Kicking off our ‘Price Vs Performance Series’ in style, we’ve conducted extensive in-house testing that you can benefit from when it comes to choosing the optimum graphics card options in your custom Apple computer.
Why Benchmark GPUs?
Benchmarking graphics cards gives an accurate representation of just how well a GPU will survive against an assortment of graphical requests and effects. If you’re thinking of making upgrades, or you’re just generally looking to see how your current GPU works against other models and brands, then we’ve got what you need!
What Did Our Tests Involve?
Our extensive graphics card benchmarking put eleven GPUs through their paces, and the results we gained paint a nice picture of which graphics card offers you the best in terms of price versus performance.
Below are the GPU units that were involved in our testing:
- AMD RX570
- AMD RX580
- AMD Vega64
- Nvidia 1070
- Nvidia 1080Ti
- Nvidia 1080
- Nvidia Titan X (P)
- Nvidia 980Ti
- Nvidia Titan Xp
- Nvidia Titan X (M)
- iMac Pro Vega64
The test systems involved in our testing were:
- iMac Pro 1,1 (2018)
- 3.2Ghz 8 Core
- 32GB RAM
- AMD Vega 64
- 1TB Flash SSD
- Mac Pro 5,1 (2012)
- 3.46Ghz 12 Core
- 96GB RAM
- 1TB PCI-e SSD
So, what are we looking for in our test? Well, the point of selecting the right graphics card is to ensure that your workflow is efficient, and you are able to get a good return on investment. Running tests to measure aspects such as latency, long render times and poor playback frames per second are all things that an incorrect graphics card choice will cause.
We broke the testing into two sections; overall performance including Final Cut Pro results and overall performance excluding Final Cut Pro results. This gives us the chance to identify if there are any areas that the cards struggle with once Final Cut Pro was introduced to the fold. This is especially poignant following an exclusive by the Create Pro team who compiled an assessment that reveals Apple could well have been slowing down not only the iPhone but Macs as well! Of course, we wanted to highlight to you which cards offered you the best bang for your buck, for want of a better phrase.
So, let’s get to the juicy stuff, shall we? We’ve picked out the main talking points from the following applications:
Geekbench 4 Compute OpenCL
Looking at the results from Geekbench 4 Compute OpenCl, AMD’s strong point is in its ability with OpenCL. Nvidia can be pleased with the improvements they saw with the OpenCL performance in the Pascal Architecture, while despite AMD’s efforts, the Vega64 flagship still cannot outperform the top Nvidia models.
The iMac Pro Vega64 version is only slightly slower despite being inside an iMac, and this can be, in part, tied to the fact that Apple have massively improved on its thermal design. Interestingly, the RX570 was only marginally slower than the 980Ti and Titan X (M), which shows there is some massive value for power in the RX570; the best of the bunch, in fact!
Davinci Resolve Standard Candle Light Test
Graphics cards are the main workhorse in this application, so who prospered here? Well, the Nvidia Cards are highly optimised with their use of the more mature CUDA platform. Interestingly the iMac Pro Vega64 outperformed the standard Vega64. We assume this is probably due to the iMac Pro’s excellent optimisation of the card.
Unfortunately, despite the most valiant efforts from AMD, their RX cards came dead last!
LuxMark 3.1 LuxBall HDR Benchmark
This benchmark is geared more towards the OpenCL architecture and the AMD Vega was the one GPU that took the bull by the horns and came into its own here. It outperformed the 1080Ti and came exceptionally close to the Titan Xp.
The iMac Pro Vega was a middle of the pack average runner in this benchmark, coming just above the 980Ti from the discontinued Maxwell line-up from NVidia.
The Titan model cards are designed for this workflow and its clear from the results.
Alas, performance from the RX cards was quite underwhelming but to be expected in this case.
Unigine Heaven Extreme Benchmarks
*This tool provides two scores so there are two tables
Why Unigine prefer the older Titan X (P) over the newer Titan Xp is unknown. There is a suggestion that it might be optimisation changes, but this is yet to be confirmed or denied.
The iMac Pro came in on this benchmark and proudly held its own, highlighting that it has great 3D performance.
Moving on to the RX cards, they were significantly worse on these tests; but it’s worth touching on the fact that they are by no means unusable. They still show they would be great as entry-level graphics cards.
Final Cut Pro X Bruce X Benchmark
We found that the order of performance was massively different in this benchmark. Here’s what the results threw our way:
With the introduction of Final Cut Pro X, this particular application was massively in favour of the AMD and their OpenCL performance.
Some users will be intrigued and possibly surprised to see that despite the 1080Ti crushing the AMD cards in normal OpenCL benchmarks, they are much slower when running Final Cut.
Conclusively, if you use Final Cut Pro, there’s nothing better from the GPUs in the test than the AMD line-up and the iMac Pro absolutely tore through the tests.
Create Pro’s Additional Notes
- You can upgrade the graphics card in a Mac Pro 5,1 as technology updates.
- You have access to PCi-e slots which you can use to expand the system and run up to three graphics cards simultaneously.
While you may have a favourite brand of GPU that you have a soft spot for, you may be looking at other options after seeing how they performed in our latest set of tests. Will you be looking to test new waters? If you feel you need a little more advice on what’s best for your requirements, then feel free to get in touch with us here at Create Pro today!