The Best Computer for Rendering & Animation

by Will Gibbons | Last Updated: August 20, 2020 •  16 min read

Articles » Technology

So, you need a new computer, but you don’t know what to buy. Or perhaps, you aren’t sure if you need a new computer, but you’re wondering if your current machine is holding you back? Maybe, you’re beginning your career as a freelancer, and want to make sure your computer can keep up with your creative output. 

Sound familiar? I know. I’ve been there. I’m going to simplify the process of choosing your next computer. I want to remove the stress as well as save you time and money when you’re ready for a computer upgrade. 

By the time you reach the bottom of this article, I want you to be able to confidently choose your next computer. You’ll be able to see through all the marketing fluff and know how to spend your money to get everything you need and nothing you don’t. And it won’t be stressful. Oh, yeah, and if you’re not the ‘reading-type’, I’ve made videos that summarize the contents of this article.

Disclaimer: I want to extend an incredibly sincere thank you to Puget Systems for providing me with my new machine free of charge. When I reached out to them with the idea of a partnership, they were enthusiastic. I firmly believe Puget provides the best computers and services for people like me and that I want to share them with everyone I know. Any content I have created featuring Puget Systems would have been created even if I had not partnered with Puget Systems in any capacity.

Don’t have time to educate yourself?

Just want my recommendation? Buy a custom PC from Puget Systems. Hands down, they have the fairest pricing and best customer support you’ll find when it comes to serious computers for professionals.

Click Here To Visit Puget Systems

Who Needs A Workstation?

For most people, an off-the-shelf laptop will be more than capable of running an office suite, sending emails, web browsing, streaming movies, editing photos and even basic video editing. But, when you get into 3D graphics, 3D modeling and photo-real rendering, many off-the-shelf computers will quickly become a bottleneck in your productivity, or in some cases, simply not be able to perform at all. 

Whether you’re a student, enthusiast, or a professional, if you’re using 3D applications for professional media creation, you need a beefier computer than most people. Today, software is constantly improving to take advantage of the latest and most advanced hardware as it is released, so if you want to make sure your computer isn’t holding you back, this series is for you.

Computers I’ve Owned

• 2007 – 2011 – Macbook Pro
• 2011 – 2013 – Dell XPS desktop
• 2011 – 2015 – Macbook Air
• 2013 – 2018 – Custom computer from Boxx
• 2016 – 2019 – Macbook Pro for travel (supplied by employer)
• 2016 – 2019 – Mac Pro (supplied by employer)
• 2018 – Current – Built my first DIY workstation
• 2019 – Current – Dell Precision laptop for travel
• 2020 – Current – Puget Systems workstation 

I share this because I’ve had professional experience with both Mac OS and Windows-based systems, laptops and desktops, custom builds and off-the-shelf workstations. They all come with pros and cons.

Step 1: Decide HOW To Buy Your Next Computer

Off The Shelf vs. DIY vs. Custom Built

Before getting into configurations and components, decide which path to take in making a purchase. In the video below, I address how I came to make this decision myself and the pros and cons of each approach.

Here’s a summary if you don’t have time to watch the video:

If you want to get the cheapest computer possible, go with an off-the-shelf system from a big company like Dell, HP, or Lenovo. Keep an eye out for seasonal sales or small business discounts. For longevity and reliability, a mac is a great option, but I don’t think you get great value for your money here.

If you want to get the absolute best performance for your money, or if you’re on a limited budget, building your own PC is a great option. There’s lots to learn from the process and you’ll be able to get exactly what you want. The downside of course is a lack of support or warranty.

If you value reliability and customer support, then doing a custom configuration with a company that offers custom-built PCs is the best way to go. You get only what you need, but you get individualized care and support that is often missing from larger manufacturers. 

Why I Ordered A Pre-Built Custom PC

As mentioned, I’ve been working on a custom-built machine that I configured and assembled myself almost 3 years ago. I was able to make it exactly how I wanted and save a lot of money, which was my priority at the time. However, I’ve since started my own business and after losing some business due to some issues with my DIY machine, I decided it was time for a new computer.

As a business owner, who creates renderings and animations for a living, I need a computer to be reliable first and foremost and high performing as a close second. It’s very expensive if my computer prevents me from completing a job, so I can easily justify spending more if it means my computer is very reliable.

Issues With My DIY Computer

While my DIY computer still has some very high-end components, as mentioned, I had a few crashes that lead to costly down-time. I took it in for professional diagnostics twice, re-installed the operating system, replaced the motherboard twice, upgraded the CPU, changed from liquid cooling to air-cooling, replaced the boot drive, storage drive and replaced the PSU. In total, this cost about $3,000. It took nearly re-building my PC to make it stable and even then, I decided I didn’t want to put all my eggs in one basket if it were to fail on me again.

Another issue I had was that the CPU was optimized for one task only. When I was doing nearly anything other than rendering, my machine was quite slow. This is due to how the CPU with its many cores, were not being utilized. For example, I had a client project in which I had to add some detail to a CAD model. I was adding fillets to a bunch (about 4,500) of small holes. This one operation took nearly 3 hours to complete and lead to multiple program crashes.

I also was getting back into recording videos and tutorials for YouTube. I found that certain tasks during the video recording and editing process again, weren’t being handled very quickly by the 2nd gen Threadripper 2990wx CPU in my computer. 

