PDA

View Full Version : Dual Core PowerMac G5 Benchmarks (XBench)


MacBytes
Oct 19, 2005, 07:16 PM
http://www.macbytes.com/images/bytessig.gif (http://www.macbytes.com)

Category: Benchmarks
Link: Dual Core PowerMac G5 Benchmarks (XBench) (http://www.macbytes.com/link.php?sid=20051019201631)

Posted on MacBytes.com (http://www.macbytes.com)
Approved by arn

Chaszmyr
Oct 19, 2005, 08:09 PM
In other words, the Quad really is gonna be packing some serious power.

Bear
Oct 19, 2005, 08:59 PM
In other words, the Quad really is gonna be packing some serious power.And hopefully be a nice machine to have for at least 3 years. Which would be about the time (in my opinion) to get an Intel based PowerMac.

Santaduck
Oct 19, 2005, 09:49 PM
Thanks for the link!

Does anyone know of a quickie URL for xbench numbers for other machines, just for reference to this article?

Can't wait until www.barefeats.com gets a hold of one and really runs it through its paces.

plinden
Oct 19, 2005, 09:55 PM
Thanks for the link!

Does anyone know of a quickie URL for xbench numbers for other machines, just for reference to this article?

Can't wait until www.barefeats.com gets a hold of one and really runs it through its paces.
http://ladd.dyndns.org/xbench/comparesubindex.xhtml?machineTypeID=24&sort=processor&minVersion=1.2

SiliconAddict
Oct 20, 2005, 02:46 AM
In other words, the Quad really is gonna be packing some serious power.


o.O

Maybe I'm reading the benchmarks wrong but that isn't how I interpreted it at all. It seems that there is a performance boost over a 2.5Ghz but nothing spectacular with a 200Mhz diff which, honestly, is to be expected.
Adding extra cores isn't going to automatically give you some massive performance boost across to board. The question is: We know how well OS X and its various API's handle dual CPU's. X has been running off of dualies for years. How are they with, for all intents and purposes, quad CPU's? 4 CPU's means nothing if the apps that run on it don't know how to use the other two in a meaningful way. I have to imagine properly tuned apps like Photoshop would SCREAM like a bat out of hell as long as they KNOW how to use all 4 CPU's in the system. :confused:

sparkleytone
Oct 20, 2005, 08:09 AM
o.O

Maybe I'm reading the benchmarks wrong but that isn't how I interpreted it at all. It seems that there is a performance boost over a 2.5Ghz but nothing spectacular with a 200Mhz diff which, honestly, is to be expected.
Adding extra cores isn't going to automatically give you some massive performance boost across to board. The question is: We know how well OS X and its various API's handle dual CPU's. X has been running off of dualies for years. How are they with, for all intents and purposes, quad CPU's? 4 CPU's means nothing if the apps that run on it don't know how to use the other two in a meaningful way. I have to imagine properly tuned apps like Photoshop would SCREAM like a bat out of hell as long as they KNOW how to use all 4 CPU's in the system. :confused:

OS X is optimized for SMP, not dual CPUs. This basically means that its going to know what to do with 2,4,8,etc processors. I'm sure there's a logical limit and a real world cutoff for performance increase, but I doubt 4 is it.

SiliconAddict
Oct 20, 2005, 08:22 AM
OS X is optimized for SMP, not dual CPUs. This basically means that its going to know what to do with 2,4,8,etc processors. I'm sure there's a logical limit and a real world cutoff for performance increase, but I doubt 4 is it.

Yes but does the OS know how to balance the load and threads from two to four processors? Itís like adding a 3rd arm. I know its there because I can see it but if I don't know how to use it while juggling its not doing me a lot of good.

dazzer21
Oct 20, 2005, 08:44 AM
I just went to the PowerMac section of the UK Apple Store - specced a fully loaded 2.5 dual core for a laugh (including £8000+ worth of 16GB ECC RAM and 2x30" screens) and it came to £2900-odd. Sounds like a bit of a bargain to me...

cwtnospam
Oct 20, 2005, 08:45 AM
Interesting article, but am I the only one who found myself correcting grammar more than reading it? If you're going to publish articles, you should at least hire an editor.

