Geekbench-results of the new iMacs (Mid 2010) compared to the old ones (Late 2009)

[blinkin182]

Just ran Geekbench on my 2009 i7 iMac, and also the OpenGL tests with the original ATI drivers for my 2009 iMac...

How do you get the 2010 iMac display drivers on the 2009 iMac?

How come the OpenGL 2.1 tests have such low results and the OpenGL 3.0 fail?

Otherwise, impressive changes! I've ordered my i7 Refurb, so will hopefully be posting my results soon

 

[Flea24]

My beauty arrived today (finally, after much ado with UPS, it got left behind in the depot and when I called chasing it, they discovered their error and sent a van especially, a 90 mile round trip!).

The screen is looking fine and there are virtually no sounds from it at all, it's almost like I've gone deaf since unplugging my aged G5

I've transferred almost all my files from the old girl and run my first Geekbench test .... results below:

Two words: well pleased!

not far behind you on that one!. i7 27" but only 4GB of ram and standard HDD: 10701

 

[Tigerman82]

I just wanted to let you guys know that someone has finally tried the 64-bit Geekbench with iMac 21.5" i5 and the score was 7801 (which is about 400 points better than previous-gen quad core i5 and only 200 points worse than current-gen quad core i5). Impressive!

My first iMac just arrived; 21.5" i5, 1TB Seagate & 512MB 5670HD. The screen is the perfect size, fits perfectly on my desk.

Whisper quiet, I'd say on a pure white screen 90% white with only minor discoloring near the bottom. This has no effect on me as I don't stare at a white screen. Normal usage, it's looking beautiful. So overall, very happy with the purchase.

Quick Geekbench scores after turning it on are in the expected range.
32 bit - 7021
64 bit - 7801

I went and did the software updates and there were a few. iTunes, java, Safari 5.0.1, something else like 6mb in size. There was no 10.6.4 Update (mid 2010) which I think I've seen someone else have a picture of.

Do you think it was already done? How can I tell what patch level I am running at?

 

[WilliamG]

I should mention by the way to anyone running geekbench that the improved updated ATI drivers aren't going to change your geekbench number - your graphics don't come into play there - in fact the only things that do seemingly are memory (how fast is it?) and processor (how good was the fab that day?).

How to install the new ATI drivers?

Here we go again:

1. Download the Mid 2010 iMac update.
2. Download Pacifist.
3. Locate System/Library/Extensions in the iMac update.
4. Highlight all ATI kexts.
5. Select "install" in the masthead of Pacifist.
6. Reboot.
7. You're done.
8. If you bricked your machine it's not my fault.

Thank you for that. Just installed the newer drivers on my 2009 i7 iMac and here are my results...

I really do think the 2010 iMac is a big disappointment after almost a year...

 

[Queso]

Mine has now arrived and going through the Transferring Information... stage of Setup. Hopefully won't be too much longer until I can run a test or two

 

[tyche]

Let the new baby sit for a few hours to make sure things have settled down and run my final tests again and uploaded the results. No tweaking or optimizing, if that's even possible. This is a 4GB machine.

iMac i5 3.6GHz

32bit - 7023
64bit - 7804

Results

 

[Felias]

Let the new baby sit for a few hours to make sure things have settled down and run my final tests again and uploaded the results. No tweaking or optimizing, if that's even possible. This is a 4GB machine.

iMac i5 3.6GHz

32bit - 7023
64bit - 7804

Results

Thanks a lot! I'll add the numbers.

 

[LAPTOPODO]

My conclusions at the moment:

1.- the i7 it's a great processor

2.- There is more difference i3-500 vs i5-680 than what the most of us supposed days ago.

3.- The i5 Quadcore it's not so good compared with the i5-680 (even in the 64bits tests where the quadcore should had been much better)

My plans:

Getting i7-860 refurb ... and if (I cross my fingers) I would get a yellow lemon .... sending it back, money refunding and get a 27'' i5-680

 

[madmonkeys]

I just wanted to let you guys know that someone has finally tried the 64-bit Geekbench with iMac 21.5" i5 and the score was 7801 (which is about 400 points better than previous-gen quad core i5 and only 200 points worse than current-gen quad core i5). Impressive!

My late 2009 quad core i5 2.66 750 clocks in at 7664 in 64bit. How is a dual core i5 beating it by so much? I thought Ghz was a myth?

