Puget Systems print logo

https://www.pugetsystems.com

Read this article at https://www.pugetsystems.com/guides/1043
Article Thumbnail

V-Ray CPU Comparison: New 14, 16, and 18-core Skylake-X Processors

Written on September 25, 2017 by William George
Share:

Introduction

V-Ray, from Chaos Group, is widely used for creating realistic 3D graphics. When rendering those graphics, both the central processor (CPU) and graphics processors (GPUs) can be utilized to increase performance. On the CPU side, 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.

Intel has just released a trio of new Core X series processors, with higher core counts than anything they have offered in this line to date: the Core i9 7940X, 7960X, and 7980XE - with 14, 16, and 18 cores respectively. We ran these chips through a few tests in V-Ray Benchmark 1.0.6 to see how they perform with rendering. We had recently put out an article using this same benchmark that covered a wider range of processors, and at the time AMD's Threadripper CPUs were found to be the fastest single-CPU option. This time we will focus just on the Core X series in comparison to Threadripper, to see if the new models can take the performance title back for Intel. For reference we included a dual Xeon system as well, to show how it compares to these single chip configurations.

Test Setup

To see how these different CPUs perform in V-Ray, we ran the free benchmark in CPU mode on the following configurations:

The main focus here is on the three new Core X (formerly code named Skylake X) processors, and specifically how they compare to AMD's Threadripper models. In our last round of testing we found that the 16-core 1950X took the lead in single-CPU performance with Keyshot, as well as many other CPU-based rendering engines. Here are some details about how we conducted our testing, but if you just want to skip straight to the results then feel free to scroll past this section.

The results presented below are from V-Ray Benchmark 1.0.6, which is a free benchmark released by Chaos Group. It is designed to test CPU and GPU performance within V-Ray without requiring a full installation of that software. Since the focus of this article is on CPU performance we only ran the CPU portion of the benchmark, which gives a time in seconds for how long it took to render a single scene. The rendering can also be watched in real-time during the benchmark, and the benefit of additional cores can be seen visually that way.

It is also worth noting that there are some differences in the amount and speed of RAM across the various test platforms. We prefer to use the speed of memory that each CPU is rated for, according to its manufacturer. For the current crop of Core X and Threadripper processors that is DDR4-2666, while the older Xeon uses slightly slower 2400MHz memory. Some of these platforms could be run with even faster RAM modules, but that is pushing the memory controller built into the CPU past its rated speed - overclocking it, effectively. That may lead to slightly increased performance but we have also found it to lead to stability issues and higher rates of memory failure. For that reason we stick with the manufacturer specs when it comes to selecting RAM for our systems.

Benchmark Results

Here are the results for the various CPUs we tested in V-Ray Benchmark 1.0.6. The new Intel processors are shown in light blue:

V-Ray Benchmark 1.0.6 Results with New Skylake X Processors

Since rendering in V-Ray is a heavily threaded application, there is a clear spread between the different processors based on core count. The dual Xeon, with a total of 28 cores, definitely wins out - but that is a much more expensive system and is really just included here as a point of reference. Among the single CPU workstations, the new 14 to 18-core models from Intel take back the lead that AMD's 16-core Threadripper had since its own launch last month. Intel has a 10-20% lead, depending on which Core i9 you look at, but it is worth noting that the 1950X is still less expensive... and outperforms the Intel processors which are in its price range.

It is also worth noting that 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 latest Xeon Platinum line. With their focus on the multi-CPU server segment those may not come into play for V-Ray, 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

Based on these results, it looks like the tables have turned since our last V-Ray CPU performance article. Intel took back the lead, but their new Core X processors are also substantially more expensive than AMD's Threadripper. If you want the fastest single-CPU rendering speeds then go for the i9 7980XE - but if you want a great value, the 1950X has strong performance while costing $1000 less.

V-Ray Workstations and Render Nodes

1 CPU / 1-2 GPU
Compact

Configure


1 CPU / 1-4 GPU
Tower

Configure


2 CPU / 1-4 GPU
Tower

Configure


1 CPU / 1-4 GPU
1U Rackmount

Configure


Tags: CPU, Rendering, Chaos, Group, V-Ray, Performance, Processor, Intel, AMD, Core, i9, Threadripper
Adam

Can you confirm if the official VRay benchmark uses the Embree raytracing kernel when operating on Intel systems (who developed it)? It's optional in a typical rendering scenario, e.g. VRay in 3DS Max, and by default is turned on. It can improve render times by a fairly significant %. It's my understanding that on Intel systems VRay will by default use Embree, an accelerating raytracing kernel that uses single precision fp rather than VRay's default double (and other stuff like AVX which is processed differently on AMD's architecture) to improve rendering speeds, which may skew the results in favour of Intel (it also means it's no longer an apples-apples result as Chaosgroup state in their documentation that the reduction from dp to sp floating point can introduce artifacts).

I'm not just trying to split hairs here as this could have a meaningful impact on buying decisions.

Posted on 2017-10-04 20:18:51

I actually don't know - there are no options for anything like that in the benchmark, and I don't see any documentation in a quick Google search that addresses this topic either. I would hope that, since it is a benchmark, they would leave out any architecture-specific stuff like that... but you might want to ask Chaos Group directly, as they will be the only folks that can say for sure.

We are also wrapping up some V-Ray RT testing in 3ds Max now, which will be published soon, but that is focusing on using CUDA based rendering - so I don't think your concern about Embree will be an issue in that article.

Posted on 2017-10-04 20:26:21
Jonathan Emms

William M George did you follow up on this? Would be curios to know.
Also found this info http://renderstuff.com/embr...

Posted on 2018-06-20 02:16:25

I wasn't able to find anything specific to the V-Ray benchmark, saying whether it did or did not use Embree. But looking at the results folks reported in the thread you linked to, it looks like both Intel and AMD procs (recent ones, at least) can benefit from it - so I'm confident, at least, that the benchmark is being "fair" (either using or not using Embree across the board, rather than giving a benefit to some chips but not others).

Chaos Group is supposed to be updating the V-Ray benchmark soon, to bring it up to date with the release of V-Ray Next. I'm looking forward to that new benchmark, but if it doesn't come soon I may try to do some manual testing as well in the near future.

Posted on 2018-06-20 16:15:30