G5 vs. Intel speeds

Question

Level 1

81 points

G5 vs. Intel speeds

First, I heard how PowerPC was so much faster than Intel. Now, I'm hearing the reverse.

I bought my wife a dual-core Intel iMac (with 1 GB RAM). I ran XBench 1.2 (which is a univeral binary app). Here are the results, compared with my Dual Processor 1.8 GHz, 2 GB RAM.

Both computers have 7200 RPM SATA hard drives.

G5: 89
iMac: 54

Power Mac G5 Dual 1.8GHz Mac OS X (10.4.6) 2GB RAM, Barracuda HD

Posted on Jun 2, 2006 3:56 PM

Reply

Answer 1

Jun 2, 2006 4:25 PM in response to Dunestrider

XBench isn't the best benchmark out there. It's inconsistent (not good when trying to do comparisons) and the baseline often changes between versions. However, it does test many different aspects of your configuration.

I'd recommend using CineBench to get a fairer comparison on the relative CPU and video card power. I'd expect your computers to be relatively close in terms of CPU power, but I expect the iMac will have better graphics performance, due to the fact that it's much newer (assuming you haven't replaced the video card in the PowerMac).

It really depends on what kind of software you're running, though. For certain apps, the Intels will be faster, for others, the PPC will be faster.

Reply

Answer 2

ds store

Level 7

30,855 points

Jun 2, 2006 5:45 PM in response to Dunestrider

You would most likely be interested in this, it's more accurate.

http://www.geekpatrol.ca/blog/74/

The PowerPC G5 is still a good processor. In fact, it’s still a great processor. Apple isn’t switching to Intel chips because Intel chips perform better, but rather because a G5 would melt through the bottom of a laptop

Says it all right there.

Since the iMac G5, Powerbook G4, and iBook G4 were all single processor machines, upgrading them to dual cores is a pretty big improvement.

Still Intel needs to come out with something better that will take on the Quad and the Core Duo is only 32 bit with a 4GB or so RAM limit for the processor, the present MacTels only be upgraded to 2GB of RAM.

Right now the only thing I could see is some sort of 64 bit Quad core with two of those (8 total cores) in a new Mac Pro.

Reply

Answer 3

Dunestrider Author

Level 1

81 points

Jun 3, 2006 12:13 PM in response to Steve Boultbee

XBench isn't the best benchmark out there. It's
inconsistent (not good when trying to do comparisons)
and the baseline often changes between versions.

I respectfully disagree. I get reasonably consistent scores.

However, it does test many different aspects of your
configuration.

Yes, it tests the hard drive as well. Given two identical systems (CPU, memory, video card), the system with the faster hard drive will appear much faster to the end user.

I'd recommend using CineBench to get
a fairer comparison on the relative CPU and video
card power. I'd expect your computers to be
relatively close in terms of CPU power, but I expect
the iMac will have better graphics performance, due
to the fact that it's much newer (assuming you
haven't replaced the video card in the PowerMac).

I'm not interested in just the CPU and video card, but in the "total package".

It really depends on what kind of software you're
running, though. For certain apps, the Intels will
be faster, for others, the PPC will be faster.

That's why a benchmark is useful.

Power Mac G5 Dual 1.8GHz Mac OS X (10.4.6) 2GB RAM, Barracuda HD

Reply

Answer 4

Jun 3, 2006 1:34 PM in response to Dunestrider

XBench isn't the best benchmark out there. It's
inconsistent (not good when trying to do

comparisons)

and the baseline often changes between versions.

I respectfully disagree. I get reasonably consistent
scores.

I suspect the truth may be somewhere in the middle. Xbench can produce different scores on the same hardware, even on consecutive runs or after a restart, but they don't seem to be wildly inconsistent. Some people do recommend running it 3 times and taking an average though. Similarly as far as I'm aware there have been 2 baselines, a dual 800MHz G4 and a (dual ?) 2.0GHz G5 if I remember correctly.

Reply