And does this actually mean it the dual core i5 will be faster at completing multi-thread tasks then the 750 i5 in real world tasks?

How much (if any) of a difference would it make switching the ram from 1066 to 1333 as the i5 750 supports that speed according to Intel?

 

[GyroFX]

I thought Geek Bench only tests up to 2 physical cores, that's why you're not seeing better results with your quad core i5 compared to dual core i5 at a higher Ghz?

 

[LAPTOPODO]

I thought Geek Bench only tests up to 2 physical cores, that's why you're not seeing better results with your quad core i5 compared to dual core i5 at a higher Ghz?

Would be interesting to confirm that, for giving GeekBench the proper importance.

 

[iamthedudeman]

Let the new baby sit for a few hours to make sure things have settled down and run my final tests again and uploaded the results. No tweaking or optimizing, if that's even possible. This is a 4GB machine.

iMac i5 3.6GHz

32bit - 7023
64bit - 7804

Results

Wow. Was I wrong. I am surprised. I was one of the early doubters. A 1000 point difference is impressive. It looks like the extra 200.00 is worth it after all.

 

[Hellhammer]

Wow. Was I wrong. I am surprised. I was one of the early doubters. A 1000 point difference is impressive. It looks like the extra 200.00 is worth it after all.

Yeah, it seems to be a nice bump. It costs ~13.3% more than 3.2GHz i3 does (1499$ vs 1699$) but provides 12.5% better multi-threaded performance (3.6GHz vs 3.2GHz) and ~20,6% better single-threaded performance (3.86GHz vs 3.2GHz). In 32bit GB, i5-680 is ~15% faster and in 64-bit it's ~18% faster than i3-550 is so mathematically and theoretically, it's definitely worth the extra $.

 

[GyroFX]

Would be interesting to confirm that, for giving GeekBench the proper importance.

well, geekbench states that it's multicore aware so i guess what I heard isn't true. For the most part i gathered that 3.6 i5 dual will out perform 2.8 i5 quad in general use till more, newer software is able to harness the raw power of more physical cores. I think that's right? please correct me if I'm wrong

 

[kernkraft]

Yeah, it seems to be a nice bump. It costs ~13.3% more than 3.2GHz i3 does (1499$ vs 1699$) but provides 12.5% better multi-threaded performance (3.6GHz vs 3.2GHz) and ~20,6% better single-threaded performance (3.86GHz vs 3.2GHz). In 32bit GB, i5-680 is ~15% faster and in 64-bit it's ~18% faster than i3-550 is so mathematically and theoretically, it's definitely worth the extra $.

In that case, can I just have half the performance for half the price?

I don't think that some performance benchmarks could justify the entry level price. From that point, matching performance and pricing only misleads.

 

[Hellhammer]

In that case, can I just have half the performance for half the price?

What does this have to do with what I said? Apple does not offer something like that

I don't think that some performance benchmarks could justify the entry level price.

Of course not, it's a Mac. You can get a PC with same power for half of the price but I was simply comparing the i3-550 and i5-680 what Apple offers and is the upgrade worth it or not. I never talked about is the iMac worth the price, only about the i5 upgrade.

From that point, matching performance and pricing only misleads.

So you think performance and price shouldn't be compared? I simply provided some mathematical figures to show is the i5 upgrade worth it or not. It's completely normal to compare price and performance to see are you overpaying for minor upgrade and to get the biggest bang for your buck. People like numbers as they are easy to understand. 13% price increase for 10-20% performance increase, that sounds like a decent deal.

I don't really get what you mean....

 

[corvus32]

I think the i5 680 is a good upgrade for the 21.5" imac, but for the 27" imac it's not.

The 27" imac starts at $1699. Add $200 for the i5 680, and you're at $1899. Add $150 more the 5750 GPU, and you're looking at $2049. At those prices, your into i5 760 and i7 870 territory both of which already come with 5750 graphics.

 

[kernkraft]

"The cynic knows the price of everything and the value of nothing"

What does this have to do with what I said? Apple does not offer something like that



Of course not, it's a Mac. You can get a PC with same power for half of the price but I was simply comparing the i3-550 and i5-680 what Apple offers and is the upgrade worth it or not. I never talked about is the iMac worth the price, only about the i5 upgrade.



