His dual G5 isn't looking so good any more and I suspect he's going to want to get a good hard look at the machine I've got coming.

As far as longevity goes, I've been in IT for 25 years. The one thing you can count on is constant change. You buy hardware to stay ahead of the curve, but the curve always catches up. Usually about a year before you get budget for the inevitable upgrade.

I guess everyones needs are different some can last as long as 8 years before upgrading and others probably around every 4-5 years.

Anyway it seems like the op has made up his mind to keep this machine for 8 years for his usage and I really think the 2.66ghz model is the way to go.

If he would upgrade in 3-4 years I would recommend the 2.26ghz.

 

[rylin]

I guess everyones needs are different some can last as long as 8 years before upgrading and others probably around every 4-5 years.

Anyway it seems like the op has made up his mind to keep this machine for 8 years for his usage and I really think the 2.66ghz model is the way to go.

If he would upgrade in 3-4 years I would recommend the 2.26ghz.

Let's put this into perspective, shall we?
A photographer can easily go over $5000/year on equipment.
A car enthusiast is very likely to spend >$5000/year on his hobby.
A junkie can and will (if possible) spend > $5000/month on his hobby.

Now, $5000 over five to eight years?
Peanuts!

(Yes, I'm still trying to convince myself to get a MP.. and a new TV.. and possibly at LED Cinema Display.. and a new camera.. and a new vacation)

 

[jjahshik32]

Let's put this into perspective, shall we?
A photographer can easily go over $5000/year on equipment.
A car enthusiast is very likely to spend >$5000/year on his hobby.
A junkie can and will (if possible) spend > $5000/month on his hobby.

Now, $5000 over five to eight years?
Peanuts!

(Yes, I'm still trying to convince myself to get a MP.. and a new TV.. and possibly at LED Cinema Display.. and a new camera.. and a new vacation)

I agree, might as well spend the 5k+ and since it's for the op's business the machine will pay for itself probably in the matter of a few months.

I know what you mean, I really really want a mac pro preferrably the 2.66ghz model and can afford it but I also just bought a 40" Sony Bravia xbr6 and a mac mini + I do already have a 24" led acd but would love to use it with the mac pro and move the mac mini over to the Sony Bravia xbr6 (amazing HDTV btw) for use of an htpc.

 

[Toronto Mike]

I guess everyones needs are different some can last as long as 8 years before upgrading and others probably around every 4-5 years.

Anyway it seems like the op has made up his mind to keep this machine for 8 years for his usage and I really think the 2.66ghz model is the way to go.

If he would upgrade in 3-4 years I would recommend the 2.26ghz.

Is it possible that the pace of improvements for print work are reaching a leveling off? Printers can only lay down so many dots on paper, and with the recession unfolding - and threatening to last much longer than previously thought - printing companies are going to be stuck with the current technology because they cannot afford new machines if their margins are being drastically reduced.

A Mac Pro for print work purchased today could conceivably last longer than what was considered as an average lifespan until recently. Only the individual can decide what they need to accomplish with the software they buy. Maybe if Adobe's next CS release utilizes the multiple cores fully - then the machine that handles that well will be the one that will last far into the future for print work.

I'm just throwing this out as a basis for further thought.

Mike

 

[sigmadog]

I guess everyones needs are different some can last as long as 8 years before upgrading and others probably around every 4-5 years.

Anyway it seems like the op has made up his mind to keep this machine for 8 years for his usage and I really think the 2.66ghz model is the way to go.

If he would upgrade in 3-4 years I would recommend the 2.26ghz.

It's not a hard and fast, cast-in-concrete decision. Obviously the market can change, the technology WILL change, and my needs may change as well. However, assuming my needs remain the same, my business continues in the print-design model (and I have no reason to think otherwise), and the equipment performs consistently at it's top speed and power, I see no reason for it NOT to last eight years at least.

Wants and needs are different things. While I may want to have the fastest, most powerful computer every year, what I need is a computer that can handle all my work tasks efficiently and without compromising my creativity.

In five years, will my computer be useless in completing my work? Will it be slower than it was out of the box? Probably not. Therefore it will still be a good machine and, barring radical changes in software requirements, will last a few years more.

Speed is relative. In five, six, or eight years of course any computer will be slow by comparison to newer computers. But will it be slower getting my work done? I don't think so. My G5 Dual is just as fast as it was back in 2004 when I bought it, and honestly, if I had to, I could keep this for another couple years. But I committed to five years on this computer because I knew that faster computers were on the horizon and I would benefit from the speed they offered. Now, I'm not so sure I would benefit all that much from faster computers down the road, which is why I'm considering expanding the lifetime of my next system.

Anyway, this is an interesting conversation. But I got to go hang out with the wife, who, by the way, I have committed to for 20 years of service at least. She's still working good at sixteen years, so we'll see.

 

[Toronto Mike]

One other thing... nobody knowingly depreciates over 8 years for PC hardware as you can't guarantee that spare parts will be available over that period so if the machine breaks you'll have to write down the remaining years as a loss.

I know that I want to keep my machine going for a long as I can - which might be as long as the machine physically lasts. So I've bought spare parts that I anticipate breaking down within the time frame that I want my machine to continue in operation (power supply, spare graphic card) - items that might not be available in five years. To think that I'd have a dead machine if my graphic card goes and I cannot find another one would be a waste. If I never use the spare graphic card - at that far distant time there will another like minded individual that will be glad to take it off my hands.

For me, it is not only replacing a computer, but updating the software that is the killer. So if my software serves me then, as well as it does now - then my machine will always be fast enough. If only I don't buy the latest software at some point.

Mike

 

[sigmadog]

Is it possible that the pace of improvements for print work are reaching a leveling off?...

A Mac Pro for print work purchased today could conceivably last longer than what was considered as an average lifespan until recently.... Maybe if Adobe's next CS release utilizes the multiple cores fully - then the machine that handles that well will be the one that will last far into the future for print work.

Yes. I quite agree. You can't make print-work any faster. There are advances still to be made with regard to proofing and such, but for the creator of the artwork, I think the technology is pretty much where it needs to be. This is essentially my point in this thread.

Let's hope Adobe gets moving on multi-processing upgrades. I probably won't even consider upgrading from Adobe CS4 until they show some movement on the multi-processing front.

 

[MCHR]

This thread is similar to the dilemma I'm going through.

I currently do the lion's share of my computer work through the Adobe Suite of software (so single thread, right?). However. . . . I'd like to grow my apps into doing Maya, Rhino, possibly Alias, as I hear it will be written to run natively on a Mac for their 2009 release.

SO. . . while a 2.8 octo "sounds" like a good deal now, and I'd likely save a thousand or more, I may actually be ahead in getting the 2.26 or 2.66 octo for the future apps I'd be adding in 2009 or 2010.

I also use my computer for 4-6 years, so spreading that expense across that timeframe brings down the yearly hardware cost if I can run with the upcoming software and OS improvements.

No, I don't like the prices on the new Mac Pros. Pricewise, I think Apple has their collective heads up their @$$e$, so I'm trying to be creative or justify this new purchase to be used over a longer timeframe.

In any case, I'm coming from a 2.5 DP PowerMac, so any choice will be warp speed faster than what I have currently. I just want to insure it will be a usable platform for several years.