More Intel iMac Benchmarks

[MacRumors]

http://www.macrumors.com/images/macrumorsthreadlogo.gif (http://www.macrumors.com)

MacSpeedZone posts (http://www.macspeedzone.com/html/hardware/machine/performance_in_the_raw/06/1_23.shtml) their benchmarks for the Dual Core Intel iMac and claims that Macworld's initial test were misleading:

This is where the Macworld "First Lab Tests" article falls a little flat ... obscuring the processor capacity vs processor usage problem inherent with mutiprocessor machines (or multi-core ... same difference). Using Macworld's logic we could argue, given the data above, the Quad G5 Power Mac is only 14% faster when running some of Apple's own applications. We think that this is misleading, as we pointed out.

They post a comparison chart, taking into account percentage of processor usage as a guide.

[nataku]

hmmm... interesting... very interesting indeed. i dont care what the numbers tell me now... because I want one!

[Josh396]

Well I have my 20 in Duo waiting for me at home with 256 VRAM and 2 GB of RAM and I'll be picking it up tomorrow so I'll see first hand how it stacks up to my Rev B iMac. One thing I found somewhat interesting was the game performance. The Duo had 2 more FPS, not really impressive, unless the game was running under Rosetta. If that's the case there may be a huge jump once the game is Universal. I can't wait to see how WoW will run on my new iMac once it's Universal. If the test done earlier are true then I may have a hard time turning WoW off (not that I don't already:eek: ;) ).

[nagromme]

The G5 is still a great CPU (while we await Intel's next generation chips this summer/fall) but it's important to be intelligent about comparing speeds. The article makes a good point about that, and about possible software bottlenecks that are making Intel Macs test slower than they might. (And no, they're not REALLY saying a Core Duo rivals a quad G5.)


"old applications, that do not run natively on the Intel processor, run about half as fast as they would run on a G5 machine"

Sounds like an older-that-a-year-ago G4 user like me will have few gripes during the (temporary) Rosetta phase :)


"There are precious few applications that take complete advantage of multiple processors"

True--but ALL apps take "INcomplete" advantage: if nothing else, OS X will give the current foreground app a CPU to itself, moving the OS, other apps, and background processes to the other CPU(s). And many single-CPU apps call on OS X features that themselves ARE multi-CPU aware.

But I like seeing consumer machines go dual core: it means app developers (and Apple) have incentive to do even more with multiple CPUs, for even more kinds of app.

[nagromme]

One thing I found somewhat interesting was the game performance. The Duo had 2 more FPS, not really impressive, unless the game was running under Rosetta.

InsideMacGames and TheMacObserver have had some game tests lately, and other reviews have mentioned games too. Results vary, but it sounds like across the board--from older games right up to Doom 3--Rosetta delivers 75%-85% of G5 speed. That ain't bad, and it means older games can run smoothly without ever needing to be ported.

It also shows the better GPU found in the new iMacs. Given that, a straight comparison can't be made to the G5 iMacs. (But comparisons can be made to Rosetta games vs. updated Universal versions.)

[adamfilip]

One thing i never understood if im doing something processor intensive

i might only see it at 80% usage or something but it takes a long time. why doesnt it use 100% and get it done faster

[plinden]

Well I have my 20 in Duo waiting for me at home with 256 VRAM and 2 GB of RAM and I'll be picking it up tomorrow so I'll see first hand how it stacks up to my Rev B iMac. One thing I found somewhat interesting was the game performance. The Duo had 2 more FPS, not really impressive, unless the game was running under Rosetta.
From the website:
Scores in dark orange are from applications that run natively on both Intel Macs and G5 Macs
The game benchmark is in white, so I guess that means it was run in Rosetta.

What's striking to me is that all but two of the native apps are at least twice as fast on the Intel and a couple are nearly 3x faster. One is close to 2x (90%) faster and one is 67% faster. So perhaps Stevie-boy didn't lie after all.

[Lazyhound]

One thing i never understood if im doing something processor intensive

i might only see it at 80% usage or something but it takes a long time. why doesnt it use 100% and get it done faster
Could be bottlenecked by the hard drive or RAM.

[mvc]

Could be bottlenecked by the hard drive or RAM.

