Geekbench on 4-core Mac Pro

[Shawn Miller]

[Michael Garber] “The local Apple Store got a 4-core Mac Pro today. Had a chance to check it out as most of the reviews I’ve seen have been on the 6, 8, and 12-core models. I ran Geekbench in 64-bit mode and Cinebench on the system. This was on the low-end 4-core with the D300.”

What was the Cinebench score?

Shawn

[Bernard Newnham]

As that didn’t mean much to me, I downloaded the trial Geekbench (32 bit only).

My Core i5-2500 3.3ghz CPU ran –

2890 single core
9240 dual core

Cinebench ran –

CPU 448cb
OpenGL 81.95fps

– the cars racing around the city very smoothly on my GeForce GTX 460 PCi. More fps than the Macpro above, don’t know why.

Not sure what all that means, but maybe it will give suitable contrast for others. I have the money to upgrade my gear, but currently I’m not rushing – though I don’t run Cinema 4d for a living. I don’t find great problems with editing 1080i.

More important to me these days is that my 60Mb/s connection which allowed me to download Cinebench at 109MB in a few seconds. Virgin are vaguely promising faster for this year.

Bernie

[Gary Huff]

[Michael Garber] “The local Apple Store got a 4-core Mac Pro today. Had a chance to check it out as most of the reviews I’ve seen have been on the 6, 8, and 12-core models. I ran Geekbench in 64-bit mode and Cinebench on the system. This was on the low-end 4-core with the D300.”

Did you have to install it yourself? I’m thinking about putting the current GPU beta of RedCineX on the display model here (if they have one) and running my typical test of a particular REDcode clip render from that. That’s the metric most important to me.

[Daniel Frome]

[Michael Garber] ”
Geekbench single-core: 3609
Geekbench multi-core: 14452″

For comparison, my retina macbook pro, i7 quad core 2.6ghz:

Single Core: 2960 (Quad Mac Pro is 14% faster)
Multi Core: 11528 (Quad Mac Pro is 21% faster)

Some other food for thought: The Mac Pro’s CPU is mathematically 30% faster by pure numbers. Interesting law of diminishing returns that only delivers 14-21% increase in performance on a cpu test… probably ever lower in real-world usage. This has nothing to do with Apple, it’s just technology.

[Michael Garber]

I didn’t get the CPU score when I was there. The FPS was 73.

Michael Garber
5th Wall – a post production company
Blog: GARBERSHOP
My Moviola Webinar on Cutting News in FCP X

[Craig Seeman]

Barefeats test which seems relevant if you’re considering speed between the two.
https://barefeats.com/tube03.html

Seems the iMac is very competitive until one sees the LuxMark test which is multi GPU aware. The dual D300s are a big advantage and this is against the iMac BTO 780M

[Michael Garber]

Yeah, I downloaded and installed it. A business specialist came over and we both bonded over it. But alas, I couldn’t leverage our new friendship to get a better discount.

They actually still had FCP 10.0.9 on there. This is because the corporate departments for the disk images are closed for the holidays. So bring a copy of 10.1 with you as well.

Michael Garber
5th Wall – a post production company
Blog: GARBERSHOP
My Moviola Webinar on Cutting News in FCP X

[Pierre Jasmin]

Actually not clear at first glance the test compares against a single Radeon 7970 on the 2010 Mac Pro(not an original card).
RE:LuxMark here across different cards outside that machine.

https://www.tomshardware.com/reviews/radeon-r9-290x-hawaii-review,3650-34.html

Notice how this compute benchmark almost linearly grows as the GPU number increases (e.g. 7990 vs 7970, 2 vs 1 GPU test, it does not really double here as they clock down the second GPU on the card for electrical reasons)- and this means that the D300 is roughly a FirePro 7000 on that test.

This is a compute only test, does not stress IO from RAM to VRAM and back etc…

Pierre

[Craig Seeman]

If only two 7970 or 7990 were possible.

[Rick Lang]

Craig Seeman:
“Seems the iMac is very competitive until one sees the LuxMark test which is multi GPU aware. The dual D300s are a big advantage and this is against the iMac BTO 780M.”

Although in the short term, the high-end iMac and high-end MacBook Pro Retina both look very good against the entry-level Mac Pro, during this year as more software is tuned to take advantage of the dual GPUs, the Max Pro will probably look much better for video work. The benchmark results are amazing and the current tuning of FCP X10.1 and Motion 5.1 mean that vendors will quickly respond to ensure their software remains competitive on the Mac Pro.

Rick Lang

iMac 27” 2.8GHz i7 16GB