Puget Systems print logo
Read this article at https://www.pugetsystems.com/guides/1661
Article Thumbnail

V-Ray CPU Rendering Performance: AMD Threadripper 3990X 64 Core

Written on February 7, 2020 by William George


AMD has launched a new top-end CPU in its Threadripper product line, equipped with a whopping 64 cores. We are putting this new 3990X chip to the test in one of the applications where its high core count should shine: CPU based rendering with V-Ray. Keep reading to find out how it stacks up to the other Threadripper models as well as Intel and AMD's various other desktop processors.

AMD Threadripper 3990X Tested in V-Ray Next Benchmark

For this article, we are focusing on just how this new AMD Threadripper 3990X compares to other current-generation processors. We want to measure CPU-based rendering performance in V-Ray, so we are using both Chaos Group's V-Ray Benchmark as well the newer V-Ray Next Benchmark. More info about these tools and how we use them are available in the Benchmark Details.

If you would prefer to skip over our test setup and benchmark sections, feel free to jump straight to the Conclusion.

Looking for a V-Ray Workstation?

Puget Systems offers a range of powerful and reliable systems that are tailor-made for your unique workflow.

Configure a System!

Labs Consultation Service

Our Labs team is available to provide in-depth hardware recommendations based on your workflow.

Find Out More!

Test Hardware

Here are the specifications of the different platforms we used for our testing:

AMD Threadripper 3rd Gen Test Platform
CPU AMD TR 3990X ($3,990)
AMD TR 3970X ($1,999)
AMD TR 3960X ($1,399)
CPU Cooler Noctua NH-U12S TR4-SP3
Motherboard Gigabyte TRX40 AORUS Pro WiFi
RAM 4x DDR4-2933 16GB (64GB total)
Intel X-10000 Series Test Platform
CPU Intel Core i9 10980XE ($979)
Intel Core i9 10940X ($784)
Intel Core i9 10920X ($689)​​​​​​​
Intel Core i9 10900X ($590)​​​​​​​​​​​​​​
CPU Cooler Noctua NH-U12DX i4
Motherboard Gigabyte X299 Designare EX
RAM 4x DDR4-2933 16GB (64GB total)
AMD Ryzen Test Platform
CPU AMD Ryzen 9 3950X ($749)
AMD Ryzen 9 3900X ($499)
AMD Ryzen 7 3800X ($399)
CPU Cooler Noctua NH-U12S
Motherboard Gigabyte X570 AORUS ULTRA
RAM 4x DDR4-2933 16GB (64GB total)
Intel 9th Gen Test Platform
CPU Intel Core i9 9900K ($499)
Intel Core i7 9700K ($385)
CPU Cooler Noctua NH-U12S
Motherboard Gigabyte Z390 Designare
RAM 4x DDR4-2666 16GB (64GB total)
Shared Hardware/Software
Video Card NVIDIA GeForce RTX 2080 Ti 11GB
Hard Drive Samsung 960 Pro 1TB
Software Windows 10 Pro 64-bit (version 1909)
V-Ray Benchmark 1.0.8
V-Ray Next Benchmark 4.10.06

A quick note about memory specifications: the latest CPUs from AMD and Intel support varying speeds of RAM depending on how many modules are installed. In past articles, we've sometimes tested at multiple speeds, but to help keep things more uniform going forward we are standardizing on the speed that each processor/platform supports when all memory slots are used (even if we aren't filling them completely for our testing). Along with using the same total RAM capacity, that should help to keep results fair and consistent.

Additionally, it is worth pointing out that the CPU cooler we used for testing the new TR 3990X is somewhat borderline for how much heat this processor can potentially put out. In the systems we build with this chip going forward, we will be utilizing a larger heatsink and fan to ensure maximum performance even under extended loads - but in the open-air testbeds we use here in our lab, we found that the cooler we used did not lead to throttling.

Benchmark Details

As previously mentioned, we used two versions of Chaos Group's V-Ray Benchmark for this comparison:

  1. V-Ray Benchmark 1.0.8 is older, but some folks may still be utilizing V-Ray plug-ins from a year or two ago - and the test is short enough that we decided to go ahead and run it this time around to have an extra data point. It includes both CPU and GPU tests, but we are only listing the CPU score here since GPU performance is not our focus nor is it impacted by the CPU in this version.
  2. V-Ray Next Benchmark includes tests for both CPU-only and GPU+CPU rendering. This is somewhat novel, as most GPU rendering engines do not use the CPU at all, but in V-Ray Next Chaos Group has implemented CUDA emulation on the CPU to improve performance a bit. Up until recently, even the fastest CPU we had tested didn't add as much performance in this mode as a single high-end video card... but the 3990X's 64 cores may have changed things.

Benchmark Results

With the information we have gathered, then, we have put together three charts. The first is the CPU's native performance when V-Ray Next is running in CPU mode, while the second is CPU performance when running in V-Ray Next GPU mode. The GeForce RTX 2080 Ti card in the test systems has been excluded from that result, so you can see just the contribution that the processors provide in this mode. And finally, the third chart shows CPU render time, in seconds, from the older V-Ray Benchmark 1.0.8.

