Contribute
Register

Laughable FCPX performance despite decent specs

Status
Not open for further replies.
Joined
Oct 17, 2012
Messages
8
Motherboard
Mac OS X 10.9.4 + Windows 7
CPU
Intel i7 4790K
Graphics
Asus GTX 780 Ti
Hi everyone,

I've recently started to feel like my Hac Mac is getting slower in terms of graphics performance. Especially, since the last few Nvidia Web Drivers updates for my GTX 780Ti. The slowness is mostly seen in the apps that benefit from OpenCL architecture, just like Final Cut Pro X, which I use for 4K film editing.

For total clarity, my specs are:

Intel 4790K (running at stock speeds, with working PM), Gigabyte Z87-UD3H, 16 GB RAM 1600 MHz DDR3, NVIDIA GeForce GTX 780Ti and Samsung 840 EVO SSD. My SMBIOS is set for iMac14,2

Everything in my hackintosh setup works pretty fine, except the desired performance in FCPX. I've recently took the famous FCPX benchmark — BruceX, and my results were even below the lowest expectations — around 63 seconds. I've seen some people on FCPX forums showing much better results with inferior hardware setup.

The question is — do you guys have any idea what can cause such performance decrease? Is my only way out of this is getting into Radeon game? Maybe I'm doing something wrong. Any feedback will be very appreciated. I will gladly provide any system info that's needed for a diagnosis.

Thanks in advance!
 
Hey,
I was just looking for informations.
I upgraded my Hackintosh's CPU to i7 6700k overclocked to 4,8Ghz.
I have a GTX 980 Ti.
Just ran the test tonight and I get 60s every time. I have much higher specs than yours for the same score...
I wonder if the problem comes from me or from last FCPX update...

Any ideas in the community?
 
Sierra gives you worse performance in FCPX but Nvidia drivers just don't cut the OpenCL mustard. If FCPX is your main objective bite the bullet and buy a RX480/580 8Gb.

I got 20 sec on my first test with my setup
 
Im getting average of 13.6 each time with the 1080ti. Some of my big 4k stuff I tested with the old 980ti and now with the 1080ti and saw tremendous improvements.

 
Im getting average of 13.6 each time with the 1080ti. Some of my big 4k stuff I tested with the old 980ti and now with the 1080ti and saw tremendous improvements.

Ok but do you mean that 980ti and 780ti are 2 cards specifically bad with FCP X compared to older and less powerful Nvidia graphic cards?
On the forums I see people with GTX 970 getting 20s, GTX 960, 26s...
Why would we get 60s with better cards and why would we get that same score? That's what I find weird :/
 
Ok but do you mean that 980ti and 780ti are 2 cards specifically bad with FCP X compared to older and less powerful Nvidia graphic cards?
On the forums I see people with GTX 970 getting 20s, GTX 960, 26s...
Why would we get 60s with better cards and why would we get that same score? That's what I find weird :/

First card I ever had was the 980ti. From last year. Friday I got my 1080ti and simply unplugged the 980ti and put in the new 1080ti. FCPX is Night and Day different. My master file outputs decreased in time tremendously. I do a lot of 4k 60fps and it handled that all great. My only complaint is still when reviewing the editing in the timeline I still get dropped frames and a lot of jitter. I can live with that. The master file plays fine. Looks great. I hope this helps. Be glad to answer anything that I can.
 
So, I just updated Sierra from 10.12.1 to 10.12.4 and installed the latest Nvidia Web drivers (the ones that support Pascal).
Results are there! Not crazy, but it takes now 45s instead of 60s.

I do a lot of 4k 60fps and it handled that all great. My only complaint is still when reviewing the editing in the timeline I still get dropped frames and a lot of jitter.
You should work with proxy.
 
So, I just updated Sierra from 10.12.1 to 10.12.4 and installed the latest Nvidia Web drivers (the ones that support Pascal).
Results are there! Not crazy, but it takes now 45s instead of 60s.


You should work with proxy.
I was honestly ready to send my 1080ti back after I ordered it. From reading all the reviews of it. But I'm glad I did try it. Brucex is a test and for benchmarking. But to go from 28-32 seconds to 13 or less. Simply amazing. And to see the encode times shrink on a 30-45 minute video I was ecstatic to pay for the cost of the 1080ti. I wonder if two would make it any faster?
 
I was honestly ready to send my 1080ti back after I ordered it. From reading all the reviews of it. But I'm glad I did try it. Brucex is a test and for benchmarking. But to go from 28-32 seconds to 13 or less. Simply amazing. And to see the encode times shrink on a 30-45 minute video I was ecstatic to pay for the cost of the 1080ti. I wonder if two would make it any faster?

As ekwipt said, if you really need tremendous performance in FCPX you should use an AMD card instead of buying 2 1080ti. A cheap Radeon will destroy the best Nvidia anytime in FCPX.
I'm a professionnal video editor and if I chose a 980ti it's because I wanted the right balance between gaming performance on windows and good performance in video applications on OSX (probably your case too?)
 
As ekwipt said, if you really need tremendous performance in FCPX you should use an AMD card instead of buying 2 1080ti. A cheap Radeon will destroy the best Nvidia anytime in FCPX.
I'm a professionnal video editor and if I chose a 980ti it's because I wanted the right balance between gaming performance on windows and good performance in video applications on OSX (probably your case too?)

I tried the AMDs and just honestly didn't care for them. But that's my personal opinion. I actually don't game. Unless it's poker. And the 1080ti can handle that in windows and OS X . I mainly do drone services for the community within a reasonable distance from me. I provide aerial video and pictures for anyone needing them. My YouTube page is for helping others learn how to properly use them. I've owned several mac pros and finally decided I wanted something that I can do what I want with it for less than half the price of a pro.
 
Status
Not open for further replies.
Back
Top