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

nMP/new Mac Pro/Mac Pro 6,1 Reviews [MERGED]

Cubemmal said:
You seriously think Apple would design it that way?
Click to expand...

It is designed that way. All the modern major components clock down when stressed with heat that can't be dissipated.

The single fan and shared cores is all about trying to leverage the abilities to dynamically adjust and a heavy bet that all won't be busy at the same time. When do find loads that will heavily stress the machine will slowly give. It simply doesn't have resources to plow ahead with brute force.
 
deconstruct60 said:
It is designed that way. All the modern major components clock down when stressed with heat that can't be dissipated.

The single fan and shared cores is all about trying to leverage the abilities to dynamically adjust and a heavy bet that all won't be busy at the same time. When do find loads that will heavily stress the machine will slowly give. It simply doesn't have resources to plow ahead with brute force.
Click to expand...

Well maybe we're talking at cross purposes. My expectation, and initial reports support, that the nMP will be silent under full load. Because of the enormous 130W power budget the chip will spank any i7 you care to compare it to, per this article

http://www.marco.org/2013/11/26/new-mac-pro-cpus
 
slughead said:
The German article already said it was noisy under full load (LuxMark).
Click to expand...

Yeah and that article made no sense, and was a machine translation. When I converted their Sones number to dB it was over 100.

Everybody else has said that even under extreme load you can't hear it unless you put your ear close, so ignore the German article.
 
Hey all,
This my first post here but I was hoping to clarify a few things.

I can only share information based on the stock 6 core second tier system as its all I've used (and 64gb of ram), but hopefully that will give you ball park information.

Running Resolve 10 I was getting butter smooth performance on a 24 fps grade on a D500 equipped Mac Pro running 4k dpx scanned film frames off of the internal ssd with a full set of grading going for a small commercial project.

Things went off without a hitch and adding a Pegasus raid as the source drive to test thunderbolt performance also resulted in similar output. I didn't see any issues.

Resolves also uses the first gpu for GUI and the second for grading as needed.

Mari 2.5 test case:

Computer 1: mid 2013 iMac 27" fully loaded with 32gb ram, top tier gpu, etc on a fusion drive system

Computer 2: 64 gb second tier six core new Mac Pro

Scene contained 5 million faces over 12 objects running about 40 8k textures real time. Total file size for Mari archive before decompression is about 4 gb.


Computer 1 ran scene at 2fps with everything turned on.

Computer 2 ran scene at 12fps withe very thing turned on.

Mari also recognizes both gpus but uses only one by default. Changing it to dual gpu doesn't seem to affect anything at the moment and I'm assuming that the code isn't implemented yet but will be in the future builds.

With main character asset and ground props enabled:

Computer 1 ran scene at 6 fps

Computer 2 ran scene at 21fps.

The software that is built leveraging the Multiple gpus in the new Mac Pro shows a distinct speed difference from what you're all used to. For professional applications in the post and VFX world this machine is phenomenal.

I'd go into vray rt and multicore statistics, but I don't have the figures handy at the moment.

Nuke screams, especially Kronos and ocula!


Maya, media composer 6.5.3 and 7, and other avid and Autodesk software doesn't take advantage of the full multicolor power at the moment in osx (but new builds are in the works), however they all run wonderfully in boot camp.

The performance on the Autodesk and avid apps are similar to the iMac due to the lack of multicore and multiple gpu support.

Hopefully this helps.
(Apologies for any typing errors, using an iPad at the moment)
 
Last edited:
Mgmx said:
Hey all,
This my first post here but I was hoping to clarify a few things.

I can only share information based on the stock 6 core second tier system (and 64gb of ram), but hopefully that will give you ball park information.

Running Resolve 10 I was getting 24 fps on a D500 equipped Mac Pro running 4k dpx scanned film frames off of the internal ssd with a full set of grading going for a small commercial project.

Things went off without a hitch and adding a Pegasus raid as the data repository also resulted in similar workload.

Mari 2.5 test case:

Computer 1: mid 2013 iMac 27" fully loaded with 32gb ram, top tier gpu, etc on a fusion drive system

Computer 2: 64 gb second tier six core new Mac Pro

Scene contained 5 million faces over 12 objects running about 40 8k textures real time. Total file size for Mari archive before decompression is about 4 gb.


Computer 1 ran scene at 2fps with everything turned on.

Computer 2 ran scene at 12fps withe very thing turned on

With main character asset and ground props enabled:

Computer 1 ran scene at 6 fps

Computer 2 ran scene at 21fps.

