All Macs In-Depth Tests

The Performance Edge: A Dueling, Dual Processor Shoot Out - A First Look At The Performance Of The New Gigahertz Tower Compared To The Last Generation Dual Processor Machine

Friday, February 15, 2002

This is our first look at the performance of the new Power Mac Dual G4/1 GHz machine. We decided to first compare it the the previous generation's dual processor machine, the Dual G4/800. Below are just a few preliminary tests, each of which stresses both of the machine's processors. All the tests below are processor intensive tasks. We will be doing much more extensive testing in the following week and hope to give you a well rounded look at all the performance facets of this new machine.

Difference and similarities in processor and memory systems of each machine

Additional Resources

Check prices at:

 

Difference and similarities in processor and memory systems of each machine


  Power Mac Dual G4/800 Power Mac Dual G4/1000
Processors 2 x G4/800 2 x G4/1000
L2 Cache 256k @ 800 MHz 256k @ 1 GHz
L3 Cache 2 MB @ 200 MHz 2 MB @ 250 MHz DDR
Memory 256 MB PC 133 512 MB PC 133
Price $3,499 $2,999

"Real World" Tests

All of the tests below were timed with a stopwatch. The times are then converted to percentages relative to the Power Mac Dual G4/800, which is set to 100%. For all scores, higher numbers are better.

Number Crunching & Rendering Tests

In the test above we are running a Ripple Effects filter on a exported QuickTime movie. The file size is about 175 MB. In the Gigahertz machine both processors were utilized but only about 60% of their processing potential was being used. In other words, 40% of the processing capacity of the machine was sitting idle. Some bottleneck is preventing data from reaching the processors fast enough to saturate them.

The Fractal program has been highly tuned to take advantage of the G4 and is precisely the type of work that the G4 was made for. It also takes good advantage of dual processors. Both processors were kept completely saturated with data.



Encoding/Decoding Tests

For the test above we launched two separate copies of QuickTime and had them each do an encode on individual files. Again in this test, both processors were kept completely saturated with data

Converting QuickTime movies to DV allows you to import them into iMovie. Of all the tests this was the best result we had for the Gigahertz machine over the 800 MHz one.

Here we are runing the QuickTime Sorenson encode and iMovie Ripple effect at the same time. Both processors were completely saturated when both applications were running. However the Ripple Effect finished before the Sorenson Encode. When only the encode was running, processor usage dropped to only a little over 50%. If I am not mistaken the Sorenson encode capability that is built into QuickTime is hobbled and does not utilized dual processors.

The beauty of dual processors and OS X for Power Users is that even if a program you are running does not itself take advantage of dual processors, if you plan your work strategy well, you can always launch another application to soak up whatever horsepower is being left on the table.

All in all the Gigahertz machine turns in about a 30% speed improvement over the dual 800 MHz when doing processor intensive tasks. A 30% speed improvement and a 14% price drop ... not too bad.