Or the user...
:p

[simie]

When running the QuickTime encodes the Power Mac Quad G5/2.5GHz took 86.25 seconds.The Intel iMac Core Duo 2.0GHz took 176.60 seconds

The article is rubbish - they start out comparing the intel iMac Core Duo 2.0GHz against the Power Mac Quad G5/2.5GHz. Now I thought that this was interesting, until you get down to the actual comparisons at the bottom and they are then comparing the timings between the iMac G5/1.8GHz and the iMac Intel Core Duo 2.0GHz.

What the hell happened to the Power Mac Quad G5/2.5GHz that the article started out with. It should have been this that they used in the comparison chart at the bottom of the article and not the iMac G5/1.8GHz.

[jope76]

The article is rubbish - they start out comparing the intel iMac Core Duo 2.0GHz against the Power Mac Quad G5/2.5GHz. Now I thought that this was interesting, until you get down to the actual comparisons at the bottom and they are then comparing the timings between the iMac G5/1.8GHz and the iMac Intel Core Duo 2.0GHz.

What the hell happened to the Power Mac Quad G5/2.5GHz that the article started out with. It should have been this that they used in the comparison chart at the bottom of the article and not the iMac G5/1.8GHz.

Why? They are comparing two different versions of the same computer that sells for the same price. It tells us what the switch to intel did for the iMac.
The Quad G5 was just to prove their point a little more.

[simie]

Why? They are comparing two different versions of the same computer that sells for the same price. It tells us what the switch to intel did for the iMac.
The Quad G5 was just to prove their point a little more.

I know what you are saying and I agree with you but the article states at the top.

We are pleased to report that our testing results show that the new Dual Core Intel iMac, which clocks in at 2X 2.0GHz is almost as fast as the current high-end Power Mac that has two Dual Core G5 processors running at 2.5GHz.


Surly the comparison should have been against iMac G5 - Intel iMac and the Quad. We would then not only see how much faster the Intel iMac is over the G5 iMac but also how fast it is against the G5 quad.

[BenRoethig]

Making a fair comparison isn't going to be possible until rosetta is mostly out of the picture.

[Peace]

To even BE comparing an Intel iMac to a Dual G5 is amazing in itself..

But the article confirms my own benchmarking I've been doing between my Intel iMac and my Dual G5..
Sorry folks,when all apps have been ported to Intel taking Rosetta out of the picture,the new Intels indeed will stand up on their own right..

[nospleen]

To even BE comparing an Intel iMac to a Dual G5 is amazing in itself..

But the article confirms my own benchmarking I've been doing between my Intel iMac and my Dual G5..
Sorry folks,when all apps have been ported to Intel taking Rosetta out of the picture,the new Intels indeed will stand up on their own right..

Have you posted your benchmarks somewhere? If not, please do!

[plinden]

The article is rubbish - they start out comparing the intel iMac Core Duo 2.0GHz against the Power Mac Quad G5/2.5GHz. Now I thought that this was interesting, until you get down to the actual comparisons at the bottom and they are then comparing the timings between the iMac G5/1.8GHz and the iMac Intel Core Duo 2.0GHz.

What the hell happened to the Power Mac Quad G5/2.5GHz that the article started out with. It should have been this that they used in the comparison chart at the bottom of the article and not the iMac G5/1.8GHz.
If you think about it, you'll realise they were doing that to point out the problems in the macworld benchmarks - using the same methodology, they could show that the Quad was only 14% faster than the Intel iMac, which was patently wrong.

See, it's like this - the Quad has 2x the cores in the Intel iMac, the Intel iMac has 2x the cores in the G5 iMac. Doing something that taxes only one of the Intel cores and comparing it to the G5 is the same as doing something that taxes only two of the cores on the Quad and comparing it to a dual core machine.

Then they go on to do the real comparisons, with the previous version of the iMac. Everyone is comparing the new iMacs with the G5, so doing so is perfectly fine.

[Peace]

Have you posted your benchmarks somewhere? If not, please do!


Keeping with my scatter-brained way of doing things my benchmarks have been posted all over the place :)

But if I can be shown a topic where benchmarks ONLY are posted I would be happy to do so..

