2011 iMac Geekbench Scores Sub-Par

[Justin Debakey]

My 2011 MBP 2.3ghz gets a Geekbench score of 10526 (32-bit), while Everymac.com shows a score of 11602 (32-bit) for the 2011 iMac 3.4ghz.

Both are Core-i7, both have 4GB of 1033mhz DDR3, both have 8mb cache. I expected a lot more from the new iMac 3.4 core-i7.

Furthermore, today there is finally a posting of a 32-bit Geekbench score for a 2011 iMac 2.8ghz Core-i7. The score is only 9746, even though the base/overclock speed of the CPU is 2.8/3.8.

The facts that 1) the 3.4 iMac score is not more significantly higher than the score of the MBP with a base/overclock CPU speed of 2.3/3.4 and that 2) the 2.8 iMac Core-i7 score is actually a lot less than the MBP seems really strange to me.

Here is a small table of these scores, all 32-bit:

10526 Core-i7 base speed=2.3 turbo speed=3.4ghz cache=8mb RAM=4gbDDR3 (MBP)
11602 Core-i7 base speed=3.4 turbo speed=3.8ghz cache=8mb RAM=4gbDDR3 (iMac)
9746 Core-i7 base speed=2.8 turbo speed=3.8ghz cache=8mb RAM=4gbDDR3 (iMac)

Being that Geekbench measures only CPU and memory throughput, these comparisons should be 'apples-to-apples' ... right?

Can anyone shed some light on this please?


_________________________________________

Mac Pro: 2008 3ghz 8-core

MacBook Pro: 2011 i7 2.3 4-core wSSD (dual HD)

MacBook Pro: 2008 C2D 2.4 wSSD (dual HD)

iMac: 2010 i3 3ghz

iPad 2

iPhone 4

 

[roland.g]

The new iMac comes in an i7 with HT at 3.4/3.8

It also comes in an i5 variant at 3.1/3.x?
And an i5 variant at 2.7/3.7.

What is the 2011 i7 2.8 you are talking about?

 

[Chris5488]

Scores are normal.

The Macbook Pro you compare it with is this years Macbook Pro, while the 9746 score of the 2.8Ghz i7 iMac were from the previous iMac models (2010), which was an amazing score! No-one complained about it, especially since the cheapest Mac Pro from 2010 (which was still more expensive than the i7 iMac, and because of being an Mac Pro still needed an external display) only got a score of 9723 , the same as the cheaper iMac!

You see my point? You can always complain about those scores being not high enough or complain about the fact one Mac has a better price/performance ratio than the other. You can't blame Apple or Intel because their i7's aren't fast enough or their notebook i7's are too fast

Besides, the 2.3Ghz i7 Macbook Pro (scores 10526) costs 2,749 dollars, the 3.4Ghz i7 iMac (scores 11602) costs 2,199 dollars...

 

[kevin2223]

I see many of the 2.3GHz MacBook Pros maxing out (64-bit) in the 11,200 to 11,500 range, while the iMacs max out around 12,700-12,900. Both are very close in performance, the iMac comes out just 10-15% faster in most tests, whether through Passmark (mostly Windows) or Geekbench.

The 2.8GHz i7 2600S on the build-to-order 21.5" iMac is shown to be only 2% faster than the 2.3GHz in the MacBook Pro. Both Geekbench and Passmark for the 2600S processor need to be looked at carefully, as there is only one sample for each.

We will likely see the iMacs pull ahead again when 6 or 8-cores (at or below 95 watts) are introduced on the high-end. This will probably happen within the next two Intel release cycles.

I'd also think that when you are doing processor-intensive tasks. Cinebench shows this a little more than Geekbench, with a 20% difference (5.44 on 2.3GHz MacBook Pro and 6.8 on 3.4GHz iMac). Heat will become a problem in the MacBook Pro before the iMac, so the processor will likely down clock to reduce heat.

What we are seeing could also indicate that Geekbench puts more emphasis on the Turbo Frequency than raw GHz. All-in-all, I think we were expecting too much from the i7-2600 processor. It is certainly 20-25% higher in performance than last year's 2.93GHz i7-870, although synthetic benchmarks don't always show this.

