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.
2009 Mac Pro in hand - unboxing etc.
- Thread starter WonderSausage
- Start date
- Sort by reaction score
Nothing optimized, came out to 47 seconds on 2.26/6GB on CS3
http://www.retouchartists.com does not allow for submission.
Well, since I will be doing compiling, raytracing, scientific computing, etc - I am a happy camper. Also, this is the year apps will be redesigned, I predict, to finally take advantage of all these cores, so no worries.
In a year, it should be cake to replace the cores with 6-8 core chips for 12-18 physical cores, or 24-36 virtual cores - running at over 3Ghz each for just an extra grand, assuming pin-compatibility with the next tock - with no RAM limitations.
The advantage is a year from now the 3.2 owners will be sad, and I will be happy
What a nice machine.
Yes, this is a little disappointing if you ask me. My old Q6600 PC with 4GB RAM did it in 45 seconds.
My 2008 2.8 Octo does 25 seconds . . . ran it again tonight (10GB RAM and 6400AAKS RAID0).
Maybe Geekbench isnt very efficient towards the new nehalem architecture and a new modified version is in order?
So did anyone get a quad core 2.66 09' mac pro? Any results at all? I am ordering first thing in the morning and am not sure whether to get the 2.66 quad or the 2.26 octomomma.
I know everyone has a hard on for the octo but if someone has some numbers on the quad 2.66, I would really appreciate it.
Thanks
Macsimus
I know everyone has a hard on for the octo but if someone has some numbers on the quad 2.66, I would really appreciate it.
Thanks
Macsimus
Mine:
CINEBENCH said:
So my ageing 2006 2.66GHz Quad is actually 20% faster than the new entry level Octo for single threaded tasks. That is insane. We all know that Apple has increased their profit margins by about $1000 per machine with these new Mac Pros but here we can see the true effect of these extra profit margins. Apple could have made the 2.66GHz Octo the base model at the same price point or even less but instead they chose to do this. Not cool.
New and Improved! Now With 2007 Mac Pro 3.0 Octad!
I added Spanky's, two 2007 3.0's, and a 2008 3.2
Some others can be found here: http://www.tabsnet.com/index.php?option=com_benchmark&task=list&bid=7&sysid=0
- Enjoy!
.
- 2009 2.26 Octad
Rendering (Single CPU): 2309
Rendering (Multiple CPU): 18088
Multiprocessor Speedup: 7.83
- 2008 2.8 Octad
Rendering (Single CPU): 3244
Rendering (Multiple CPU): 18907
Multiprocessor Speedup: 5.83
- 2008 2.8 Octad
Rendering (Single CPU): 3232
Rendering (Multiple CPU): 18245
Multiprocessor Speedup: 5.65
- 2008 2.8 Octad
Rendering (Single CPU): 3186
Rendering (Multiple CPU): 17464
Multiprocessor Speedup: 5.48
- 2007 2.66 Octad
Rendering (Single CPU): 2873
Rendering (Multiple CPU): 16615
Multiprocessor Speedup: 5.78
- 2006 2.66 Quad
Rendering (Single CPU): 2806
Rendering (Multiple CPU): 8832
Multiprocessor Speedup: 3.15
- 2008 3.2 Octad
Rendering (Single CPU): 3682
Rendering (Multiple CPU): 21221
Multiprocessor Speedup: 5.76
- 2007 3.0 Octad
Rendering (Single CPU): 3200
Rendering (Multiple CPU): 18829
Multiprocessor Speedup: 5.88
- 2007 3.0 Octad
Rendering (Single CPU): 3228
Rendering (Multiple CPU): 18132
Multiprocessor Speedup: 5.62
[*]2009 2.93 Octad
Rendering (Single CPU): 4074
Rendering (Multiple CPU): 25644
Multiprocessor Speedup: 6.29
- 2006 2.0 Quad
Rendering (Single CPU): 2184
Rendering (Multiple CPU): 7457
Multiprocessor Speedup: 3.41
I added Spanky's, two 2007 3.0's, and a 2008 3.2
Some others can be found here: http://www.tabsnet.com/index.php?option=com_benchmark&task=list&bid=7&sysid=0
- Enjoy!
.
Except that Apple have the CPU(s) and RAM on a daughter card that plugs into the main board. Who is to say that single CPU and dual CPU daughter cards can not plug into the same main board if designed appropriately? I am not saying that is the case, but it is a possibility. We won't know until someone gets a single CPU system.
S-
From what I read so far, considering quad 2.93 ~ octo 2.26 in price, the quad is a much better deal if you aren't doing lots of multi-threaded stuff. It seems from the scores that clock for clock, the new system is only slightly faster than the old with single threaded apps. In which case, your better off spending the money to get a fast 2.93 quad that will be on par with the old 3.2 for a little cheaper.
