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

Is the new Mac Pro a Failure for traditional Mac Creative and Professional customers


  • Total voters
    417
Status
Not open for further replies.
Factory OC is OC.

Nominal clocks for GTX 970 are 1050, yours have 1117 MHz, according to the stats of Luxmark.

I counted Compute performances based on scores and the clocks. Mac Pro with Dual D700 which one is on 150 MHz and the second is on 850 has 4.1 TFLOPs of compute performance.

Dual GTX 970 on 1177 MHz has 7.8 TFLOPs. Scores in Luxmark now have sense.
 
Last edited:
Factory OC is OC.

Nominal clocks for GTX 970 are 1050, your have 1117 MHz, according to the stats of Luxmark.
It's factory reference spec from the manufacturer. My cards are reference cards. FYI boost is similar to the turbo in a Core series processor by Intel. I don't see you minimizing Intel, but your all knowing Apple hasn't told you otherwise yet. It's controlled by the driver and NVidia's software.

I guess this is more of that "you're doing it wrong" mentality.
 
So, tell me why Mac Pro 5.1 with 3.33 GHz Six core CPU, 24 GB of RAM, and R9 280X is 3.5 times faster in Final Cut Pro X using OpenCL, than GTX 970 in the same job and the same app using the same thing?

This is my experience of working with the GPUs.

P.S. I admit Im baffled by the results of Lexmark. Because they contradict my experience in work with both brands in OpenCL.
 
So, tell me why Mac Pro 5.1 with 3.33 GHz Six core CPU, 24 GB of RAM, and R9 280X is 3.5 times faster in Final Cut Pro X using OpenCL, than GTX 970 in the same job and the same app using the same thing?

This is my experience of working with the GPUs.

P.S. I admit Im baffled by the results of Luxmark.

I don't have a Mac Pro. I can't say for sure.

If I had to guess it could be a driver issue or maybe a CPU bottleneck. It might also be something in the Final Cut Pro X software that is the culprit.

P.S. I admit Im baffled by the results of Lexmark. Because they contradict my experience in work with both brands in OpenCL.

:D I guess you just took the red pill.
 
Last edited:
FCPX uses only OpenCL. Nothing else, so it will run the faster on hardware that is the fastest in it. It is as simple as it can be. By looking how people are getting results in BruceX test, on Hacks with Haswell i7 4770K and R9 280X faster than people with Xeon's and Titan X there is nothing strange at all. The problem is with the way FCPX offloads CPU, RAM and hard drive and puts all the hard work on GPU. If the hardware is capable of managing itself(Asynchronous Compute) it will be working properly. If not: it won't. That may be a reason why we see the results in Lexmark like they are. With no Asynchronous compute. FCPX uses it heavily.

P.S. I want to understand what is going on with these benchmarks.
 
FCPX uses only OpenCL. Nothing else, so it will run the faster on hardware that is the fastest in it. It is as simple as it can be. By looking how people are getting results in BruceX test, on Hacks with Haswell i7 4770K and R9 280X faster than people with Xeon's and Titan X there is nothing strange at all. The problem is with the way FCPX offloads CPU, RAM and hard drive and puts all the hard work on GPU. If the hardware is capable of managing itself(Asynchronous Compute) it will be working properly. If not: it won't. That may be a reason why we see the results in Lexmark like they are. With no Asynchronous compute. FCPX uses it heavily.

P.S. I want to understand what is going on with these benchmarks.
It sounds like a CPU bottleneck or OSX not doing a good job of using the hardware it has.

How much does the CPU and memory take into account for the benchmark even when in GPU only mode? Keep in mind I'm sporting DDR4.
 
What all your holy benchmarks are missing, in favor of AMD and their mighty OpenCL which no one uses except FCP X, which again, makes maybe 1-3% of the pro market, is that all NLE use a combination of CPU and GPU, not strictly one component.
 
