Mac vs PC benchmarks

Hi guys, I am no mac basher, in fact I am in the market for a Mac mini to introduce myself to OSX.... BUT

I have a mac-ophile friend who claims that the 1.4 G4 PPC will beat a 3.06 P4 or 3000+ A64. He also believes that the dual 2.5 G5 is faster than 3.6 Dual Xeons or similar Opteron systems.

I have googled for it, but does anyone have any solid Mac vs PC benchmarks that I can send him?

I am sorry to tell you that no such thing exists as far as I am aware

I don't really see any fair comparision between two inherently different platforms I am sure people can try to come up with something. Besides look through the forums you will see all sorts of interesting mac vs. pc threads

THis one's from a PC website, I didn't read the whole thing, but it could be biased:
http://www.pcworld.com/news/article/0,aid,112749,pg,1,00.asp
DIscussion at ace's
http://www.aceshardware.com/forums/read_post.jsp?id=105045929&forumid=1#TEXT

This one definately sounded mac-biased:
http://www.barefeats.com/g5op.html

Hey, if you want to donate one of each to me, I'll run some benchmarks

I wouldn't call the last one biased aside from the fact that they're using cinamerealistic photo and video rendering software which the MAC is made specificly for. The P4 actually fares better than the Opteron in that category.

theMonster said:

I wouldn't call the last one biased aside from the fact that they're using cinamerealistic photo and video rendering software which the MAC is made specificly for. The P4 actually fares better than the Opteron in that category.

Click to expand...

It's the rest of the article that seems biased (all the rubbish about Mac's being cheaper than PCs) Which does put into question the benchies.

I didn't saay that it was biased, I said it sounded biased.

Like this part:

The Dual Opteron was running Windows XP Professional, not Windows XP 64 bit edition... which is officially avaialable only for the Itanium. Several of you pointed out to me that it's available in BETA form for the Dual Opteron, but...
... there is no Windows 64 drivers for the Radeon,
... and none of the software we used is optimized for Windows XT 64,
... so we decided to wait until the next round in February to test with it.

OS X isn't really a 64 bit OS... so I guess you could say we were fair.

Click to expand...

First, it's not a 64-bit OS, but it does have some 64-bit portions, I believe.

Second, appliations compiled for 64-bit mode on the Opteron should benefit form the additional registers, I don't think that running in 64-bit on the PPC-970 gets you extra registers, so it would only speed up 64-bit integer operations.

Third, I strongly dislike windows, and want to see linux benchmarks.

edit:
He also says "Both had 2GB of DDR400 memory." Then, when listing the configuration, he says the Opteron box used only DDR333: " MEMORY? 4 x 512MB DDR333 Registered ECC (2 in each channel)"

I'm also not sure why he lists two sticks on each channel. 2x Opterons should have 4 memory channels (2 for each cpu). So either he set it up wrong, halving the memory bandwidth, or just didn't really know what he was talking about.

First off this is a Dual 2gig Apple against a Single processor on page 8 the tests show apple wine BUT gets beat by a dual Opteron ..
http://www.pcworld.com/news/article...749,pg,1,00.asp

Also i just ran CineBench on a syatem I have about $1200.00 Into lol..It renders the image in 38 seconds In there test it was 50 for mac and 51 for the Opteron..


Bang for buck and power relation to the dollar the P/C usually wins hands down..

Wait my time included having , Norton running , a firewall,Aim , Yahoo , MSN , and 6 web pages open as well as out look expres 2003 ...OOPS I forgot to turn things off...

yeah i know but my friend claims otherwise.

How does everyone reckon the San Diego and Dothan cores compare performance per clock with the PPC-970?

>HyperlogiK< said:

yeah i know but my friend claims otherwise.

How does everyone reckon the San Diego and Dothan cores compare performance per clock with the PPC-970?

Click to expand...

No matter if your friend claims other wise,I just showed you a Xeon with 2 x $50.00 Chips walk that Mac and Opteron.I am gonna run the bench test again and get a scren shot of it so you can show your friend that for around 1200.00 ya can get performance close to the G-5.

And just because ya cant self build a MAC does not mean a tester should disallow pricing based on a PC being self buiilt matter of fact he should have shown the Pc both ways but his love for Mac wont let him..

I lowerd my clock speed to 3200,And also as you can see,have a bunch of crap open and or running.Bottom Left corner time 38.6 Seconds after I removed 52MHz,Yesterday at 3252 i was 38.0 exactly.
http://img.photobucket.com/albums/v87/diehrd/Cinebench2003.jpg

I am gonna surf around and look at the times for a G-5 On this bench test.

Here is a true dual to dual shoot out with CineBench test 2003.Let me say also I chose this as a base comparison BECAUSE macs are supposed to be so much better then a P/C in renderring and video production..Ha
http://www.barefeats.com/g5sum02.html

Most of the tests I can not do my self but 1 item is weird,This tester shows the G5 @ 2.5 gigs smoking the P/C entries BUT he only tested the Xeon 3.0 in CineBench where it beat the G5 2.5 gig rig,In all other tests he used a Dual 2.4 gig Xeon which lost to the G5..I am not sure why he would do this,My opinion is he has a point to make so he flip flops the systems in order to nake his point.Also he spends a lot of time testing a Dual Cpu rig against Single Procesor P/C rigs BUT is using Dual Cpu bench mark software.Very weird guy.
http://www.barefeats.com/pentium4.html

As a proud Mac fan, let me tell you that your friend is full of crap

Don't get me wrong--AMD has done everything necessary to prove to the x86 world that the "megahertz myth" crap Apple has spewed for years really is true, but it's not that true.

Completely out of my ***, I'd say that a 1.4Ghz G4 is comparable to an x86 in the lower 2Ghz range. When it comes to applications that make full use of the AltiVec extensions, it may squeeze out a little more, but a 1.4Ghz G4 is certainly not going to touch a 3Ghz P4. By the time Apple finally put it to rest in their professional grade machines, the G4 was a dinosaur--just look at the FSB on most of them.

Similarly, while the G5 will perform better, I wouldn't expect a 2.5Ghz G5 to outperform a 3.6Ghz P4. The G5 is still a marvel of efficiency and all, and it benefits from a more modern architecture, but I don't believe for a second that it's near as efficient as the G4 was. IBM was willing to sacrifice its performance/mhz ratio in order to increase the overall clockspeed (just as Intel has in the past).

Also, I don't believe anything in OS X is 64-bit yet. They're currently developing a 64-bit version for the Tiger release I think. Not really sure though, I haven't kept up with this so much lately. Down to only 1 Mac in the household

This article:
http://www.theregister.co.uk/2003/07/07/mac_os_x/

claims that 10.3 was planned to make some use of 64-bit addressing and data paths. I assume this is what happened.