In the case of multi-threaded, go for the octo because clock for clock it creams the old ones... but you are starting with a much lower clock for the same price. In terms of ultimate speed, the new machines are way faster than the old but at a considerable price.
It looks like speed/price hasn't changed all that much unless you are working with single threaded stuff... which in that case get a quad and it's still a good speed/price. You just don't get the best of both worlds for cheap. I'm going with the quad 2.93 cuz I couldn't care less about using 16 cores, since I wont be.
That would be great if Apple opted to use the same main board / chipset for both models... Giving quad core buyers a simple upgrade path when Gainstown is below stratospheric pricing levels. However, I suspect that if they did use the same main board, they might as well have used the same cpu daughter board and just populated one socket. Given they didn't do this, leads me to believe they used a different main board, but we should know for certain soon enough.
From all the groaning on this board about the lack of RAM and expansion on the quad, they should probably launch a dual socket quad... that would give them a more attractive entry level system that could easily be expanded to an octo with tons of RAM.
Thanks for the reply but the only thing stopping me from ordering the quad anything is the fact that the RAM can only be upgraded to 8GB. The way I look at it is that 8GB will last me a year, maybe 1 and 1/2. The 32GB max on the octo is more appealing for longevity. I used to have a 2006 2x2.66 dual cores system (quad) and it did pretty much all that I needed it to do. With new software constantly coming out and taking advantage more and more of multi-cores and threading, I almost want to buy the octo 2.26 for the long run. What do you all think? Thanks
If RAM is a real big factor, then you don't have much choice. But I still strongly believe that the quads will be able to use 16GB of RAM.
Would you buy a 2.66 quad core with 16GB of ram for $5499 ($3000+ for RAM) or a 2.66 octo with 16GB of ram for $5199? hmmmmmmmm, real hard decision there.
Thats my guess why they don't include the option, no one would go for it. Of course in 2 years when you need that 16GB, the price will drop from $3000 to about $500 I'm sure. And that is apple prices BTW.
Would you buy a 2.66 quad core with 16GB of ram for $5499 ($3000+ for RAM) or a 2.66 octo with 16GB of ram for $5199? hmmmmmmmm, real hard decision there.
Thats my guess why they don't include the option, no one would go for it. Of course in 2 years when you need that 16GB, the price will drop from $3000 to about $500 I'm sure. And that is apple prices BTW.
It seems odd to me that the Quad can't take 16 when the Octo can take 32. I know this is what Apple has said, but let's wait a few days and see what people can do in practice...
What to do?
I'm in a bit of a quandry here as I need to run 2 monitors, XP on Fusion and Dragon voice recognition. I've run Dragon both on my MBP 2.8 (latest) as well as my home Mac Pro 1.1 2.66 with infinitely better results on the MP 1.1. I'm trying to decide between the 2.26 octo vs. 2.9 quad for work (or even the 2.66 quad). The MP is the only mac that'll run dual monitors and the photo/movie management is better for me on the Mac. Any thoughts?
I'm in a bit of a quandry here as I need to run 2 monitors, XP on Fusion and Dragon voice recognition. I've run Dragon both on my MBP 2.8 (latest) as well as my home Mac Pro 1.1 2.66 with infinitely better results on the MP 1.1. I'm trying to decide between the 2.26 octo vs. 2.9 quad for work (or even the 2.66 quad). The MP is the only mac that'll run dual monitors and the photo/movie management is better for me on the Mac. Any thoughts?
Wait for Snow Leopard. This "geekbench" program will be recompiled on different architecture and the numbers WILL be different.
Makes me wonder how much faster mine would have been.
Only marginally, I'm sure. I had your exact specs plus a GT 120.
Huh?
You mean after months and months of talking about waiting to buy this revision you didn't ?
DAMN she's got a nice ass!!!
Congrats OP...I will be forever jealous.
Yep! By one of the screws you can see a reflection of light that look beautiful!
Actually, they shouldn't change much. It's testing CPU and RAM. It's already multithreaded correctly.
The only reason they'd significantly change would be due to Snow Leopard being less bloated and pissing away fewer CPU cycles on heaven knows what.
Unless they have it test OpenCL, at which point the results are no longer valid when compared to non-Snow Leopard setups, which defeats the whole purpose since it's hardware-dependent, not software-dependent, and is "theoretically" supposed to give an independent assessment of performance.
Possibly but I've been testing some "software" on Snow Leopard and I can say this much. On my iMac 3.06 w/NVidia 8800GS QT-X uses about 10-18% CPU in SL. playing an H.264 960X540 video. On Leopard it uses on average 30% CPU. I'm guessing Grand Central has something to do with this not just a new Quicktime app because the memory usage is actually more in SL than in Leopard. But that may be due to debugging.
So I'm still guessing when geekbench is recompiled under SL the numbers will change. Especially if they decide to take advantage of the multi-Threading capabilities and Grand Central.