Today we are finally starting to see the GeekBench scores and it apppears that the quad 2.93 is going to be faster then the 2.26 8 core MP. For someone that mainly just does photoshop and dreamweaver and some gaming do you think I would be better off with getting the Quad 2.93 or the 2.26 8 core MP?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Should I get 2x2.26 or Quad 2.93 Mac Pro?
- Thread starter Greenhoe
- Start date
- Sort by reaction score
Absolutely! Unless you use massively parallel multi-threaded applications or lots of VM machines all doing stuff at the same time, you will NEVER leverage the power of an octo-core.
Whereas EVERY application benefits from increased clock speed!
Keep in mind that even the entry level quad machine can process 8 threads in parallel... something an octo core could only do before.
More benchmarks will confirm it, but for most people, a higher clocked quad-core will do them much better than a lower clocked octo-core.
The more I think about it I think the Quad core is the best option although I'm just worried about only being able to upgrade to 8GB for Ram, which I would like the option to upgrade to more later in the future.
Also if I was to get a 2.26 octo would I be able to upgrade the processors later on down the road? Are you able to upgrade the old Mac Pro processors to a higher clock speed processor?
Unless you are a professional GeekBench tester, Geekbench scores don;t mean anything.
Relax,
count to ten
Wait 2-3 weeks for reputable benchmarks,
keep an eye on sites like barefeats.com or diglloyd
This machine will last for 5+ years as a workhorse, so please don;t rush into a purchase based on initial, non-relevant posts on some forum.
BTW, here is what diglloyd says in response to your question
"What about the 4-core vs the 8-core model when you need more than 8GB memory?
The crippled 4-core model has only 4 slotsgoing beyond 8GB therefore requires 4GB modules at 8X the cost per GB of 2GB modules.
You can get an 8-core model with 8X2GB for about $3600 or a 4-core model with 4X4GB for about $5000 (and thats assuming the 4-core model even works with 4GB modules). Sometimes decisions are easy: if you need more than 8GB of memory, the 8-core model is far less expensive than the 4-core model! So shun the 4-core model. Its a lousy investment if your needs ever change."
If you use photoshop, you will eventually benefit from more than 8GB of RAM.
Relax,
count to ten
Wait 2-3 weeks for reputable benchmarks,
keep an eye on sites like barefeats.com or diglloyd
This machine will last for 5+ years as a workhorse, so please don;t rush into a purchase based on initial, non-relevant posts on some forum.
BTW, here is what diglloyd says in response to your question
"What about the 4-core vs the 8-core model when you need more than 8GB memory?
The crippled 4-core model has only 4 slotsgoing beyond 8GB therefore requires 4GB modules at 8X the cost per GB of 2GB modules.
You can get an 8-core model with 8X2GB for about $3600 or a 4-core model with 4X4GB for about $5000 (and thats assuming the 4-core model even works with 4GB modules). Sometimes decisions are easy: if you need more than 8GB of memory, the 8-core model is far less expensive than the 4-core model! So shun the 4-core model. Its a lousy investment if your needs ever change."
If you use photoshop, you will eventually benefit from more than 8GB of RAM.
I'm also interested in this as I'm in a similar situation with similar needs, although I have one more scenario for all of you... If a quad would be better suited to someone who doesn't use a lot of multi threaded apps, which quad is more worth it at this moment? I mean, how much faster do you think the 2.93 quad will be over the 2.66 quad? It would cost me another $450 just for this speed bump. Do you think it would be worth it?
The quad is a waste IMO.
You're better off going with a workstation/desktop from Dell or HP that offers a Core i7 rig for a cheaper price. You're limited to 8GB of RAM, and I don't call that a workstation at all.
Only the octo is worth it, even though it may not perform as well as the quad. Why? Because it can handle more RAM. In a few years, having more RAM will be much more important that having a higher clock speed.
You're better off going with a workstation/desktop from Dell or HP that offers a Core i7 rig for a cheaper price. You're limited to 8GB of RAM, and I don't call that a workstation at all.
