[Todd Madson]

I saw the news in the macnn.com main site about the 2.0 ghz Freescale G4
upgrade chip and thought:

If anyone gets this, would a challenge be do-able with a 2.0 ghz G5 tower?

Just install CHUD tools on the G5 and disable one of the CPUs and do a
full-on head to head benchmark and application challenge.

I'm unsure if the 2.0 ghz upgrades are shipping yet but is anyone game
for this?

Could be quite interesting to see what arises.

[Salty]

Pretty expensive card to get just to have fun on MacNN

But if someone bought it could be fun to watch... I should hope the G4 gets spanked.

[Eriamjh]

According to Barefeats.com, the G5 wins. Here it is...

Originally Posted by Barefeats.com

PERFORMANCE INSIGHTS
1. Though the 2.0GHz G4 upgrade produced respectable performance, it's no match for a dual G5 Power Mac running at the same clock speed. The argument is not settled, though. Until you can put a G4 CPU in a motherboard with the same frontside bus speed and same memory bus speed as the G5, we can't really say whether a G4 running at the same clock speed can match it. The reality is that G5 systems have a much faster frontside bus, memory bus, and graphics slot.

Plus, the G5 Power Mac can be expanded to either 4GB or 8GB of memory, depending on the model. Most G4 Power Macs are limited to 1.5GB. Therefore, the meaningfulness of the comparison will remain questionable.

2. Spending extra for dual processors makes sense if you are running multi-processor aware applications like Photoshop (see "MP" test above). Even then, not every Photoshop function uses dual processors as you can see by our "SP" test. For most users, a fast single processor should be sufficient.

One the other hand, if you run multiple active applications (like a long FileMaker script or X-Code compile or Final Cut Pro export while checking your email and editing something in your word processsor), then having dual processors can be very helpful.

3. Note that the Dual G4/1.4GHz is faster than the Dual G4/1.8GHz when running Photoshop MP actions and all four 3D Game tests. That's because the Dual G4/1.4GHz Power Mac has 2MB of L3 cache per processor. The Dual 1.8 only has 512K L2 cache per processor.

4. Spending $1000 on a modified FireGL is a waste when the G4 Power Mac can't run 3D apps much faster than with a $212 Radeon 9800 Pro. On the other hand, a G5 makes big gains with the X800 XT (which costs half as much as the modified FireGL).

Satisfied?