G5 faster than intel pro

[gpp]

I have a 2000 g5 pro and the first intel pro that came out in 2006.i ran a test to transcode a 1.7 gb video using toast 8 and 10 on the intel machine.both are the basic version but i dont know the specs.the older one transcoded it in 4 min and the newer intel one took 12.seems to me it should be the other way around.since i was looking to buy a new one this kinda put a stop on that.could there be something wrong with the intel machine .just for kicks i did the same with a 2011 basic mac mini and it took 35 minutes.

 

[simsaladimbamba]

Do you have a 2005 G5 PowerMac? Or do you have another PowerMac?
What about posting the model identifier of each machine and its CPU speed and processor count?
To find out, what exact Mac model you use, click > About This Mac > More Info > Hardware Overview:

To see a more elaborate image explaining the above steps, click here.
To see a more detailed video tutorial showing the above steps, click here (2 MB, 23s, .mp4).​

Strangely enough, the G5 is a good CPU, but PowerPC CPUs use a different architecture than x86 CPUs (RISC vs. CISC), thus maybe you have seen something easily explainable, but for that we might also need to know, what kind of video you choose to transcode and to what kind of format you transcoded it with.

 

[burnout8488]

PowerMac G5, no "Pro" about it. You'll just confuse others.

Mac Pro = Intel. PowerMac = PowerPC G3/G4/G5

 

[gpp]

Ok,its a macpro1.i,dual core intel xeon,2.66ghz,2processors,4core and 5 gb of memory.the other one ispower mac g5,1.6 ghz .hope this helps.

 

[simsaladimbamba]

Ok,its a macpro1.i,dual core intel xeon,2.66ghz,2processors,4core and 5 gb of memory.the other one ispower mac g5,1.6 ghz .hope this helps.

Spaces do help a lot with readability.

What about the video? What was the source, what was the setting you chose?

In order to find out, at least about the source:

• VideoSpec (free)
• MediaInfo (free)
• Media Inspector (2.99 USD € in the Mac App Store (MAS))

When done, you can use the "Report" (VideoSpec) and "Export" (Media Inspector) button to export a report,
attachable to your next post via the

button.
In MediaInfo you have to go to the Menu Bar, select View > Text and copy the text via CMD+A > CMD+C and paste it via CMD+V into your next post.​

 

[deconstruct60]

I have a 2000 g5 pro and the first intel pro that came out in 2006.i ran a test to transcode a 1.7 gb video using toast 8 and 10 on the intel machine.both are the basic version

Are you sure you are comparing the same transcoding codec's ? There is a version difference of 2 between the two pieces of software. So it isn't the 'same' software.

I don't think Toast 8 is a universal binary. It likely is an 'end of the line' highly PPC optimized app. If targeting an older codec the Altivec code is likely more optimized. The newer x86 targeted on probably not so optimized ( especially if it was a code in decline without hardware support).

If trying to expose differences based on hardware should be targeting a code that has hardware support. Otherwise is a software or at least software+hardware combo at the heart of the issue.

If engaged the Intel Quick Sync Video transcode infrastructure of the mini should crush the other two.

 

[GermanyChris]

Are you sure you are comparing the same transcoding codec's ? There is a version difference of 2 between the two pieces of software. So it isn't the 'same' software.

I don't think Toast 8 is a universal binary. It likely is an 'end of the line' highly PPC optimized app. If targeting an older codec the Altivec code is likely more optimized. The newer x86 targeted on probably not so optimized ( especially if it was a code in decline without hardware support).

If trying to expose differences based on hardware should be targeting a code that has hardware support. Otherwise is a software or at least software+hardware combo at the heart of the issue.

If engaged the Intel Quick Sync Video transcode infrastructure of the mini should crush the other two.

Toast 10 is fat..I used it on my Quad but it's the last 11 is not