 

[theSeb]

^ those scores show that there is something wrong with the benchmark and not with the sandy bridge CPUs. They go against all real application benchmarks from toms hardware and anandtech.

 

[CoryBoyUSA]

What is the 2011 i7 2.8 you are talking about?

I'm pretty sure he's talking about the upgrade on the high-end 21.5 model, which can be upgraded from a 2.7 i5 to a 2.8 i7.

 

[Kendo]

$2500 for a portable 17" MacBook Pro.

$2000 for a 27" iMac with a gaming GPU.

I don't get what is the issue? Even if they have similar scores, one is 20% cheaper and you get a giant 27" LED screen and the 2nd most powerful mobile GPU on the market.

The other is classified as a prosumer laptop where it is expected to be immensely powerful but is much more expensive due to the portability.

Are we expecting the iMac to not only have a larger screen and a better GPU at a much lower price, but also have twice the benchmarks of the Macbook Pro (which was created to be a portable beast in the first place)?

 

[kevin2223]

Geekbench (32-bit scores they have posted on their blog and on the benchmark chart) show a 10.5% performance gain between the high-end MacBook Pro and high-end iMac, Passmark shows a 15% difference, and in AnandTech's 3D rendering tests, the iMac *using i7-2600K* scores 20-27% better. In the Photoshop test done by AnandTech, the 2600K scored nearly 30% higher than the MacBook Pro processor (11.3 vs. 16.1 seconds).

The high-end iMac's processor will perform better in multi-threaded tasks than the MacBook Pro, and this could be a reason why Geekbench/Passmark scores are only showing a 10-15% difference between models that should be showing higher scores.

Real-world performance will be higher than these synthetic benchmarks show.

 

[Justin Debakey]

Thanks everyone!

Thank you, all for your input on this. I sincerely appreciate it.

I like Geekbench for two reasons: 1) It tests only processor and memory performance, and 2) Geek scores translate consistently into Logic Pro and Pro Tools virtual instrument tracks, my main use. I use top-drawer SSDs on everything, and don't really care about things like Photoshop, Word, how fast documents open up, and gaming issues.

The reason I started this thread is that I sold my Core-i7 iMac on eBay (after seeing the Logic Pro (and associated Geek scores!) performance of my MacBook Pro 2.3 Quad Core-i7, awaiting the new SB Core-i7 models. I was expecting significantly higher scores on the new i7 models, (like 25-30% or more), but have not seen that. The 18% increase in the top 27" model (over the previous 2.93ghz model: 12877 vs 10830), is not bad, and is probably where I'll go with my purchase.

The model I'm still disappointed in is in the 21.5" Core-i7, because this gives one the opportunity of getting a quad core-i7 Macintosh for the price point of $1699 retail. We're finally seeing a few more Geekbench scores posted for the 21.5" Core-i7. I still find it odd that every score is lower (not 10-15% higher, but lower) than my 2.3ghz Macbook Pro even though the clock speed, both base and overclock, is higher on the iMac.

Again, thank you, everyone for your comments.

Justin
_________________________________________

Mac Pro: 2008 3ghz 8-core

MacBook Pro: 2011 i7 2.3 4-core wSSD (dual HD)

MacBook Pro: 2008 C2D 2.4 wSSD (dual HD)

iMac: 2010 i3 3ghz

iPad 2

iPhone 4

 

[iSayuSay]

Benchmark tools usually used by people to brag about their own computer, rather than using it for a real test.

Each computer only need benchmark test once, and it has to be in idle to make sure benchmark using almost 100% of your computer power.

But oftenly, we use benchmark to compete with other computer.

"Whoa my macbook pro i7 scores 11000 while your precious iMac scores 12000 .. boooo .. you paid too much"

Or "My $1000 custom built PC scores 16000 while your $2000 iMac not even reach 10000 .. you've been fooled by dumbass"

And there will be blood shed after ... so yeah, let's back to basic. What do you use with your computer? For sure, you didn't buy it only for running geekbench, right?