Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

usptact

macrumors regular
Apr 2, 2011
157
0
Oh my... oh my... this test is just pure nonsense. What is the baseline? What do they compare to? 27inch display with MacPro gray metal casing?

If you compare processors then launch comparison on applications that are mostly processor hungry. Don't show that your macX starts faster than macY.

If you compare hard drives then do benchmarks with file operations. There are many parameters to compare to. Just showing application opening is the last test to do...

In comparison one should be super precise about what is compared! First making sure that apples and oranges are comparable!
 

chrmjenkins

macrumors 603
Oct 29, 2007
5,325
158
MD
I'm pretty sure a pointless claim is pretty dumb. But I'm pretty sure its also incorrect. Even a MP with a westmere processor with a couple of striped vertex 3s will beat a sandy bridge imac. Its pointless to compare an SSD equipped imac with a MP that only has a regular HD. Especially using a benchmark that relies heavily on drive speed.

Which only underscores how pointless it is. Given that the mac pro is infinitely more configurable than the iMac, the CPU is the only basis of comparison. The fact that the CPU is faster is a foregone conclusion and the fact that they didn't attempt to match storage medium only adds insult to injury.
 

wiz329

macrumors 6502a
Apr 19, 2010
509
96
Which only underscores how pointless it is. Given that the mac pro is infinitely more configurable than the iMac, the CPU is the only basis of comparison. The fact that the CPU is faster is a foregone conclusion and the fact that they didn't attempt to match storage medium only adds insult to injury.

Haha well I'm glad we'll we're all pretty much agreed on how pointless this test was =)
 

mabaker

macrumors 65816
Jan 19, 2008
1,209
566
Still rocking a G5 here. It's laughing all the way. :D that new one is smoking though!
 

Lone Deranger

macrumors 68000
Apr 23, 2006
1,895
2,138
Tokyo, Japan
For those rare "properly threaded" tasks. ;)

The problem is that "properly threaded" somewhat implies that it's the skill of the programmers. Some applications have serial dependencies in their data which makes arbitrary multi-threading impossible.

A "properly threaded" application could be a video encoder. Many (most?/all?) video codecs have the concept of "keyframes" - a complete frame without any dependencies. These typically are placed every few handfuls of seconds in the stream. From one keyframe to the next, temporal compression is used to reduce the data rate.

Such a data flow is easy to parallelize to an arbitrary number of cores. You break the input into chunks based on the keyframe interval (if you have keyframes every 5 seconds, a one hour video is 720 chunks). You then start 720 threads to encode, and combine the results. (Realistically, you'll schedule the 720 threads a small number at a time, based on the number of cores that your have.)

Other applications are very serial. If you have a spreadsheet cell that uses the standard deviation of a set of cells as an input - regardless of how many cores you have you must wait for the standard deviation to be computed before you can start the next set.

Sometimes a "properly threaded" application is a "single core serial" application, because each bit of the application depends on an earlier calculation - and it cannot be sped up by adding cores. (Or, a huge amount of programming complexity can be added that produces a barely noticeable improvement in the overal speed.)

All very true AidenShaw. The 3D applications in my arsenal that I alluded to (Maya, modo, Mudbox, etc.) certainly do not completely escape the problems you speak of. Many modeling operations conducted in 3D are still (frustratingly) in the land of single threads.
Dynamics and fluid simulations are also notoriously difficult to multi-thread (if possible at all). In that case a faster single core will indeed outperform a meatiest of multi-core systems.
Rendering on the other hand scales very nicely indeed, which is where a Mac Pro shines brightly in the glossy face of an iMac.
 

Hellhammer

Moderator emeritus
Dec 10, 2008
22,164
582
Finland
Quite the usual - MR is misleading people. It's only the fastest Mac MacWorld has tested. That does not mean that it's the fastest Mac that has ever been manufactured, especially when considering all aspects.

2.93GHz 12-core Mac Pro with 3 SSDs in RAID 0, NVIDIA Quadro 4000, 96GB RAM etc. would bash this iMac in everything else besides single-threaded app performance.
 

iSee

macrumors 68040
Oct 25, 2004
3,539
272
The iMac vs. Mac Pro comparison is pretty pointless.

