Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
What I meant was in multi-threading alone the score would be higher not because of the speed of the processor, but because of the fact that the applications are using all 12 cores. Same would be for the 6-core, once again.. using all cores. But the 12-core would be faster in rendering since it is using more than 6-cores and is multi-threaded.

For normal non-multi threaded apps, the 6-core would be higher since thats due to the speed of the processor. Take 3.33/2.40 -1 * 100 = 38.75 or close to 40% percent faster in overall performance(non multi-threaded apps only).

Photoshop is not multi-threaded like Final Cut, therefore the performance of the 12-core would be slower compared to the 6-core since the 6-core's speed of 3.33 is higher than 2.40 of the 12-core.

Probably the reason why you are not seeing any benchmarks for the 2.4 is because its embarrassing. its a much lower clock speed compared to the 3.33 6-core and the other 12-core based mac pros.


Please explain! That doesn't sound too logical to me.
 
Last edited:
Thank's goMac and PowerPCMacMan. Don't want to hi-jack here, so I'm off now, good luck with your decision OP...I still don't know what to do (and am luckily not forced to right now).
 
Of course there is.

But all that matters is what the user will be doing with the system, as that's what performance that matters to that particular person.


This is a blind, inaccurate statement.

It will be the case under specific conditions, but not ALL. Some still benefit from higher frequencies rather than core counts.

Keep in mind, that software is quite a bit behind hardware, and there will be applications that cannot be threaded to multiple cores (impossible/impractical due to the type of program).

Then there's the financial implications of such a statement for those that a DP core count are useless or nearly so (either don't use applications that can leverage the core count, or only a very small portion of their time is spent in them).

The "DP/high core counts all the way" comments have financial implications as well for those that can't really benefit from them.

The point was that the 4-core machine is limited in its potential compared to the 12-core. The 12-core can be updated with faster processors if needed and in the long run it will have more flexibility. The 4-core machine's only advantage is its clock speed and is easily overcome. The VM point is an example only to point out that you could also use the machine for purposes beyond original intent. If any of the software the op uses gets updated to use as many cores as possible, the 4-core is a limiting factor too.

Your opinion may be different but it doesn't make my opinion blind or inaccurate.
 
Probably the reason why you are not seeing any benchmarks for the 2.4 is because its embarrassing. its a much lower clock speed compared to the 3.33 6-core and the other 12-core based mac pros.

That's reading a lot into it. I don't know why it would be embarrassing. You could easily guess what it would be given the 8 core 2.4 or 12 core 2.66 scores that are available.

In fact, its about exactly what you would expect. Our lab just recently got a 2.4 12 core Mac Pro and I just GeekBenched it right now (only in tryout mode). In 32-bit mode it scored 17626. Compared to the 3.33 hex at 13808, the 2.4 8 core at 12599, and the 2.66 12 core at 20007.

That's almost exactly what you'd get from 1.5 times the 2.4 8 core score, or 2.4/2.66 times the 2.66 12 core score. And its a very nice number. Of course, you need to use all 12 cores somewhat frequently to make it worth it, but even at 50/50 single/multi threading its probably worth it. Each configuration holds a similar advantage over the other in either single or multithreading tasks (of about 40%), so you really can't go wrong.

However, the upgrade path for the CPU and RAM would make me lean towards the 12 core if usage was close to 50/50 for single/multi threading. I agree with another poster that said prices for the 1366 are not likely to drop. After all, Nehalem 1366s are still as expensive as their Westmere equivalents today. However, it wouldn't cost you $2000 to go from the 2.4 to the 2.66, because you could sell those pair of 2.4 hex cores for probably ~$800. So the real upgrade cost is likely closer to $1200, or if you bought used X5650s it might be only ~$800. Also, the 12 core may come in handy if you ever find yourself wanting more than 48 GB of RAM. Plus, you make wake up one day and find those single threaded apps you bought the hex for are now multithreaded...

