PDA

View Full Version : Apple G5 vs PC: Cinebench results


Onno
08-25-2003, 07:21 PM
Cinebench test of the G5 versus PC's.

Striking result...

http://www.barefeats.com/g5sum02.html

PS: Keep in mind though:
BIG CAVEAT: Cinebench 2003 is NOT optimized for the G5... yet. Maxon is looking seriously into G5 optimizations using the new Xcode compiler. They predict the G5/2GHz MP (after optimizations) will equal or surpass the Dual Xeon 2.4GHz.

Byla
08-26-2003, 05:25 AM
Very very lame test I say. Someone had 5 minutes to spare and that is the result. Lame as..:thumbsdow

Still waiting for a GOOD test. Like PDF handling benchmark, optimized Cinebench benchmark, audio handling, photoshop filters benchmark, Shake, After Effects, Maya rendering... Just loading a benchmark test (and unoptimized!!!) is not a true test. Just a proof how lame the author is.

Joviex
08-26-2003, 06:19 AM
While I would agree that more specificly tuned benchmarks are in order, there is some logic to this test.

All the hype around this new box might obviously fall flat in the area of older applications. Thus, why would joe nobody with a tons of already owned software want a machine that is slower then what he currently has.

My point being, this is a new machine which in all forms should be at least slightly faster then a G4 when it comes to older software.

Sad to see it is not.

Byla
08-26-2003, 06:36 AM
how do you mean its not faster then g4? Only 1.6 g5 single has some problems, the rest of the g5 family are doing fine against g4. But comparing them to PC is another story, not a very pleasant for Apple right now:surprised .

But I agree, why would anyone right now buy a new g5? The only reason I see is that investment in G5 is kinna investment into future. But, for non americans, an extremly expensive investment, I have to add. With a present software, there is just no point of buying g5 Iam afraid. Still, lets wait a couple of weeks for software updates. Better, lets wait for better tests.

Joviex
08-26-2003, 06:41 AM
Originally posted by Byla
how do you mean its not faster then g4? Only 1.6 g5 single has some problems, the rest of the g5 family are doing fine against g4. But comparing them to PC is another story, not a very pleasant for Apple right now:surprised .


Well, there is something else I would like to point out about that test.

While CB hasn't been tuned for a G5 yet, neither has it been tuned for an Opteron 64. So horses for courses I suppose.

MacRonin
08-26-2003, 03:51 PM
And, if you actually look at the specs for the 1.6GHz G5, you will see that it is slower in many areas than the single 1.8GHz G5 or the dual 2.0GHz G5 boxes...

Slower RAM, slower FSB, slower PCI slots...

All in all, NOT the box to buy...

But with OS X, why would anyone (especially DCC folks) want to get anything BUT a minimum of dual CPUs...?!?

chadtheartist
08-26-2003, 03:54 PM
Ummm.... after looking at the graph at Barefeats, I don't see any mention of the Opteron in those test results. So can you explain how the Opteron not having an optimized Cinebench program even relates to the G5 tests?

One thing to note, also, is the G4 units they tested were all Dual Processor Machines. And an interesting "FEAT" was the DP G5 performed rather well without any sort of optimizations.

So from what I gather, a Single G5 is faster, or almost as fast as Apple's previous DP G4 units. That sounds pretty good to me. Although from this test it seems that the G5 still has a ways to go from matching, or exceeding the Intel and AMD chipsets.

richcz3
08-26-2003, 08:20 PM
There is something inherently wrong with all the Benchmarking going on.

Apple marketing peeps fired the 1st salvo in their media blitz. So turn around is fair play. That being written, two wrongs don't male a right.

When real practical benchmarks in real world apps show up, then and only then will we get a good reasonable comparison.
The little "astericks" in most benchmark schemes basically invalidate tests for any but the Market and PR guys who rig them.
That applies to all companies, AMD, Apple, Intel, nVidia, ATI and the whole lot.

Give me MAX, Lightwave, Maya, Photoshop, AE, render times and then there will be a more level playing field. Of course even then people will decry these as invalid because some of that software is optimized for certain hardware. :surprised

Oh well. So don't bet the farm on applications you don't use everyday.


Rich

Marc Andreoli
08-26-2003, 10:01 PM
Originally posted by richcz3
Give me MAX, Lightwave, Maya, Photoshop, AE, render times and then there will be a more level playing field.


just give me MAX on OSX and I will be a happy camper...:)

Limbus
08-26-2003, 11:14 PM
Originally posted by Byla
Very very lame test I say. Someone had 5 minutes to spare and that is the result. Lame as..:thumbsdow
Why are they lame? Because you dont like the result?
Right now Cinema 4D (and most other apps) are not optimized for the G5 (BTW Cinema 4D isnt the most P4 SSE2 optimzed app either) and it can take a while till they are. And even with the optimization it will barely be as fast as a dual Xeon 2.4 Ghz. And this isnt even the fastest Xeon out there.


Still waiting for a GOOD test. Like PDF handling benchmark, optimized Cinebench benchmark, audio handling, photoshop filters benchmark, Shake, After Effects, Maya rendering... Just loading a benchmark test (and unoptimized!!!) is not a true test. Just a proof how lame the author is.

He clearly stated that Cinebench isnt optimized for G5 and that he will test many other apps and even a opteron dual rig. What is lame about it?

Florian

chadtheartist
08-26-2003, 11:21 PM
I guess the reason why he said the test was lame was because there are other benchmarks, like Photoshop benchmarks, that might be of more interest. Especially since Adobe has already released an update for the G5. IMO, a Photoshop test would have been the first thing I would have done. But still, who cares! None of us are getting one anyway! At least for now. :D

Marc Andreoli
08-27-2003, 05:33 AM
some photoshop benchmarks on the (single) 1.6 G5... (http://www.chaosmint.com/benchmarks/powermac-g5-ps7bench/)

It is running on a 'G5 enabled' version of Photoshop, so I guess it is not fully optimized yet (but who knows if/when it is going to be). A test with the dual 2Ghz machine might be interesting...and of course the usual Opteron comparison to fuel the flame wars ;)

CGTalk Moderation
01-16-2006, 12:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.