Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
From the GPU point of you are right, but concerning CPU performance the new iMac i7 should be on par with the 2010 Hex MacPro...


The Hex still kills it, don't get it twisted. Geekbench isn't the best but there is about 4000 pooints between the two.

Pass mark same thing.
 
The Hex still kills it, don't get it twisted. Geekbench isn't the best but there is about 4000 pooints between the two.

I said should...

According to this http://www.9to5mac.com/65186/new-imacs-get-benchmarked/

The Geekench values difference is quite smaller than you think...

The iMac i7 is on par with the MacPro Octocore and a little slower than the HEX.

MacPro HEX: 14049
MacPro OCTO: 12722
iMac i7: 12697

And all this on multi-threaded tasks (AFAIK GeekBench is multi-core optimized)...

On a single threaded task (real world performance) iMac and HEX should perform nearly identically, the OCTO should lag behind....
 
I said should...

According to this http://www.9to5mac.com/65186/new-imacs-get-benchmarked/

The Geekench values difference is quite smaller than you think...

The iMac i7 is on par with the MacPro Octocore and a little slower than the HEX.

MacPro HEX: 14049
MacPro OCTO: 12722
iMac i7: 12697

And all this on multi-threaded tasks (AFAIK GeekBench is multi-core optimized)...

On a single threaded task (real world performance) iMac and HEX should perform nearly identically, the OCTO should lag behind....

I honestly doubt these numbers. The average score for the 3.33 hex is about 16000, the current octad should score about 15000 considering that my 2.26GHz octad (2009 model) scores 14000 (average as well).

Still, 12000 for an all in one computer is astonishing! I really can't image what a dual hex Sandy Bridge Mac Pro will do to these numbers. 35 to 40000 sounds reasonable. :eek:
 
I honestly doubt these numbers. The average score for the 3.33 hex is about 16000, the current octad should score about 15000 considering that my 2.26GHz octad (2009 model) scores 14000 (average as well).

Still, 12000 for an all in one computer is astonishing! I really can't image what a dual hex Sandy Bridge Mac Pro will do to these numbers. 35 to 40000 sounds reasonable. :eek:

They need to offer the option for dual 3.33 GHz hexacores for the MP.
 
I honestly doubt these numbers. The average score for the 3.33 hex is about 16000, the current octad should score about 15000 considering that my 2.26GHz octad (2009 model) scores 14000 (average as well).

Still, 12000 for an all in one computer is astonishing! I really can't image what a dual hex Sandy Bridge Mac Pro will do to these numbers. 35 to 40000 sounds reasonable. :eek:

You have to check what version of Geekbench they are posting. 32-bit will net the 14000 and 64-bit test will get you into the 16000+.
 
I have a 5870 in my Mac Pro currently and I wonder what kind of GPU is going in the 2011 Mac Pro maybe 6970 or higher? I hope this year Apple will have the upgrade kit available soon after the release of the new Mac Pro's
 
Those CPUs are 130W, the current ones are 95W

Haha yes, I'm aware. But it's not like the PSU can't handle it. OWC does upgrades for it and as far as I know they don't replace the PSU. I just think they should at least have the option for it.
 
They need to offer the option for dual 3.33 GHz hexacores for the MP.

The 3.33GHz is such a slouch. Go 3.6! W5687. :D

A $1600 CPU that offers roughly 20 to 30% more than a $200 3.4GHz Sandy Bridge processor. Way to go Intel!

I can see why Apple doesn't offer the most powerful versions of processors. They are just too expensive for the little extra speed you get.
 
The 3.33GHz is such a slouch. Go 3.6! W5687. :D

A $1600 CPU that offers roughly 20 to 30% more than a $200 3.4GHz Sandy Bridge processor. Way to go Intel!

Uh, hate to break it to you, but the i7-2600 beats the X5687 (4-core 3.6GHz) by a good margin. So it is a 1600.00 cpu that is bested by a 200.00 cpu:)
6-core nehalem/westmere only chips that keep up. The 4's get embarrassed.
I am assuming this is the chip, there is no W5687.
http://ark.intel.com/Product.aspx?id=52578
 
I have a 5870 in my Mac Pro currently and I wonder what kind of GPU is going in the 2011 Mac Pro maybe 6970 or higher? I hope this year Apple will have the upgrade kit available soon after the release of the new Mac Pro's
The 7000 series is going to debut this summer, so my hope is that Apple already is working on drivers for those for a release later this year.
 
I said should...

According to this http://www.9to5mac.com/65186/new-imacs-get-benchmarked/

The Geekench values difference is quite smaller than you think...

The iMac i7 is on par with the MacPro Octocore and a little slower than the HEX.

MacPro HEX: 14049
MacPro OCTO: 12722
iMac i7: 12697

And all this on multi-threaded tasks (AFAIK GeekBench is multi-core optimized)...

On a single threaded task (real world performance) iMac and HEX should perform nearly identically, the OCTO should lag behind....

I just checked my base Mac Pro 2.26 with the ATI 5870 and Geekbench = 13,938.
 
Something unexpected

The 6970m is a better card, it takes the lead by 2% according to game-debate.com, since it is among the latest cards, it proves to be a better one.
 
Uh, hate to break it to you, but the i7-2600 beats the X5687 (4-core 3.6GHz) by a good margin. So it is a 1600.00 cpu that is bested by a 200.00 cpu:)
6-core nehalem/westmere only chips that keep up. The 4's get embarrassed.
I am assuming this is the chip, there is no W5687.
http://ark.intel.com/Product.aspx?id=52578

Crap, I totally missed that. Thought it was a 6-core chip. Thanks for the heads up!
 
Tests indicate that the Radeon HD 6970M shows an impressive performance. AMD's old top model, the Radeon HD 5870, is surpassed in every benchmark by about 50% – a testament to the new model's formidable performance. Even Nvidia's rugged GeForce GTX 480M winds up a whopping 33% behind the Radeon HD 6970M.

Hmm. I was just stumbling out of here into the iMac forum but... this seems to contradict some of the things being said over there.

I can't find the link right now, but a guy who seems to know what he's talking about asserts that the 6970m is very similar in performance to the desktop 5850 - which, obviously, is lesser than the desktop 5870.


EDIT: Ah, I see there's been some clarification of this, since.
 
The 6970m is a better card, it takes the lead by 2% according to game-debate.com, since it is among the latest cards, it proves to be a better one.

Stop with the misinformation. The 6790m is based on the 6850 architecture. It is never, ever going to beat a 5870 or a 5770 for that matter. It is not one of the latest cards it is a repackaged mobile version of existing middle of the road tech.
 
We will need to see some in game figures, I think theres something to be said as regards monster resolutions like 2560x1440p and 1600p. These resolutions need more than 1gb vram for sure! I've been asking around for performance specs for the 5870 in the latest games to compare to the 6970m at 1440p. One thing about mac displays is gaming does not look good at 1080p res.
 
We will need to see some in game figures, I think theres something to be said as regards monster resolutions like 2560x1440p and 1600p. These resolutions need more than 1gb vram for sure! I've been asking around for performance specs for the 5870 in the latest games to compare to the 6970m at 1440p. One thing about mac displays is gaming does not look good at 1080p res.

Just go to Toms Hardware and look at 6850 (6790m) vs. 5870 @ 2560x1600 and you'll see they both are equally insufficient. The core on the 5870 can move the memory in and out much faster than the 6970m. Slow card with tons of memory or a faster card with less memory tends to net similar results. There are exceptions, of course. I'd personally rather have a faster core for shaders vs. memory for more AA.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.