So you think performance and price shouldn't be compared? I simply provided some mathematical figures to show is the i5 upgrade worth it or not. It's completely normal to compare price and performance to see are you overpaying for minor upgrade and to get the biggest bang for your buck. People like numbers as they are easy to understand. 13% price increase for 10-20% performance increase, that sounds like a decent deal.

I don't really get what you mean....

Don't take it seriously, my point was purely about having pretty much the same functionality and the same physical attributes apart from a marginal price difference. Personally, I think the best strategy with iMacs is to buy the cheapest possible one with the feature that you require. If you need the smaller screen, get the basic one and if you want the larger screen, get that. 10 or 20% price increase for similar or even slightly better speed increase in my opinion is not a good deal at all.

I wouldn't go as far as using mathematics, but to me it seems that in the overwhelming majority of the cases, what most people use is not even close to the capabilities of the entry-level models. By the time that your hardware meets its match in software and workload (usually not in a few years); the higher value of the stronger models almost completely disappear.

Personally, for an extra 12-15% in price, I would expect much better performance or some really useful extra features. Otherwise I wouldn't even hesitate to go for the cheapest. It's the same with the RAM upgrades. As long as the memory can be upgraded, I tend to stick within reasonable range. Anything above 4GB and any SSD at this stage would be an overkill for the overwhelming majority. Which is what the iMac is aimed at. Graphic designers, web developers, software engineers and geeks should get something else.

 

[Hellhammer]

*snip*

Of course if you don't need speed, any Intel Mac is fine for you but there are people who can really benefit from that extra 10-20% in performance. If you use the machine for work, 15% shorter waiting time means more work can be done thus more money can be made. Graphic designers, web developers, software engineers and geeks do use iMacs too, it's not just for people who read their email and use Safari. No upgrade is worth it unless you can take advantage of the extra speed it offers

For you and for an average Joe, the i5 won't make a difference but that doesn't mean that it won't for anyone. For majority of people it isn't worth it but as this thread and discussion is based on how much faster it is, I didn't think it's needed to mention that it's not worth it unless you can take advantage of it.

 

[xraydoc]

I think the i5 680 is a good upgrade for the 21.5" imac, but for the 27" imac it's not.

The 27" imac starts at $1699. Add $200 for the i5 680, and you're at $1899. Add $150 more the 5750 GPU, and you're looking at $2049. At those prices, your into i5 760 and i7 870 territory both of which already come with 5750 graphics.

I agree in part, but if the user will do nothing that takes advantage of those 4 cores, then he's better off getting two faster cores than 4 slower ones.

Personally, I think for email, web, 3D games and iLife, the average user would be better off with the upgraded 3.6GHz dual-core i5 plus the 5750 upgrade rather than the base 2.8GHz quad i5. But I suppose we have to wait for benchmarks to prove one way or the other.

The main reason the quads aren't available in the 21" iMac is heat management in the smaller case.

 

[Felias]

I thought Geek Bench only tests up to 2 physical cores, that's why you're not seeing better results with your quad core i5 compared to dual core i5 at a higher Ghz?

On their webpage they write:

Multicore

Whether you're running Geekbench on a single-core Pentium, a quad-core Mac Pro, or a sixteen-core Sun server, Geekbench is able to measure the performance of all the processor cores in your system. Every processor benchmark is multi-threaded and multi-core aware to show you the true potential of your computer.

So i guess it should be... although it seems that the difference between 1,2,4,8 cores really does show in the 64bit-test only.

 

[beeinformed]

Hi!

I would appreciate it if there are any members who have not posted their geekbench tests results for their I5 3.6 processor and would kindly do so. Thanks.

 

[Sammy Cat]

After buying my lady the base model i3, I am thinking of getting the dual core i5 for myself, selling her i3 in two years, and then giving her my i5.

It is either that or just buying myself an i3. My P4 3.0ghz is starting to feel very slow.

Somehow my gut instinct told me that the i5 3.6 would fly. It reminds me of my P4 3.0ghz, which at the time was a very good processor. I've been using it the past 6 years, albeit with older software.

 

[PsyD4Me]

top of the line i7 matches my Early 2008 Quad-Core MacPro....
i guess i'll hold on to it for at least another year....

 

[Hellhammer]

top of the line i7 matches my Early 2008 Quad-Core MacPro....
i guess i'll hold on to it for at least another year....

Maybe in specs but in GeekBench for example, the iMac is over 50% faster due newer architecture