Only the octo is worth it, even though it may not perform as well as the quad. Why? Because it can handle more RAM. In a few years, having more RAM will be much more important that having a higher clock speed.
A quad would be faster probably.
http://www.tomshardware.com/charts/desktop-cpu-charts-q3-2008/compare,826.html?prod[2179]=on&prod[2180]=on&prod[2271]=on
Or your existing 3Ghz machine.
http://www.tomshardware.com/charts/desktop-cpu-charts-q3-2008/compare,826.html?prod[2179]=on&prod[2180]=on&prod[2271]=on
Or your existing 3Ghz machine.
It's a philosophical debate at some level...
Like most purchase decisions, there is a diminishing returns curve to consider and where you need/want to be on that curve and how often.
Buy a quad core now and get the next latest Mac Pro in 2 years or buy an octo-core and stuff more RAM in it in 2 years to try and buy yourself some more time... it's really a personal investment decision.
I laugh at the person who referred to the quad core as "crippled"...
That person has lost perspective.
Again, I think the only ones who can really justify the cost of an octo core are the ones who's business and living depends on it. If every minute it takes a video to encode or a scene to render costs you money, then you can justify the obscene cost of the octo-core.
In my opinion, the quad core is the perfect Mac Pro for the enthusiast or recreational PC user who wants state of the art computing technology in a kick-ass package. Not everyone in here works for Pixar or Disney.
Like most purchase decisions, there is a diminishing returns curve to consider and where you need/want to be on that curve and how often.
Buy a quad core now and get the next latest Mac Pro in 2 years or buy an octo-core and stuff more RAM in it in 2 years to try and buy yourself some more time... it's really a personal investment decision.
I laugh at the person who referred to the quad core as "crippled"...
That person has lost perspective.Again, I think the only ones who can really justify the cost of an octo core are the ones who's business and living depends on it. If every minute it takes a video to encode or a scene to render costs you money, then you can justify the obscene cost of the octo-core.
In my opinion, the quad core is the perfect Mac Pro for the enthusiast or recreational PC user who wants state of the art computing technology in a kick-ass package. Not everyone in here works for Pixar or Disney.
Fair enough
I laugh at the person who referred to the quad core as "crippled"...
Except that you just admitted that the Quad only has a 2 year lifespan before needing to upgrade...?
And in terms of perspective, when my Mac was brand new, while it came with a mere 512MB ... it nevertheless had 8 RAM slots and an 8GB ceiling.
And here we are, FOUR Moore's Law cycles later...and yet today's Quad Mac Pro ... still has only an 8GB ceiling.
If its RAM could be expanded to 16GB, I'd be willing to agree.
-hh
Order
I have an order in for the Dual 2.26 --
I'm just about ready to cancel and order the Quad. Is it worth $500 to upgrade to the 2.93? Anyone else going up to that or staying with the 2.66?
I have an order in for the Dual 2.26 --
I'm just about ready to cancel and order the Quad. Is it worth $500 to upgrade to the 2.93? Anyone else going up to that or staying with the 2.66?
Why do we have to answer this question 500 times?
find our yourself. (hint, close your eyes and randomly click on any topic on the first page of this forum, and then read it)
I read other topics and what I can read is this: quad 2.92 is better for single thread application and octo 2.26 is better for multi-thread application.
My question was different! I asked if MAYA during rendering has some advantage with more core.
the other thing to remember is that Snow Leopard will make better use of more cores, and as time goes on, 3rd party applications will as well. So just because a higher-clocked quad beats a lower-clocked octo in photoshop today doesn't mean it will three years from now when you still have the same mac pro but Adobe is on CS 6.
Fair Question:
His question was fair and unique due to the new Mac Pro being released.
Instead of wasting time being rude -- why not answer his question or at least ignore his question?
His question was fair and unique due to the new Mac Pro being released.
Instead of wasting time being rude -- why not answer his question or at least ignore his question?
.......or 2008 2.8 - 8 core with the ability to add 32 GB of RAM and can upgrade the video card......seems like a better deal.
