Mac Pro GeekBench-Whats your score?

[h9826790]

Any idea how I would be impacted by that in the real world?

I'd try both 3x8G 1333MHz (triple channel) and 4x8G 1066MHz (Dual channel) config.

When running GeekBench, yes, you will see a 25% (or more) performance difference for the RAM test. But overall, it's only about 5% difference for the whole machine's score.

I made a 10G RAM disk in both config, and run the Blackmagic disk speed test. There is about 7% difference (3000MB/s vs 2800MB/s) in both read and write speed.

Cinebench, Photoshop, and handbrake speed test, almost no difference (triple channel 1333MHz config always run a little bit better, but only 1-2% improvement). Also, can't feel any difference for gaming, or other normal use.

Therefore, I will say there may be <3% performance difference in most real world application (unless some software is extremely sensitive to the memory bandwidth), but virtually no difference.

My personal opinion, if you want both best speed and max out the RAM size, 3x16G is your best choice. However, if someone like me have 4 smaller RAM size sticks in hand, I prefer to use all 4 slots. I can't see any reason to sacrifice 25% of RAM size for only 3% performance in some applications. Also, if the machine eventually running out of RAM in the 24G config. I am sure the performance penalty will be more than 3%.

 

[flowrider]

Any idea how I would be impacted by that in the real world? I have read that the speed difference is 25% on paper/theory, more like 2/3% in practice

Down the page under "ram rules"

(http://pindelski.org/Photography/2013/07/08/mac-pro-2009-part-tbd/)

Does that impact single thread speed? Do you think I'd even notice?

IMHO, not much.

Lou

 

[CASLondon]

thanks for the feedback.

Since it seems the high end x5680/90 chips aren't coming down in price any time obviously soon, I think I might go for it. I like the idea of lower power draw for that much speed on these x5679s

 

[bxs]

Upgraded MP4,1 (early 2009) to MP5,1 with 2x 2.93 X5670s

I recently upgraded my early 2009 MP4,1 with 8core running 2x 2.26 GHz to a MP5,1 with 12cores running 2x 2.93 GHz X5670s. I also installed ATM Radeon 7950 and OWC's 960GB PCIe SSD and upgraded the 1066MHz 32GB RAM to OWC's 1333MHz 64GB RAM (8x 8GB sticks).

The Geekbench score (64-bit) after the upgrade displayed a solid 27500 consistently. I'm very pleased with this at this time.

This was the 'before' upgrade score

 

[Antoni Nygaard]

Mac Pro 2,1

3.0Ghz 8-core
8Gb Ram
AMD 6870
60Gb SSD

i get around 11500-11700

 

[m4v3r1ck]

Perhaps a noob question... How to 'prepare' a Mac before running GeekBench e.g. killing running (start-up) applications and/or processes?

 

[bxs]

Perhaps a noob question... How to 'prepare' a Mac before running GeekBench e.g. killing running (start-up) applications and/or processes?

I would ensure indexing/Spotlight, Time Machine, scheduled cloning and all Applications are not active. Normally, after a Restart the system will do some indexing (multiple mds processes active), so wait for this to complete. Wait for fans to spin down. Check with Activity Monitor or Terminal's top command to wait for CPU activity to die down.

I would also run Geekbench 3 times and take the average of the 3 scores. Make sure the same 32-bit or 64-bit modes are consistent for the 3 runs as well.

 

[m4v3r1ck]

I would ensure indexing/Spotlight, Time Machine, scheduled cloning and all Applications are not active. Normally, after a Restart the system will do some indexing (multiple mds processes active), so wait for this to complete. Wait for fans to spin down. Check with Activity Monitor or Terminal's top command to wait for CPU activity to die down.

I would also run Geekbench 3 times and take the average of the 3 scores. Make sure the same 32-bit or 64-bit modes are consistent for the 3 runs as well.

Thank you very much! Let's see what my MP 3.1 is made of.

 

[sonicbrute]

Mac Pro 2006 pimped for 10.9.0 SSD 2 quad core 3ghz



geekbench 10747

 

[m4v3r1ck]

 

[m4v3r1ck]

Upgraded to MP 5.1 (2012) DODECA 2.66 Ghz.

Again strange that GB3 sees it as a mid 2010

~ Cheers

 

[englishman]

http://browser.primatelabs.com/geekbench3/588789

Mac Mini

SC - 2552 - compared to your MP!
MC - 5314

 

[flowrider]

Upgraded to MP 5.1 (2012) DODECA 2.66 Ghz.

Again strange that GB3 sees it as a mid 2010

~ Cheers

It's really not strange. GB3 reads the processor, number of cores and speed and determines the model description from that information. Both the 2010 and 2012 Mac Pros were available with the X5650s. So, it appears they recognize the earlier model information. I have a 2010 5,1 and have dual W5590s. Since that was never an Apple option, GB3 only recognizes it as a MacPro5,1 without any year designation.

Lou

 

[m4v3r1ck]

It's really not strange. GB3 reads the processor, number of cores and speed and determines the model description from that information. Both the 2010 and 2012 Mac Pros were available with the X5650s. So, it appears they recognize the earlier model information. I have a 2010 5,1 and have dual W5590s. Since that was never an Apple option, GB3 only recognizes it as a MacPro5,1 without any year designation.

Lou

Thanks for clearing that out Lou!

 

[Ludacrisvp]

MP3,1
GB3 score post processor upgrade:

GB2 score stock processor and less RAM a few months after purchase:

 

[lexR]

2009 4.1 8 core@2.66 upgraded to 5.1 12 core@2.93

Just recently upgraded my MP 2009 with 2 X5670's and very happy with the results.

and the results are from my point of view are astonishing as i also upgraded my SATA II SSD to a SATA III PCI-e mounted SSD and installed a GTX 770 so the Mac Pro certainly does rip along nicely now (not that it didn't before)

 

[flowrider]

Resurrecting this to add that Geekbench issued V3.2 this morning.

http://www.primatelabs.com/geekbench/

And to show the results of my Upgrade path from dual E5620s to W5590s to X5677s. I stayed with 8 fast cores because for what I do, I felt that 12 cores would be a waste. Also, I believed that less cores running at the same speed would run cooler with less of a power draw.

Lou

 

[Thomaspin]

From an X5690 upgrade on a customer's machine, dual CPU 2009 Mac Pro:

 

[CASLondon]

Naive question here - do other factors besides CPU play a part in this score here? i.e read/write speeds on the boot disk, gpu, etc?

Or is this just a raw processor test?

 

[flowrider]

^^^^Processor and RAM. Disk I/O or GPU does not effect the score of Geekbench.

Lou

 

[Ironjer]

Here we go.

 

[Synchro3]

Upgrade from Xeon W3540 to W3690

Single Core old W3540 2.93 GHZ
2'126
Multi Core old W3540 2.93 GHZ
8'036

http://browser.primatelabs.com/geekbench3/670044

Single Core new W3690 3.46 GHZ
2'656 (+ 25%)
Multi Core new W3690 3.46 GHZ
14'664 (+ 82%)

http://browser.primatelabs.com/geekbench3/727972

 

[maplingstorie]

Upgraded from E5-1650v2 to E5-2697v2.

 

[rueyloon]

Upgraded from E5-1650v2 to E5-2697v2.

You DIY the upgrade ? time for me to show my hand too.

 

[rueyloon]

Upgraded from E5-1650v2 to E5-2696v2.