I would agree that it would probably be a better option to just sell your machine and buy what ever is the latest and greatest at the moment instead, but that's not known for sure. The 2013 Mac Pro may never materialize, or it may turn into the 2014 Mac Pro.
 
Last edited:
The 12-core can be updated with faster processors if needed and in the long run it will have more flexibility. .

So can the Quad and the gains are far better (You get 2 more cores plus clock speed boost) Not just clock boost.

17213 is average perfect Geekbench scale score for 12-core 2.4GHz in 32-bit according to Mactracker. It will be 18300-18900 in 64-bit. 20300 for 32-bit 2.66 and 22130 for 3.06GHz.
 
17213 is average perfect Geekbench scale score for 12-core 2.4GHz in 32-bit according to Mactracker. It will be 18300-18900 in 64-bit. 20300 for 32-bit 2.66 and 22130 for 3.06GHz.


Because today a 2.66Ghz 12-core system popped up as a refurb here:
Is my assumption right, that the speed gained regarding single-threaded programs compared to a 2.4Ghz 12-core MacPro is neglectable? I'm speaking mainly about PS usage. Maybe some gaming, too.
 
The point was that the 4-core machine is limited in its potential compared to the 12-core. The 12-core can be updated with faster processors if needed and in the long run it will have more flexibility. The 4-core machine's only advantage is its clock speed and is easily overcome. The VM point is an example only to point out that you could also use the machine for purposes beyond original intent. If any of the software the op uses gets updated to use as many cores as possible, the 4-core is a limiting factor too.

Your opinion may be different but it doesn't make my opinion blind or inaccurate.

The cpus for the 12 core will make it completely impractical. Given their cost the used supply at cost effective levels isn't going to be very good. Much of the time it's more cost effective to replace the computer. As to the quad, its gains are much better and the cpus are cheaper. Someone already mentioned that though. Right now you shouldn't suggest any of these based on upgradeable cpus. If they had Sandy Bridge E and the idea was that they could be bumped to Ivy later, that would be much more of an upgrade path than being able to go to a slightly better 2010 era cpu.
 
The point was that the 4-core machine is limited in its potential compared to the 12-core. The 12-core can be updated with faster processors if needed and in the long run it will have more flexibility. The 4-core machine's only advantage is its clock speed and is easily overcome. The VM point is an example only to point out that you could also use the machine for purposes beyond original intent. If any of the software the op uses gets updated to use as many cores as possible, the 4-core is a limiting factor too.
You're only looking at core count though, and though important, it's not the only factor.

Yes, the base DP system's CPU's can be upgraded. But so can the SP (toss in a Hex for example, which assuming all cores can be utilized, will perform better than the 2.4GHz DP for both single and multi-threaded applications).

Both systems can be used for VM, but like the original use stated by the OP, the specifics have to be taken into account. For example, if a user only needs a couple of VM's for example, an SP would certainly be up to the task without breaking the bank. Now if they're wanting to run say 10 VM's, then a DP would be the better way to go (obvious in this case).

The specific usage is absolutely critical though, which is why general/blind statements of SP or DP is better than the other aren't universally valid (this aspect is all I'm having issues with, as reality of usage patter = conditional; thus the case by case nature of which system is the better fit at the time of purchase).

Now granted, I suspect you're thinking "what if" in regard to things the user may be doing in the future. But all one can really do, is plan based on solid information at the time of purchase (i.e. don't buy a system for high VM counts if they don't expect they'll need it within the usable lifespan of the system).

There are other factors that contribute significantly to the overall performance of a system as well, such as GPU (i.e. GPGPU processing capabilities if the software supports it), as well as a sufficient storage system for whatever applications being used (i.e. don't want to choke on a single HDD when the application may be generating 1+ GB/s in data transfers).
 
6 core plus more RAM and SSD

The OP indicated he/she is on a budget - that apparently goes up to $3800 for the base 12-core but not beyond. Seems to me the best choice given the apps listed would be the 3.33MHz hex ($3000) and then use the remaining $800 for upgrading the RAM and at least one SSD via non-Apple sources.
 
