MacWorld review of 2009 MP quad and octo

[RebootD]

What I don't get is that macworld says the quad is better, while barefeats and tesslator's graph with the cinebench test here on the boards all say octo 2,26 -> octo 2,8 -> quad 2,66.

I myself am deciding which of these three but I'm quite confused as to why macworld's results are so different!

Isn't the big issue with Macworld reviews is they review them with the stock ram configs and only test it one program at a time? Sure a 3Ghz iMac will beat a 2.2 Xeon but try the same test with 8 other apps open, put 12GB in the MP and see who wins. =)

 

[jjahshik32]

OWC now has 4GB 1066MHz DDR3 DIMMs listed as supported for the Quad-core Mac Pros, with 16GB Max! If/when you need to upgrade past 8GB, prices for 4GB DIMMs would probably have fallen substantially by then.

Considering that the 2.93GHz Quad cost $2999 and that $970 bucks for 4x4GB=16GB of RAM! No thanks. For that kind of dough, I'd rather just get the 2.26GHz Octad (only $246 difference) + 16GB of RAM for $283 from transintl and I can still buy 3x2TB hdds for $179 a piece.

 

[DeepCobalt]

If you think 6, 8, 12 or even 32G of RAM is 'excessive', don't click this link.

Yeah, but Dell is using 16GB DIMMs... Presumably (after OWC's successful use of 4GB DIMMs in the new quads) we could put 16GB DIMMs into an octad Mac Pro to get 8 x 16GB = 128GB. Not 192GB, but not bad.

One annoyance with the Mac Pro, however, is it isn't clear yet whether the 2.66 and 2.93 octads are set to recognize the faster 1333 RAM. The Dell does.

 

[Ploki]

I left the EXS sampler out of my comment because I do not use it much. It can access more ram outside of Logic, but this is a fairly new (within the last year) option. I have also not noticed clocking as having much effect on my buffer setting. Core audio drivers have more to do with that.

The point is that huge amounts of ram are not necessary for all professional uses. Clock speed and hard drive access speed can many times be more important.

what do you mean that core audio drivers have more to do with that?
i dont think i understand what you mean

 

[Igantius]

Isn't the big issue with Macworld reviews is they review them with the stock ram configs and only test it one program at a time? Sure a 3Ghz iMac will beat a 2.2 Xeon but try the same test with 8 other apps open, put 12GB in the MP and see who wins. =)

It's the standard review format for MW - not saying it's a good or bad way, but this is the way it's carried out reviews for ages.

With regards to the RAM issue, it's going to be Apple that supplies the machines. Now, one might argue that MW should pay out of its own pockets for more RAM or get some from somewhere - but again, this is the way it's done it in the past and arguably, the type of buyer who MW is mainly targetting is more likely to get a machine 'off the peg' and upgrade the memory later.

Also, MW may not have not that long to test the machines - the reviewer may have had a tight deadline! And we don't know the testing conditions - I can't speak for the US, but in the UK, these haven't been great always - e.g. with the Cube, each of the 3 Mac mags were given about 2 hours each with a machine in order to do a review.

Regardless of benchmarks, reviews do tend to be subjective - this one reflects what MW merely thinks the majority of its readers are interested in.

Meanwhile.... MW UK have used the American review and have given the Quad maximum marks. PCMag.com have posted a revied for the Oct. (Quad to follow) - http://www.pcmag.com/article2/0,2817,2343855,00.asp

 

[RebootD]

Meanwhile.... MW UK have used the American review and have given the Quad maximum marks. PCMag.com have posted a revied for the Oct. (Quad to follow) - http://www.pcmag.com/article2/0,2817,2343855,00.asp

Well in this case CURSE YOU ADOBE!!! This is from that PCmag.com link,

"I tested the system under both Mac OS X and Windows Vista 64-bit and was surprised to find that the system is at times a better performer under Windows. The Mac Pro finished the Photoshop CS4 test in 1:50 under Mac OS X but took only 1:17 to finish the same tasks under Windows. The performance difference is due to the fact that the Mac version of CS4 can directly address only 3.5GB of the system's memory, while the Windows 64-bit version can access the full 6GB of memory. This can add up to a significant performance benefit."

 

[Boneoh]

Well in this case CURSE YOU ADOBE!!! This is from that PCmag.com link,

"... The performance difference is due to the fact that the Mac version of CS4 can directly address only 3.5GB of the system's memory, while the Windows 64-bit version can access the full 6GB of memory. This can add up to a significant performance benefit."

This has been a known bummer for a long time. Adobe PS won't support 64 bit until CS 5.

 

[Igantius]

Another review

http://www.reghardware.co.uk/2009/04/02/review_desktop_mac_apple_mac_pro/

El Reg gave it 80% - a decent mark for the site, and a pretty glowing write-up. The comment section should be fun to read

 

[rmpstudio]

This has been a known bummer for a long time. Adobe PS won't support 64 bit until CS 5.

And this brings me to my comments/points.... If you are correct Boneoh (and i'm assuming you are) and considering the below excerpt from the MW article, then i think that if somebody is going to be buying a mac based on performance (we all do EVEN IF WE AREN'T PROS - WE DON'T HAVE TO BE!) then we should buy based on where technology is heading and i suspect it's heading towards multi-core optimized programs.

"The new 2.26GHz eight-core Mac Pro has twice as many processing cores as the 2.66GHz quad-core model, but each core runs 15 percent slower than the cores in that 2.66GHz quad-core Mac Pro. Because many applications have a difficult time using even four processors efficiently, the advantage of having eight was not apparent in most of the application tests that make up our Speedmark benchmark test suite."

I'll be upgrading from an IMAC so i don't care about the 15% slower because i want the 16GB of RAM and i'm "betting on the come" (i.e. i will be buying the upgrade to CS5 and by then i suspect the 2.26 Octo will outperform the 2.66 quad - or maybe even 3.2 quad)

 

[Boneoh]

I'll be upgrading from an IMAC so i don't care about the 15% slower because i want the 16GB of RAM and i'm "betting on the come" (i.e. i will be buying the upgrade to CS5 and by then i suspect the 2.26 Octo will outperform the 2.66 quad - or maybe even 3.2 quad)

I'm going through a similar upgrade. My old machine is one of the original MacBook Pros. I'm seeing about 5 times the speed in my large builds. I'm running VMware with windows 2003 server. I've set it up as 4 virtual processors with 4 gb of ram. Its performing very well.