Dont Hurt Me
Oct 20, 2005, 08:49 AM
Yes but does the OS know how to balance the load and threads from two to four processors? Itís like adding a 3rd arm. I know its there because I can see it but if I don't know how to use it while juggling its not doing me a lot of good.
Plus what programs are going to know what to do with that 2nd cpu let alone a 3rd and 4th :confused: Reminds me of when they came out with the dual powermacs but there wasnt anything really using that 2nd cpu at the time. Even most of todays games ignore that 2nd cpu, i think this is a pure professional machine with little to offer a consumer. I wonder what apps will know what to do with 4 cores.

Chaszmyr
Oct 20, 2005, 11:13 AM
o.O

Maybe I'm reading the benchmarks wrong but that isn't how I interpreted it at all. It seems that there is a performance boost over a 2.5Ghz but nothing spectacular with a 200Mhz diff which, honestly, is to be expected.
Adding extra cores isn't going to automatically give you some massive performance boost across to board. The question is: We know how well OS X and its various API's handle dual CPU's. X has been running off of dualies for years. How are they with, for all intents and purposes, quad CPU's? 4 CPU's means nothing if the apps that run on it don't know how to use the other two in a meaningful way. I have to imagine properly tuned apps like Photoshop would SCREAM like a bat out of hell as long as they KNOW how to use all 4 CPU's in the system. :confused:

These benchmarks, as I am sure you realize, are not testing a quad, they are testing a single dual core processor. OSX is in no way limited to 2 processors, and almost every app that supports dual processors will also support quad processors, because of the way these apps thread.

bloodfist
Oct 20, 2005, 12:35 PM
My DP 2.3 gets 115.46 overall in Xbench, while the DC 2.3 gets 108.7 overall.

I get a 114.26 on the CPU test while the DC 2.3 gets a 114.32. I get a 116.53 on the thread tests while the DC gets a 113.74. Not a great deal of difference between DP and DC models.

Weren't these DC chips supposed to be 40-80% faster? I know its an artifical benchmark, but there should be some performance gain shown.

Maybe my not waiting for these DC machines was a good idea, now that I see the performance gains aren't that great. (I just got the DP a week before the new PM came out)

-bloodfist

whooleytoo
Oct 20, 2005, 03:06 PM
Plus what programs are going to know what to do with that 2nd cpu let alone a 3rd and 4th :confused: Reminds me of when they came out with the dual powermacs but there wasnt anything really using that 2nd cpu at the time. Even most of todays games ignore that 2nd cpu, i think this is a pure professional machine with little to offer a consumer. I wonder what apps will know what to do with 4 cores.

I reckon if you run more than one application at a time, or run any application with more than one thread (i.e. most of us), you'll see the advantage of multiple cores/cpus.

maya
Oct 20, 2005, 03:13 PM
Plus what programs are going to know what to do with that 2nd cpu let alone a 3rd and 4th :confused: Reminds me of when they came out with the dual powermacs but there wasnt anything really using that 2nd cpu at the time. Even most of todays games ignore that 2nd cpu, i think this is a pure professional machine with little to offer a consumer. I wonder what apps will know what to do with 4 cores.

With the Quad chip PM on sale, Apple is entering the "workstation" market. This says a big thing in a big way. For a workstation it still lacks a few key components, however these will b e fixed later down the road I hope. :)

Lacero
Oct 20, 2005, 03:16 PM
Dual cores and only a marginal 69% improvement? In an ideal world, I was hoping to see 90% or greater efficiency.

Too bad no comparisons based on running multiple apps. Maybe we'll see better performance this way.

SPUY767
Oct 21, 2005, 02:48 PM
Dual cores and only a marginal 69% improvement? In an ideal world, I was hoping to see 90% or greater efficiency.

Too bad no comparisons based on running multiple apps. Maybe we'll see better performance this way.

Now I have to dig through the cushions of the couch to find 4 grand for a Dual Dual.