There is a lot going on in the charts below, so before getting into it here is a reference key for the color scheme:

  • Light blue = Intel consumer CPUs (9th Gen Core series)
  • Dark blue = Intel HEDT CPUs (X series)
  • Light red = AMD consumer CPUs (Ryzen)
  • Dark red = AMD HEDT CPUs (Threadripper)
  • Dark red with glow - NEW AMD Threadripper 3990X


Unsurprisingly, AMD's new Threadripper 3990X took the top performance spot across the board in V-Ray. CPU rendering scales very well with lots of CPU cores, as long as clock speed doesn't suffer too much, and AMD has done a fantastic job of combining those two aspects of performance in their 3rd generation of Threadripper processors. The only alternatives that can provide anything close to the same performance are multi-socket systems with AMD Epyc or Intel Xeon CPUs, but those would cost far more to get similar results.

Moreover, the 3990X is the first processor we've tested that can exceed the performance of a high-end video card when used in V-Ray Next's GPU mode. For reference, a GeForce RTX 2080 Ti scores around 330-340 mpaths in that benchmark, while the 3990X pulled off 438! Of course, the 2080 Ti costs far less money as well - so real video cards still offer the best value for increasing performance in V-Ray Next GPU. If you want to max out your render speed, though, pairing multiple GPUs with a high-end CPU like the 3990X is great!

Is AMD's Threadripper 3990X Good for V-Ray rendering?

Yes: the new AMD TR 3990X offers the best CPU-based rendering performance currently available from a single-socket workstation! Only dual- or quad-socket configurations can outperform it, and those would cost far more money even just to reach the same speed.

Looking for a Rendering Workstation?

Puget Systems offers a range of poweful and reliable systems that are tailor-made for your unique workflow.

Configure a System!

Labs Consultation Service

Our Labs team is available to provide in-depth hardware recommendations based on your workflow.

Find Out More!
Tags: Intel 9th Gen, Intel X-series, Intel vs AMD, AMD Ryzen 3rd Gen, AMD Threadripper 3rd Gen, Threadripper, V-Ray, CPU, Rendering, Threadripper 3990X, Intel X-10000
Zain Fadhil

Performance/Price comparison :
3990x ..............Cinebench R20 score: 24k @ $3850. (6.23 pts.per Dollar)
3970x...............Cinebench R20 score: 17k @ $2000. (8.5 pts. per Dollar)
5950x ..............Cinebench R20 score: 10k. @ $800. (12.5pts. per Dollar)

Compared to a 3970x, the 3990x is almost double the price but only 45% faster. Question is: is the 3990x really worth buying?

Posted on 2020-11-06 14:31:04

Depends on your work really. Performance per dollar is almost always worse the higher up you go, but what you gain is increased performance in a single box. The 40% faster render times with the 3990X over the 3970X is pretty good, and for those whose livelihood depends on how efficient they can be, that $1,850 difference could pay for itself really quickly. It will be different for everyone, but if someone gets paid $50/hr and the 3990X lets them get 40% more work done every hour, that $1,850 is paid off in just 26 hours.

Of course, things don't actually work like that in real life, but for someone who does a lot of CPU rendering, the return on investment for a CPU like that can be way quicker than you might think.

Posted on 2020-11-06 17:31:11
Zain Fadhil

@Matt Bach . I agree, time is money. I apologies, I should'd explained the whole concept. So we mainly use Vray as our rendering engine (GPU+CPU) & did made some PC parts cost calculations of a 3990x system including quad 3090s and compared that to another scenario where we use 4 separate systems each with 5950x & a single 3090, all 4 systems will be on distributed rendering as render servers. Theoretically, when using this approach we should be seeing around 25% increase in CPU performance with only a $1000 difference in price, plus avoiding the noise/Heat issues

Posted on 2020-11-06 18:33:02

If you get into network rendering, you can do some pretty cool stuff like that! Either way you go has it's pros and cons. A single workstation is less complex, licensing is cheaper (when you need a license per machine at least), and just generally easier to work with. Multiple machines in a render farm setup can be more cost effective, but generally uses more power, has higher licensing costs, and is more complex, so more stuff that can break and requires additional setup.

On the other hand, I know some people like to have multiple machines when they have more employees because it lets them "break up" the hardware resources. So you can allocate more resources to one person who is in a crunch versus someone who is ahead of schedule.

Posted on 2020-11-06 18:37:40
Zain Fadhil

Yes, you do have a point with the additional licensing & complexity. From what I knew, we might go for 2x 3990x systems (for our senior animators) and keep 8x 5950x pcs as render slaves for overnight rendering.

Posted on 2020-11-06 19:00:51

Unless you are just buying a CPU alone (like for an upgrade to an existing computer) then I think that viewing the price of that component alone can be misleading. Instead, consider the performance you would get from a whole system... and the price of that whole system. Lets say, for example, that the rest of a workstation (not counting the CPU) would be $3,000 - motherboard, RAM, GPU, drives, chassis, PSU, etc. now you are talking about price and performance more like this:

3990x ..............Cinebench R20 score: 24k @ $6850. (3.5 pts.per Dollar)
3970x...............Cinebench R20 score: 17k @ $5000. (3.4 pts. per Dollar)
5950x ..............Cinebench R20 score: 10k. @ $3800. (2.6pts. per Dollar)

Now that isn't perfect - for example, a Ryzen based system would cost a little less due to using a cheaper motherboard and CPU cooler... but the general idea is demonstrated well. The 3990X is now the best for "pts per dollar", but the 3970X is comparable so if you were on a budget it would be fine too from that perspective. But Matt also made good points with the way he looked at it: if the extra money saves you time in a job, then that can pay for itself over a relatively short period of time :)

Posted on 2020-11-06 17:47:12