[plinden]

Keeping with my scatter-brained way of doing things my benchmarks have been posted all over the place :)

But if I can be shown a topic where benchmarks ONLY are posted I would be happy to do so..
I've started this thread http://forums.macrumors.com/showthread.php?p=2094210

[Diatribe]

One thing i never understood if im doing something processor intensive

i might only see it at 80% usage or something but it takes a long time. why doesnt it use 100% and get it done faster

It's mostly the software that doesn't use it completely. Once the software is optimized we should see those iMacs surpass the 2.5GHz Powermacs. And when Merom comes (at that time the software will most likely be optimized) those apps should be at least 150% as fast as the current G5 Powermacs. The Conroe and Woodcrest cpus will probably beat the G5 by 200-250% in speed.
Now who was saying again that Apple should've stayed with IBM? :rolleyes:

[Peace]

D'OH! I just noticed the title of this topic!!
I work a lot a just had my first cup of java..:D

Here's a couple benchmarks.Keep in mind my way of benchmarking is tuned more toward the average user experience..

Intel iMac Dual-Core 1.83 w/2gigs ram X1600 video w/128megs..HEY! it was free from the ADC :-) vs my rev.B Dual 2.3 G5 Powermac w/4 gigs ram Radeon 9650 video card.

I copied a 3.03 gig zip of .WAV files.180 minutes,14 songs ,from a DVD to the desktop..

Approx 8-10 minutes for each machine.

Unzip the file..
Approx 5 minutes for the G5 and 4 minutes for the iMac.

I set iTunes to encode AAC 128kbps as the default.

Drag and Drop the songs to iTunes library

Approx 3 minutes for each machine.
Convert to AAC 128kbps.
Approx 3 minutes @ 30X for the Dual G5
Approx 5 minutes @20.5X for the Intel iMac.

Thats one..

iMovie :
Connected a Canon optura 60 and recorded 2 minutes worth of wide-screen native .dv

Export to .dv :
G5 took about 3 minutes as did the Intel iMac
here's the interesting part
Export to iPod
G5 took about 5 minutes
Intel iMac took about 3 1/2..

More later since I'm moving various systems around the house..

[bugfaceuk]

It's hard enough to compare to processors with each other, but two incompatible processors coupled with different numbers of available cores? I think it resolves down into a perception argument.

[dwsolberg]

Geez, you'd think benchmarking was rocket science because no one seems to know what they're doing. I expected MacWorld to do a good job, and they seem be be technically accurate, but MacSpeedZone makes some excellent points that you'd expect a good magazine staff to know.

The iMac Duo has two processors, and at least some tests need to take this into account. For example, I currently have iTunes, Safari, Quark 6.5, Entourage, InDesign, Photoshop, iBiz, iCal, Address Book, Word, Excel, Preview and Firefox open. While I'm waiting for Quark to output a postscript and distiller to make a PDF, I often switch to Entourage to check my email accounts, and then switch over to Photoshop to check a file. There are no benchmarks for what I expect is pretty typical usage, but I know from experience that dual processors (or cores) make a big difference in speed and responsiveness.

[Fabio_gsilva]

MacSpeedZone made a very good work indeed.

I thought it was very informative.

See you.

[gmanrique]

:rolleyes: The article is rubbish - they start out comparing the intel iMac Core Duo 2.0GHz against the Power Mac Quad G5/2.5GHz. Now I thought that this was interesting, until you get down to the actual comparisons at the bottom and they are then comparing the timings between the iMac G5/1.8GHz and the iMac Intel Core Duo 2.0GHz.

What the hell happened to the Power Mac Quad G5/2.5GHz that the article started out with. It should have been this that they used in the comparison chart at the bottom of the article and not the iMac G5/1.8GHz.

[gmanrique]

;) I know what you are saying and I agree with you but the article states at the top.




Surly the comparison should have been against iMac G5 - Intel iMac and the Quad. We would then not only see how much faster the Intel iMac is over the G5 iMac but also how fast it is against the G5 quad.

[gmanrique]

So far, the Macworld test was something like comparing a BMW and a Cavalier (No offene intended to Cavalier owners) and concluding that even the BMW has double the horsepower, performance is just a little better, by testing them in the city respecting the speed limits :eek:

Geez, you'd think benchmarking was rocket science because no one seems to know what they're doing. I expected MacWorld to do a good job, and they seem be be technically accurate, but MacSpeedZone makes some excellent points that you'd expect a good magazine staff to know.

The iMac Duo has two processors, and at least some tests need to take this into account. For example, I currently have iTunes, Safari, Quark 6.5, Entourage, InDesign, Photoshop, iBiz, iCal, Address Book, Word, Excel, Preview and Firefox open. While I'm waiting for Quark to output a postscript and distiller to make a PDF, I often switch to Entourage to check my email accounts, and then switch over to Photoshop to check a file. There are no benchmarks for what I expect is pretty typical usage, but I know from experience that dual processors (or cores) make a big difference in speed and responsiveness.

