[PeterKG]

Here are Macworlds just posted benchmarks. I wish it had included Powerbooks!

Macworld Benchmarks

[wjdennen]

Is this a sentence?

"As you could probably guess, the new G5-equipped iMacs out-performed than their G4-powered predecessors."

[mkb]

barefeats compared them to powerbooks. The PB did very well.

http://barefeats.com/imacg5.html

[Pierre B.]

Now, who will explain this:

UT2004
------

iMac G5 1.8 GHz: 33.6 fps
iMac G4 1.25 GHz: 21.6 fps




It doesn't look like 3 times more to me, as Apple claim.

[Eriamjh]

Why didn't they compare the iMac to a SINGLE 1.8GHz G5 PM?

[PEHowland]

Originally posted by Pierre B.:
Now, who will explain this:

UT2004
------

iMac G5 1.8 GHz: 33.6 fps
iMac G4 1.25 GHz: 21.6 fps

It doesn't look like 3 times more to me, as Apple claim.

It's probably due to the game settings. If you lower the resolution right down, then the performance becomes more CPU limited than graphics limited. Same thing if you turn off all the graphical effects, lower textures, etc. Apple probably did this to demonstrate the difference in raw CPU power. If you turn all the eye candy back on and play at a normal resolution, then the GPU will have a far greater impact, and you'll get results like that above. So, Apple probably weren't lying, just doing selective benchmarking. As usual.

EDIT: Seems I'm partially right. I see that both benchmarks were performed at 1024 x 768 with 32-bit colour. However, Apple's used low quality whereas Barefeat's used High quality. I'm surprised the difference is so large, but that possibly explains the difference. They also used a different UT2004 benchmark - maybe that was less GPU intensive as well.

[Crusoe]

Summary, If you're not a gamer the 1.8 is a deal and the 5200 is a PoS no matter what system you put it in. (You decifer the acronym)

I like the new iMac and will be getting a 1.8 though I would have liked it with 128 MB VRAM for texture rendering and a smidge of future proofing.

[MrForgetable]

Originally posted by Crusoe:
Summary, If you're not a gamer the 1.8 is a deal and the 5200 is a PoS no matter what system you put it in. (You decifer the acronym)

hmm.. oh what could be???

[bradleykavin]

reality is if your playing games its better to go with a powermac than an imac

[MrForgetable]

reality is that if you really wanted to play games you'd build your own PC :shrug:

[DaBeav]

Or get an XBox/PS2/NGC.

[Loco Engr]

Originally posted by Eriamjh:
Why didn't they compare the iMac to a SINGLE 1.8GHz G5 PM?

I think the most important comparison is the new entry iMac G5 vs. the entry eMac G4. My guess is value (performance/price) tilting toward the eMac.

EDIT: All the numbers are there -- just not on the same page. According to MacWorld's SpeedMark tests, the current eMac G4 (1.25 GHz, 512 MB) scores 129. The new iMac G5 (1.6 GHz, 512 MB) scores 150.

This represents only a 16% performance improvement for the iMac -- while the clock goes up 24% and the base price increases 62%! Of course, people will buy what they want.

[i am yujin]

Originally posted by DaBeav:
Or get an XBox/PS2/NGC.

Yeah because UT 2K4 is out for any of these.

[BenRoethig]

IIt'd be nice if they added a 1.8ghz 17" "Special Edition" with either a 128mb Radeon 9600XT or a Geforce 5700 instead of the64bm 5200.

[DaBeav]

Originally posted by i am yujin:
Yeah because UT 2K4 is out for any of these.

Not the same version, no. You can get Unreal Championship II, and Unreal II: The Awakening for XBox however. You can even play online using XBox Live with voice chat and everything.

Consoles are becoming quite advanced. The next generation from MS and Sony will push the limits even more. You can get an XBox for less than a video card upgrade and games will be available long before you can get them for the Mac.

To be blunt, the Mac, even the PowerMac G5s, is not a great gaming platform right now. The hardware potential exists, but PCs still give a better dollar to performance ratio. That's why I have a Mac for computing (and for RTS games - they don't work too well on a console) and an XBox for gaming. Halo runs great with great framerates and lots of eye candy and I'll be able to play Halo 2 in a couple months - long before it comes out for the Mac.

[Pierre B.]

Originally posted by PEHowland:
It's probably due to the game settings. If you lower the resolution right down, then the performance becomes more CPU limited than graphics limited. Same thing if you turn off all the graphical effects, lower textures, etc. Apple probably did this to demonstrate the difference in raw CPU power. If you turn all the eye candy back on and play at a normal resolution, then the GPU will have a far greater impact, and you'll get results like that above. So, Apple probably weren't lying, just doing selective benchmarking. As usual.

EDIT: Seems I'm partially right. I see that both benchmarks were performed at 1024 x 768 with 32-bit colour. However, Apple's used low quality whereas Barefeat's used High quality. I'm surprised the difference is so large, but that possibly explains the difference. They also used a different UT2004 benchmark - maybe that was less GPU intensive as well.

Hold on, here is round 2. It is really strange.