Arnold for Maya CPU Comparison: Skylake-X vs Threadripper

Always look at the date when you read an article. Some of the content in this article is most likely out of date, as it was written on September 1, 2017. For newer information, see our more recent articles.

Table of Contents

Introduction

Arnold is the render engine included with Maya 2017, and is also available as a plug-in for several other design applications. It is made by SolidAngle and produces beautiful 3D renditions of scenes and visual effects. When rendering those graphics, the performance of the central processor (CPU) in a computer is of paramount importance. CPU-based rendering generally scales well with both clock speed and core count, but those specifications cannot be directly compared across different brand or generations of processor. Here at Puget Systems we do real-world testing to ensure we provide our customers with the right computer for their needs.

In this article we are going to look at the latest CPU options from Intel and AMD, both of which have recently released high core count processors that should be great for Arnold. From Intel we have the Core X series of processors, which were code-named Skylake X. They come in 6- to 12-core models currently, though up to 18-core variants will be coming in the future. Up against these chips are AMD's new Threadripper processors, based on a pair of Zen chips combined into a single package. We'll be looking at the 12- and 16-core models in that family to see how they stack up against Intel's traditional lead in this application.

We are also including several other CPU platforms in this article, as points of comparison. We've got Intel's lower core count Kaby Lake chips, along with AMD's similar Ryzen line. Intel's Xeon versions of the Skylake X are also present, called Skylake W or Xeon W depending on where you look. And to top things off, a dual Xeon configuration is included as well – to show what can happen if you put two powerful processors together.

Test Setup

To see how these different CPUs perform in Arnold, we ran it within Maya 2017 on the following configurations:

Skylake X (X299) and Threadripper (X399) Test Platforms
Motherboard:	Gigabyte X299 AORUS Gaming 7 (rev 1.0)	Gigabyte MW51-HP0	Gigabyte X399 AORUS Gaming 7 (rev 1.0)
CPU:	Intel Core i7 7820X 3.6GHz (4.3/4.5GHz Turbo) 8 Core ~$599 Intel Core i9 7900X 3.3GHz (4.3/4.5GHz Turbo) 10 Core ~$999 Intel Core i9 7920X 2.9GHz (4.3/4.4GHz Turbo) 12 Core ~$1199	Intel Xeon W-1235 3.7GHz (4.5GHz Turbo) 6 Core ~$835 Intel Xeon W-1245 3.7GHz (4.5GHz Turbo) 8 Core ~$1113 Intel Xeon W-1255 3.3GHz (4.5GHz Turbo) 10 Core ~$1440	AMD Threadripper 1920X 3.5GHz (4.0GHz Turbo) 12 Core ~$799 AMD Threadripper 1950X 3.4GHz (4.0GHz Turbo) 16 Core ~$999
RAM:	8x DDR4-2666 16GB (128GB Total)	4x DDR4-2133 16GB ECC Reg. (64GB total)	8x DDR4-2666 16GB (128GB Total)
GPU:	NVIDIA Quadro P6000 24GB
Hard Drive:	Samsung 960 Pro M.2 PCI-E x4 NVMe SSD
OS:	Windows 10 Pro 64-bit
Software:	Arnold for Maya

These test configurations include three sets of Intel and AMD processors. CPU-based rendering in Arnold is known to scale well across multiple cores, and ideal performance for rendering is usually found from processors with 10+ cores. Dual CPUs are often a great way to increase performance as well, though they cost more and are outside the scope of this particular article.

The results presented below are from Arnold for Maya, which comes standard with Maya 2017. That program is available from Autodesk for a 30-day trial, as well as in several subscription options for long-term use. Within Maya we opened a test file called from Solid Angle called "model village", and then navigated to the Arnold render window. The scene was rendered at two resolutions: 640×480 and 1280×960. Initially we left other settings at default, but mid-way through testing we found out that 3 threads are normally reserved for UI and other applications. This threw off our testing, especially with lower core count CPUs, so we began setting that to 0 threads reserved. A lot of test time had been lost at that point, though, so we only re-tested with the new settings on the few CPUs which we really wanted to focus on for this article.

It is also worth noting that there are some differences in the amount and speed of RAM across the various test platforms, as well as the video card utilized. We prefer to use the speed of memory that each CPU is rated for, according to its manufacturer, but in the case of the new Xeon W processors we couldn't get the rated speed running on the motherboard sample we have. Intel processors generally show little impact from small changes in memory speed, though, so that shouldn't impact the results much.

Benchmark Results

Here are the results, in seconds, for the various CPUs we tested in Arnold for Maya at 640×480. Lower times are better:

And here are the results for the same scene but rendered at 1280×960:

Both sets of results are nearly identical in terms of relative CPU performance. Since rendering in Arnold is a heavily threaded application, there is a clear spread between the different processors based on core count. AMD's 16-core 1950X comes out as the clear winner here, being both faster and less expensive than Intel's 12-core i9 7920X. Next up are AMD's 12-core 1920X against Intel's 10-core i9 7900X, which are in a dead heat. On the Xeon side, the 10-core W-2155 lags behind a bit, and the rest of the processors fan out from there in accordance with the number of cores they have.

An interesting thing to note here is that even though Intel's Xeon W line falls behind the Core X series a bit they still cost more. Intel charges a premium for the Xeon variants since they support features like ECC memory. If that matters to you then they are still a viable option in terms of performance, but be aware that you will pay extra for them. Many AMD processors also support ECC (though not registered) memory, but not all motherboards do.

It is also worth noting that Intel has 14, 16, and even 18-core models coming soon in the Core X series. When they are released later this year some of them will likely beat out AMD's Threadripper chips for the top performance spot in single CPU rendering, but at a higher price tag. AMD and Intel both have server-class processors with even more cores as well: up to 32 on AMD's EPYC and up to 28 from Intel's Xeon Platinum line. With their focus on the multi-CPU server segment those may not come into play here, but if a manufacturer puts out a single-socket workstation motherboard that is compatible with either of those platforms then they may be worth a look.

Conclusion

Here is a summary of render performance in Arnold between just the top two Skylake X and Threadripper processors:

Based on these test results, and as of the writing of this article, Threadripper is clearly in the lead for CPU based rendering in Arnold. The 16-core 1950X not only beats the 12-core i9 7920X, it also costs less. Looking to the near future, Intel's 14-18 core processors are likely to take back the performance lead – but will still cost more than AMD's offerings. We will test those when they become available and publish additional results at that time. We will also use this data to craft our recommended systems for Arnold going forward.