[Nuc]

I went and checked out the 20" intel iMac. It had 1Gb ram and the usual specs. I was surprised that it opened up word and excel so fast. Probably about 3-5 seconds faster than my PB G4 (1.5Gb ram). I couldn't test any other apps that ran on rosetta because they didn't have any installed. I was going to test out photoshop but it wasn't on the machine. Hmmm I wonder if they did that on purpose... :rolleyes:

Nuc

[MacinDoc]

But I like seeing consumer machines go dual core: it means app developers (and Apple) have incentive to do even more with multiple CPUs, for even more kinds of app.
...which is one of the reasons I'm expecting all new Apple computers to be multi-core, to encourage all developers to take advantage of the processing power that would otherwise be underutilized on the pro grade Macs. There has to be a critical mass of multi-core machines to make it worth developers' efforts.

[Homy]

Check out these numbers (http://www.macworld.com/2006/01/reviews/intelimacs/index.php)! 36 fps in Doom 3 v1.3.1303: Universal Beta Demo 1. Why is it impressive? Cause they tested with Ultrahigh graphics and shadows :eek:. That's with only 512 MB RAM.

"We ran the Demo 1 test on a beta Universal version of Doom 3, which was set to use Ultrahigh graphics at a resolution of 1,024 by 768; all advanced options were set to Yes except for vertical sync and antialiasing."

iMac G5 2.1 GHz gets 17 fps.

[bugfaceuk]

Check out these numbers (http://www.macworld.com/2006/01/reviews/intelimacs/index.php)! 36 fps in Doom 3 v1.3.1303: Universal Beta Demo 1. Why is it impressive? Cause they tested with Ultrahigh graphics and shadows :eek:. That's with only 512 MB RAM.

"We ran the Demo 1 test on a beta Universal version of Doom 3, which was set to use Ultrahigh graphics at a resolution of 1,024 by 768; all advanced options were set to Yes except for vertical sync and antialiasing."

iMac G5 2.1 GHz gets 17 fps.

Do you know if this includes the changes to the Doom 3 engine that iD have made with Intel? I was reading the latest PC Format in the UK, and apparently Intel and iD spent some time making the graphics pipeline more "Dual Core friendly" and gained a 70% performance boost which the magazine was able to see as a 30%-60% boost depending on the particular machine. I'm curious to know if the Mac version of the engine has this extra booster strapped on?

[bugfaceuk]

...which is one of the reasons I'm expecting all new Apple computers to be multi-core, to encourage all developers to take advantage of the processing power that would otherwise be underutilized on the pro grade Macs. There has to be a critical mass of multi-core machines to make it worth developers' efforts.

I must admit I agree with this, like, 160%. I'm more than a little concerned (and I know all the product differentiation arguments, and they are good ones too) that there will be some single core machines (iBook and MacMini). Although again, there's the whole cost argument. There would seem to be a real competitive advantage to be had in the Mac platform being entirely dual-core. It would allow Mac developers to say "well I know I have at least 2" and start coding for that.

Now all we need to do is for intel to chop the cost of the dual core's, and to release the faster dual-core's early, and Apple will be able to implement it.

I guess the thing that disappointed me the most was that the MacBook Pro came out at this stage, I would have prefered to see that processor in an iBook, and save the faster ones due out later this year for the Pro 'books.

[BenRoethig]

Do you know if this includes the changes to the Doom 3 engine that iD have made with Intel? I was reading the latest PC Format in the UK, and apparently Intel and iD spent some time making the graphics pipeline more "Dual Core friendly" and gained a 70% performance boost which the magazine was able to see as a 30%-60% boost depending on the particular machine. I'm curious to know if the Mac version of the engine has this extra booster strapped on?

Those changes were made for Quake 4 and were added because of the Xbox360 version than anything else. Doom 3 has no plans for an MP patch.

[bugfaceuk]

Those changes were made for Quake 4 and were added because of the Xbox360 version than anything else. Doom 3 has no plans for an MP patch.

Doh, my mistake I mistook the screen shots obviously. Not that iD games have all begun to look the same or anything...

[Homy]

More results from Barefeats (http://www.barefeats.com/imcd.html)

[wasimyaqoob]

My advice is that grab a PowerBook G4 and a PowerMac G5 while you can :D