Because today a 2.66Ghz 12-core system popped up as a refurb here:
Is my assumption right, that the speed gained regarding single-threaded programs compared to a 2.4Ghz 12-core MacPro is neglectable? I'm speaking mainly about PS usage. Maybe some gaming, too.

I wouldn't say "neglectable" as my spell check say's it isn't a word:)
It will be faster and boosts to 3.06GHz so you get "over the hump" so to speak on single threaded apps. 2.4GHz boosts to only 2.67GHz. If you are mainly concerned about PS then get the 6-core. It'll slaughter both 12-cores.
http://macperformanceguide.com/Reviews-MacProWestmere-Photoshop-CoresSlower.html
 
I wouldn't say "neglectable" as my spell check say's it isn't a word:)

:eek: mine says it's an english adjective..compromise: is negligible fine with you, then?


It will be faster and boosts to 3.06GHz so you get "over the hump" so to speak on single threaded apps. 2.4GHz boosts to only 2.67GHz. If you are mainly concerned about PS then get the 6-core. It'll slaughter both 12-cores.

Thank's for the info! As I wrote above, my workflow is about 50/50 between PS and rendering. So when I have a scene rendered in 1/3 the time I save much more time compared to some seconds or minutes in PS. I also thought that having actually more cores, I could render and work with PS at the same time..but yeah, I'm pretty messed up taking this decision! What do you think about getting the hex and eventually add 1-2 mini server (next gen) as render nodes?
 
What do you think about getting the hex and eventually add 1-2 mini server (next gen) as render nodes?

I personally wouldn't want to deal with minis as render nodes, however depending on you budget, it may make the most sense. What is your budget anyway?
 
:eek: mine says it's an english adjective..compromise: is negligible fine with you, then?

Yes. Apparently Apple spell check is not "with it" as normal internet search shows you to be very correct in use and spelling. So apologies. But same intent anyway.
If you can get it in your budget get the 2.66GHz 12-core for the render speed and overall depth. If it is between the 8-core 2.4GHz or even 12-core 2.4GHz then I'd get the Hex 3.33GHz as it has a better balance. But memory needs are paramount and 8 DIMM slots is 2x as flexible as the 4 in the single proc. Don't count on node anything making a purchase as there is always overhead that will diminish the money paid. I mean it is cool if you have stuff laying around but buying something just for that is a little premature. See if you can deal with the times after your new Mac. Then re-evaluate. You have 2 weeks to return if you buy new and you think another model would work better.
 
I personally wouldn't want to deal with minis as render nodes, however depending on you budget, it may make the most sense. What is your budget anyway?

Initally about 3000€ just for the machine, without SSD+RAM addition+maybe graphics. I assume you're in the US so what I saw, you can almost 1:1 the Euro to Dollar in terms of Apple-pricing. The 2.66 12-c is 4100 for the refurb, which seems to be interesting with the info Derbothaus gave about HT. I was hoping about the hex with 15-19% discount (for 2500€ I'd buy immediatly), but now it costs exactly the same as the new one.


I mean it is cool if you have stuff laying around but buying something just for that is a little premature. See if you can deal with the times after your new Mac. Then re-evaluate. You have 2 weeks to return if you buy new and you think another model would work better.

I hate having stuff laying around. And I wouldn't buy them right now anyway, but - even if it's a bit awkward - it would be at least a possible 'upgrade' in one or two years (for the hex). The nice side effext could be that i could use the mini as an HTPC...but that's unimportant.
Dunno about the 2-weeks return policy with refurbs in Europe yet.
 
Sorry for double post.
Refurb's come with 2-weeks return policy as well, so that hint is useful. I don't know about FCPX, I'm working with Maxwell and have 5 render nodes included.
It will be the Hex then, I guess.

to OP: why do you decide between the 4-core and 12-core? The 3.33 Hex is still cheaper than the 2.44 12-core system but a lot more appealing then the Quad!?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.