The software that is built leveraging the Multiple gpus in the new Mac Pro shows a distinct speed difference from what you're all used to. For professional applications in the post and VFX world this machine is phenomenal.

I'd go into vray rt and multicore statistics, but I don't have the figures handy at the moment.

Nuke screams, especially Kronos and ocula!


Maya, media composer 6.5.3 and 7, and other avid and Autodesk software doesn't take advantage of the full multicolor power at the moment in osx (but new builds are in the works), however they all run wonderfully in boot camp.

The performance on the Autodesk and avid apps are similar to the iMac due to the lack of multicore and multiple gpu support.

Hopefully this helps.
(Apologies for any typing errors, using an iPad at the moment)
Click to expand...
Thank you. Finally real numbers.
 
Zbrush performance is moot as it is a non gpu accelerated single core application.

The windows build isn't even natively 64 bit but instead uses a few patches to enable higher amounts of ram.

You would notice the same performance in Zbrush as you would on any PCI ssd equipped machine.

Vray, however, loves this new Mac Pro. And the opencl performance is fantastic, especially in windows.

Typically I see 10-20% faster performance on ANY mac when running boot camp as windows seems to have better written driver support for most rendering and maya work.

Octane I haven't tried yet, though I don't really see it or keyshot as production renderers like vray, renderman, or Arnold. They're great for still frame and concept work but I wouldn't run a feature or television show through them.

That being said, I'm not sure how they perform on the new Mac Pro, but I'm assuming octane won't run at all because the last time I used it the software was CUDA only.

This machine is more designed for High end production work than Photoshop or consumer level work and is a godsend as it's easy to move from job to job compared to the older mac pros.

I know there has been quite a bit of discussion about upgrade life, but most of these things will pay for themselves in on site rental (with artists) in under 3 months time. Most of the VFX and production people I know tend to buy new machines every year or so in general and if these machines can eke out 2 or more years of life then they'll be even better for that.

Best,
M
 
Last edited:
Mgmx said:
Zbrush performance is moot as it is a non gpu accelerated single core application.

The windows build isn't even natively 64 bit but instead uses a few patches to enable higher amounts of ram.

You would notice the same performance in Zbrush as you would on any PCI ssd equipped machine.

Vray, however, loves this new Mac Pro. And the opencl performance is fantastic, especially in windows.

Typically I see 10-20% faster performance on ANY mac when running boot camp as windows seems to have better written driver support for most rendering and maya work.

Octane I haven't tried yet, though I don't really see it or keyshot as production renderers like vray, render an, or Arnold. They're great for still frame and concept work but I wouldn't run a feature or television show through them.

That being said, I'm not sure how they perform on the new Mac Pro, but I'm assuming octane won't run at as the last time I used it the software was CUDA only.

Best,
M
Click to expand...

Would it be possible for you to run like some games on it. For instance diablo 3 ? There's a free d3 trial on blizzard's website.
 
I don't really run games on this type of hardware but I seem to recall benchmarks for diablo already being available on other reviews. I wouldn't expect it to be as fast as the iMac mostly because nvidia driver support and physx support is far better than amd from what I recall.

Again, the aim of the system isn't really for gaming performance through directX.

Edit: As an aside, things like cinebench and geekbench have never been accurate estimates of a machine's performance in the field, the constant request for those scores as a benchmark has always baffled me. They're never really stressing a system the way production does.

I always just run my most recent project through hardware and compare raw figures (time rendered, etc) as well as things such as the 'feel' in interactivity in a similar pipeline.


When I get back from vacation and the office opens I can install it but by then I'm sure there will be plenty of the machines in the wild.
 
Last edited:
Mgmx said:
I don't really run games on this type of hardware but I seem to recall benchmarks for diablo already being available on other reviews. I wouldn't expect it to be as fast as the iMac mostly because nvidia driver support and physx support is far better than amd from what I recall.

Again, the aim of the system isn't really for gaming performance through directX.

Edit: As an aside, things like cinebench and geekbench have never been accurate estimates of a machine's performance in the field, the constant request for those scores as a benchmark has always baffled me. They're never really stressing a system the way production does.

I always just run my most recent project through hardware and compare raw figures (time rendered, etc) as well as things such as the 'feel' in interactivity in a similar pipeline.


When I get back from vacation and the office opens I can install it but by then I'm sure there will be plenty of the machines in the wild.
Click to expand...

We have the gaming benchmarks for the d700 but we don't have the d300 and d500.

There is a mac version of d3. thats the version i am very interested in seeing the performance

Another big question is if there is crossfire on windows enabled.
 