What all your holy benchmarks are missing, in favor of AMD and their mighty OpenCL which no one uses except FCP X, which again, makes maybe 1-3% of the pro market, is that all NLE use a combination of CPU and GPU, not strictly one component.
I debunked that with the benchmarks I posted. The bench marks show that a Core i7 Extreme 5930K with two GTX 970s in SLI beats AMD in OS X in all Open CL tests against a 3.5GHz Xeon E5 V2 with D700 GPUs. :D There was not one test where the Mac Pro was faster. There where a few where performance was close but still slower.

If the next Mac Pro has updated processors and nVidia cards I think it would be a major upgrade. If they don't want Core i7 Extreme in a Mac Pro revive the Power Mac name. I personally would love to see a headless Mac with a Core i7 Extreme processor and SLI. I would snatch one up day 1 gladly, especially if Apple also releases an updated cinema display with the same monitor as the Retina iMac even if it needs a dual link.
 
Last edited:
The bench marks show that a Core i7 Extreme 5930K with two GTX 970s in SLI beats AMD in OS X in all Open CL tests against a 3.5GHz Xeon E5 V2 with D700 GPUs.

SLI & Crossfire is mostly for gaming, I don't think most professional applications can use SLI and in fact may slow down performance.
 
SLI is mostly for gaming, I don't think most professional applications can use SLI and in fact may slow down performance.
The Open CL test was using both GPUs. That seems to me like it can be used. If Apple made a dual AMD machine it can make a dual nVidia machine.

I'm honestly not too sure what the applications are using. I know 3D modeling is mostly single threaded in processing the calculations but in 3D Studio Max I believe they use DirectX. If it uses DirectX SLI can be forced on.
 
Premiere Pro rarely uses the 2nd GPU, mostly für Red footage and encoding. Resolve for example uses both GPU more frequently, but only in the "Studio" version. In the "normal" ie "free" version, only with the trashcan it uses both D300/500/700s.
 
Premiere Pro rarely uses the 2nd GPU, mostly für Red footage and encoding. Resolve for example uses both GPU more frequently, but only in the "Studio" version. In the "normal" ie "free" version, only with the trashcan it uses both D300/500/700s.
Hopefully is 2 GPU systems become more common the technology will be better used outside of games. :D
 
Right now I don't see the big benefit in video post (from an editing standpoint, VFX and color are two separate entities here), since CPUs are the bottlenecks most of the time when dealing with >4K footage and other essentials tasks for me. Hence why I always preferred dual CPU system like my 5,1.
 
Right now I don't see the big benefit in video post (from an editing standpoint, VFX and color are two separate entities here), since CPUs are the bottlenecks most of the time when dealing with >4K footage and other essentials tasks for me. Hence why I always preferred dual CPU system like my 5,1.
In 3D modeling there would be a benefit.

Dual CPU systems are awesome. I'm honestly surprised that Apple would drop them.

Edit:

I went ahead and published a few results onto the web site. I got a better result just now. http://www.luxmark.info/node/1853
 
Last edited:
There was not one test where the Mac Pro was faster
faster at what? what are you all talking about? are you saying i can work faster if i get a windows machine with nvidia?

because if so, i think i'll go buy one tomorrow.

if you're talking about 'faster' that has nothing to do with working and/or accomplishing tasks then, ok.. but please clarify.
 
faster at what? what are you all talking about? are you saying i can work faster if i get a windows machine with nvidia?

because if so, i think i'll go buy one tomorrow.

if you're talking about 'faster' that has nothing to do with working and/or accomplishing tasks then, ok.. but please clarify.
Seriously ...

Not taking the bate.
 
  • Like
Reactions: AidenShaw
Seriously ...

Not taking the bate.
it's not bait.
it's a real question.. maybe a rhetorical question but still real nonetheless.

-----
a faster computer will only help speed things up if your workflow involves a lot of waiting.. (and even then, only certain waiting scenarios will make use of a faster computer).

what the heck are people doing where they're just sitting around waiting all the time?
 