The iMac has an SSD in it and the Mac Pro doesn't.

Is it really notable that the file copy test is faster on an SSD than a HDD?

C'mon Mac Rumors, *think*
 

RedTomato

macrumors 601
Mar 4, 2005
4,155
442
.. London ..
Wirelessly posted (This odd thing that I hold: Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_2_1 like Mac OS X; en-gb) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8C148 Safari/6533.18.5)

Huh, my 4 year old 2007 white MacBook can get pretty near this video, and it doesnt even have a SSD.

Just reboot, open all the apps, close them, start recording the video and reopen. All the apps will still be in memory and should reopen pretty fast.

Did i mention I have 4GB ram in my MacBook? That can hold several dozen apps with some tweaking (ie no content in the apps).

A slightly newer 2008 MacBook with 8GB should be able to reopen all the apps from the video at more or less the same speed. RAM's wonderful :)
 

likegadgets

macrumors 6502a
Jul 22, 2008
775
340
US
Kind of makes me wish I was patient enough to wait for the SSD model rather than just getting a standard HDD. Hopefully there will be some bootable Thunderbolt drives soon that will achieve the same effect.

I too have the 2011 27" 3.4GHZ without the factory SSD (have a 2TB). I am just waiting for the dust to settle on either external Thunderbolt or the best third party SSD, many of which claim to be much faster than the 256GB one Apple uses. Once the temperature issues, TRIM, cables, etc. are clearer, I will have an Apple authorized service center add a 512GB SSD to my iMac.
 

La Porta

macrumors regular
Dec 15, 2006
241
0
Why is it that they don't compare it to the mid-2010 i7 iMac? I'd like to see those numbers.
 

Cavepainter

macrumors regular
Apr 26, 2010
203
109
Los Angeles
Not impressed. Stick an SSD in the MacPro hexacore and try again.

Not a fair comparison either. That Mac Pro is $800 more expensive. (And that's not even including taxes.)

That article is terribly done though.... they should have emphasized something more along the lines that an imac with a SSD can beat a comparably priced Mac Pro. Not as good headlines, however.
 

Lesser Evets

macrumors 68040
Jan 7, 2006
3,527
1,294
Semi-ridiculous. However, despite the oft observed problems with the tests here, it is impressive nonetheless. Can't wait for prices on SSDs to drop, because I am sick of having slower drives that bottom-out yearly.

Geekbench rates that same iMac at half the processing power of the top Pro. I'd consider that a better benchmark. Again, it is impressive nonetheless. Being well over twice as fast as my tower impresses me greatly, since this 2006 tower is still extremely capable.
 

Digital Skunk

macrumors G3
Dec 23, 2006
8,097
923
In my imagination
Can't wait to see the next Mac Pro revision.

It's either going to be the maxed out 27" or a nicely spec'd Mac Pro. The next Mac Pro being an option ONLY because of it's expandability.
 

wiz329

macrumors 6502a
Apr 19, 2010
509
96
Not a fair comparison either. That Mac Pro is $800 more expensive. (And that's not even including taxes.)

That article is terribly done though.... they should have emphasized something more along the lines that an imac with a SSD can beat a comparably priced Mac Pro. Not as good headlines, however.

Price isn't in the picture at all. You can "fairly" compare an imac and a mac pro, even if the price points are completely different. But if you're going to, you should at least put the same drive in both of them. Or, if you're looking to find the "fastest mac ever", at least put the best BTO configs possible in both of them.
 

TimeArrow

macrumors member
Mar 2, 2011
39
0
Looking forward to getting mine. My last Mac was a 24" 2.16ghz core2duo iMac so this beast sounds like it'll run just a little bit quicker ;)

Agree. By fast, it's simply mean the launch time brought by SSD. Apple does NOT use the best SSD, so they compare SSD with HDD. For those who have no patience waiting for a SSD version, just buy an HDD one and upgrade the hard drive yourself. Cheaper and even faster.
 

DouglasDolde

macrumors newbie
Jun 15, 2011
4
0
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_1 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8G4 Safari/6533.18.5)

I'd rather buy a better monitor for my Nehalem Mac Pro, I don't need the speed and glossy displays bother me with all the reflections
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.