Ah. Understood about the benchmarks.
I would have to check the mac build.

As to the crossfire support.. I honestly am not sure.

Mari, resolve and the like tend to use the GPUs outside of the SLI type of environment so I never thought to check.

I didn't have too much time with it, and pulled windows off after testing a few apps. I'll reinstall when back in town.
 
Last edited:
Mgmx said:
Zbrush performance is moot as it is a non gpu accelerated single core application.
Click to expand...

I don't know how you came to this but it's wrong. Zbrush isn't single core, it finely used multiple cores. It just doesn't scale linear like a raytracer. The speedup on my quad early 2008 is *2,34


The windows build isn't even natively 64 bit but instead uses a few patches to enable higher amounts of ram.
Click to expand...

Zbrush is 32 on both platforms. Zbrush 5 will bring 64bit.
 
Dranix,
You're correct, my mistake.

I know they're building for 64 (and have been for awhile now) but it involves a big rewrite of the code.

Anyhow, the point I was making was that Zbrush is not software that would perform drastically different on the new Mac Pro hardware.
 
Last edited:
Mgmx said:
Hey all,
This my first post here but I was hoping to clarify a few things.

I can only share information based on the stock 6 core second tier system as its all I've used (and 64gb of ram), but hopefully that will give you ball park information.

Running Resolve 10 I was getting butter smooth performance on a 24 fps grade on a D500 equipped Mac Pro running 4k dpx scanned film frames off of the internal ssd with a full set of grading going for a small commercial project.

Things went off without a hitch and adding a Pegasus raid as the source drive to test thunderbolt performance also resulted in similar output. I didn't see any issues.

Resolves also uses the first gpu for GUI and the second for grading as needed.

Mari 2.5 test case:

Computer 1: mid 2013 iMac 27" fully loaded with 32gb ram, top tier gpu, etc on a fusion drive system

Computer 2: 64 gb second tier six core new Mac Pro

Scene contained 5 million faces over 12 objects running about 40 8k textures real time. Total file size for Mari archive before decompression is about 4 gb.


Computer 1 ran scene at 2fps with everything turned on.

Computer 2 ran scene at 12fps withe very thing turned on.

Mari also recognizes both gpus but uses only one by default. Changing it to dual gpu doesn't seem to affect anything at the moment and I'm assuming that the code isn't implemented yet but will be in the future builds.

With main character asset and ground props enabled:

Computer 1 ran scene at 6 fps

Computer 2 ran scene at 21fps.

The software that is built leveraging the Multiple gpus in the new Mac Pro shows a distinct speed difference from what you're all used to. For professional applications in the post and VFX world this machine is phenomenal.

I'd go into vray rt and multicore statistics, but I don't have the figures handy at the moment.

Nuke screams, especially Kronos and ocula!


Maya, media composer 6.5.3 and 7, and other avid and Autodesk software doesn't take advantage of the full multicolor power at the moment in osx (but new builds are in the works), however they all run wonderfully in boot camp.

The performance on the Autodesk and avid apps are similar to the iMac due to the lack of multicore and multiple gpu support.

Hopefully this helps.
(Apologies for any typing errors, using an iPad at the moment)
Click to expand...

Finally! Now these are comparisons i think alot of us where waiting for.

Thank you Mgmx
 
That about wraps it up. Surprise surprise, Apple said it would be a quiet performer, and it's a quiet performer.
 
theSeb said:
That's not what the article said.
Click to expand...
And that isn't true either because it is somewhere in the middle ;) This is what the German article says:

Während des ersten Tags war er bei den allermeisten Benchmarks schlimmstenfalls mit einem leisen Säuseln zu hören (0,1 bis 0,2 Sone), beim längeren Spielen mit Dirt II haben wir immer noch wenig unangenehme 0,5 Sone gemessen. Allein beim Herumprobieren mit dem neuen Final Cut Pro und mit dem kostenlosen Open-CL-Test Luxmark hörten wir den Lüfter, der die warme Luft durch den dreieckigen Kühlkörper nach oben abzieht, deutlich mit recht lauten 2,7 Sone.
Click to expand...
With most benchmarks and even with Dirt II they didn't quite hear the fan (aka need to put ear very close to the fan) but they could definitely and noticeably hear the fan when running FCP and LuxMark. They measured 2.7 Sone and defined that as being loud. This, however, is exactly the same as what other reviews have concluded.

Wikipedia has a nice article about Sone. There they have this nice table of various Sone levels. If the amount of Sone is 1~4 it is defined as "people talking 1m away". For a fan that indeed is loud. The others are around something they call "very calm room".
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back