Seriously ...

Not taking the bait.

Wise choice.

If you point out that cMP can house drives that are 4x faster than nMP, you will be told that drive speed doesn't matter.

If you prove that more CPU cores are important you will be told that only rubes use CPU cores when render farms are a dime a dozen.

If you point out that the 4 (yep, 4) year old GPUs are a joke you will be told that real Pros don't use GPUs for some reason or other.

The Gibberati will straw-man and goal-post move indefinitely. Like the movie "The Never Ending Story" it has become the "The Never Ending Excuses".

Apple apologists are a motivated bunch, logic and reason can't be used on them. When cornered they will revert to glorifying shiny polished surfaces, or friendly pix of the design team. Or claim that GPU updates are coming since Apple used Security Torx screws to lock them to the heatsink. They are, after all, "the most user friendly standard screws available".

Trying to debate people who will never admit obvious defeat is very ungratifying.
 
it's not bait.
it's a real question.. maybe a rhetorical question but still real nonetheless.

-----
a faster computer will only help speed things up if your workflow involves a lot of waiting.. (and even then, only certain waiting scenarios will make use of a faster computer).

what the heck are people doing where they're just sitting around waiting all the time?
Having used both platforms in a professional capacity since going to Savannah College of Art & Design I can say that switching between Windows and OS X is trivial. The vast majority of the software is available on both platforms except of course for the first party software like Final Cut and Visual Studio.

In the real world if whatever machine stays out of my way and allows me to manipulate my work as best as possible and then renders/encodes/compiles it in the least amount of time is the fastest machine. I have shown that on Windows nVidia cards outperform the Mac. Setting aside the what ifs and looking at this in terms of a slower machine and OS X and a faster machine with Windows there is only one question. Can the machine run the software I need to use. In my case if it runs on Windows then my machine is faster in every sense including your question about work flow.

I don't know if you have ever worked on a XCode project and felt the thing chugging along and ever thought about how much you loath that IDE but I have.
 
Having used both platforms in a professional capacity since going to Savannah College of Art & Design I can say that switching between Windows and OS X is trivial. The vast majority of the software is available on both platforms except of course for the first party software like Final Cut and Visual Studio.

In the real world if whatever machine stays out of my way and allows me to manipulate my work as best as possible and then renders/encodes/compiles it in the least amount of time is the fastest machine. I have shown that on Windows nVidia cards outperform the Mac. Setting aside the what ifs and looking at this in terms of a slower machine and OS X and a faster machine with Windows there is only one question. Can the machine run the software I need to use. In my case if it runs on Windows then my machine is faster in every sense including your question about work flow.

I don't know if you have ever worked on a XCode project and felt the thing chugging along and ever thought about how much you loath that IDE but I have.

so i ask again-
faster at what?

you've given a complete non-answer other than the complete obvious 'i need software that only runs on windows therefore a windows machine is faster in every sense'

if your benchmark battle post had any actual meaning to them, you'd easily be able to point out something like "i could draw a stick figure on windows in 1 minute but if i do the same thing on mac, it will take me 2 minutes"..

problem is, those numbers you keep posting as example of macs being crushed have about zero meaning.. like-- do you even know what those numbers mean? 15000 to 10000 ? cool
15000 what?
 
so i ask again-
faster at what?

you've given a complete non-answer other than the complete obvious 'i need software that only runs on windows therefore a windows machine is faster in every sense'

if your benchmark battle post had any actual meaning to them, you'd easily be able to point out something like "i could draw a stick figure on windows in 1 minute but if i do the same thing on mac, it will take me 2 minutes"..

problem is, those numbers you keep posting as example of macs being crushed have about zero meaning.. like-- do you even know what those numbers mean? 15000 to 10000 ? cool
15000 what?
Faster at everything. A all encompassing everything.
 
Status
Not open for further replies.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.