How long do you plan to keep the computer? It is my belief that in a few years most apps will be able to take advantage of multiple cores. If you plan to keep the computer for more than a couple years, you will probably be glad you have the extra cores.
this isn't as simple as it might have seemed. There appears to be a problem with mental ray and hyperthreading...
https://forums.macrumors.com/showthread.php?p=7245202&posted=1#post7245202
This is indeed the "Read the Tea Leaves" question...
...although it arguably could be as soon as Photoshop CS5, as Adobe is IIRC currently porting to 64 bit and other OS X stuff....and in sniffing around, it does seem that CS4 "Stonehenge" already had some GPGPU enhancements already in 2008...see link.
The bad news - good news is that since historically Photoshop has been on an 18-24 month release cycle, this would put the release of OS X based Photoshop CS5 to the March-September 2010 time frame.
That's "good" from the aspect that CS5 still has roughly a year from today for OS X's Snow Leopard, Grand Central, OpenCL, and whatever GPGPU technologies ... to be incorporated by Adobe's development team.
But its also "bad" if one is hoping for some more near term "which new Mac Pro?" decision guidance for anyone considering a near term (eg, 2009) purchase, which then puts us 400 days downstream into the old "don't buy - upgrades soon!" merry-go-round.
What would be tremendously helpful ...and not all that unreasonable to ask for... would be a non-committal comment from Adobe on how they think that their Betas seem to be forming up on the basic question. Obviously, it can't be any sort of firm promise, but merely a take on their observations so far.
Maybe after the NDA for the Mac Pros drop...my personal guess would be that late March (24th?) date that's been floating around.
-hh
My advice: wait and save your money
People seem to forget that Apple and Intel are corporations and their raison d'etre is to make lots of $.
I suspect that one of the goals of getting Snow Leopard out the door is to optimize the performance of the best Mac Pros (the 2009 ones) so that Apple can increase hardware sales. It may be that when Snow Leopard is released, the 2.26 model will beat the 2008 2.8 model in many programs (especially in a SL-optimized version of FCP). But for now it does not appear to be the case. Also, rarely do new OSes perform as well as promised; it takes a revision or two to get things right. So if SL 10.6 was released for WWDC in June, it may take until August to get 10.6.1 out the door and September for 10.6.2. By then the processor prices will have decreased, the PC competition will have been releasing their own benchmarks, and Apple and Intel would have tweaked their designs one more iteration.
Unless rendering time is crucial and you can justify spending >$5K for a decent 2.93 model, I think you are better off waiting at least three to six months to get more proven value. On the other hand, given that Snow Leopard's goal is to stimulate new hardware sales, the value of the 2008 and earlier models will drop even more, especially if Apple only minimally optimizes SL for them.
People seem to forget that Apple and Intel are corporations and their raison d'etre is to make lots of $.
I suspect that one of the goals of getting Snow Leopard out the door is to optimize the performance of the best Mac Pros (the 2009 ones) so that Apple can increase hardware sales. It may be that when Snow Leopard is released, the 2.26 model will beat the 2008 2.8 model in many programs (especially in a SL-optimized version of FCP). But for now it does not appear to be the case. Also, rarely do new OSes perform as well as promised; it takes a revision or two to get things right. So if SL 10.6 was released for WWDC in June, it may take until August to get 10.6.1 out the door and September for 10.6.2. By then the processor prices will have decreased, the PC competition will have been releasing their own benchmarks, and Apple and Intel would have tweaked their designs one more iteration.
Unless rendering time is crucial and you can justify spending >$5K for a decent 2.93 model, I think you are better off waiting at least three to six months to get more proven value. On the other hand, given that Snow Leopard's goal is to stimulate new hardware sales, the value of the 2008 and earlier models will drop even more, especially if Apple only minimally optimizes SL for them.
In two years time, isn't it safe to assume that DDR3 4GB dimms will be a lot cheaper than they are today? And if so, is there any reason to believe that the quad Nehalem Mac Pro couldn't be upgraded to 16 GB of RAM?