Finally, when I did have a project rendering, my computer was tied up, so I couldn’t keep working. Having to schedule all your rendering to happen overnight isn’t the most convenient solution. 

So, I decided to start looking for new computer solutions. And since I decided that stability and support was a top priority, I decided to go with a custom-configured computer that was purpose-built for me and my specific workflow.

Step 2: Decide What Components You Need

Too Lazy To Read? Here you go, you creature of comfort:

What Should You Prioritize?

So, what’s the most important consideration? 

• Budget?
• Operating system?
• Form factor? 

Nope! It all starts with what tasks you’ll be doing and in what software you’ll be doing it in. There are a couple of key components within your computer that are responsible for how it performs. Disclaimer! I’m going to make a few generalizations here to keep things as simple as possible.  

• The CPU, or central processing unit is what determines how fast certain tasks are performed like simulations, calculations and rendering
• The GPU, or graphical processing unit is responsible for moving pixels around on the screen as well as certain specialized rendering tasks
• RAM, or random access memory makes certain data quickly accessible to the CPU to speed up processes like launching an application, saving a file or presenting lots of data at once, like all those browser tabs
• Storage, comes in a few formats, many capacities, holds files, but also has a speed associated with it which affects how quickly software can access data
• Cooling, keeps components running efficiently and protects them from damage due to excessive heat
• The Motherboard, is what all these components are connected to and determines what components are compatible with one anther
• The PSU, or power Supply Unit sends power to each component and needs to be large enough to accommodate the total power your computer can draw under a heavy load

Tasks: What They Are And Why You Should Care

When you perform a certain task on the computer, you give it a command and it performs some calculations and tasks. How quickly your computer performs this task depends on the complexity of the task and the capabilities of the specific components that are carrying out that task. 

For example, if I am doing some 3D modeling, and I extrude a cube, I’m instructing the computer to take a sketch of a square and turn it into a 3D object. The CPU has to take my order, create the thing I asked it to do and the GPU needs to create a correspondence 3D cube on screen. This simple task will happen quickly on almost every modern computer. 

Now, let’s say I’m creating a simulation of an explosion. I’ve set lots of parameters like how big I want the explosion to be. I’ve set velocity, smoke, turbulence, voxel size, resolution and many more properties. The CPU will often handle a task like this due to its high complexity. There are so many variables and calculations that need to happen in order to complete this task with accuracy. A CPU often handles the most complex calculations. Then, once the simulation has been completed, the GPU (which is responsible for graphics) may be used to actually render out this data and make it look like a real explosion. 

So, depending on the kind of tasks you’re doing, different components will be used. By understanding the kind of tasks you’ll use your computer for, you’ll be ready to identify which components you should prioritize, which in turn will determine what computer will be best for you. 

CPU vs GPU

CPUs

When it comes to rendering, there’s a lot of confusion around CPUs and GPUs. The most common question I get is, which is better? With many rendering applications now supporting both CPU and GPU rendering, you may want to know which method you should use and thus, which to spend your money on. 

Historically, the CPU was used for rendering due to its ability to handle complex calculations like multiple indirect light bounces, refraction and caustics. One downside to performing all of these calculations on a single CPU is that it could be slow. 

There are two main factors that determine how fast a CPU can do things. There’s a core count and a clock speed. Let’s say we’re building a pyramid. How fast the pyramid depends on how many workers we have and how fast each one works. The cores are like workers and the clock speed is how fast each worker performs. 

If you’re rendering an image with a CPU that has 2 cores and one that has 10 cores, the one with 10 cores will be approximately 5 times faster. Thanks to advancements in software, a thing called hyperthreading exists. With hyperthreading enabled in your operating system, for every physical core a CPU has, a theoretical core exists. The total number of physical and theoretical cores equals the total thread count of a CPU. So, an 8-core CPU has 16 threads.

This is important to understand because some software takes advantage of hyperthreading and some do not. To make matters worse, only SOME operations within a given software package will take advantage of all those threads. 

Finally, the clock speed determines how fast each core can operate, and is measured in GHz. The more cores your CPU has, the slower each core will perform.

Understand, a CPU with just a few cores and a high clock speed will out-perform a CPU that has many cores and a lower clock speed if a task or software does NOT support multi-threaded operations. 

Imagine Company A has 4 employees who can make a widget in one hour. Company B has 16 employees who each take 2 hours to make widgets. If they are both making widgets, but only one employee is allowed in the factory at a time, Company A will make the widget in 1 hour and Company B will take 2 hours. But, if ALL employees are allowed in the factory at a time, in just 2 hours, Company A will have made 8 widgets and Company B will have made 16 widgets. Even though Company A’s employees are faster than Company B’s employees, eventually, the higher number of employees yields greater results. This is pretty much the same when it comes to rendering. 

GPUs

Over the years, GPUs have become more similar to CPUs. GPUs also have a speed as well as a number of cores. They even have their own dedicated RAM too, which we’ll get to. 

If you’re not using software that uses the GPU for rendering, you don’t need to spend lots of money on a fancy GPU. While a core on a CPU isn’t exactly the same as a core on a GPU, modern GPUs have hundreds of cores that tend to perform simpler calculations all at the same time. This, coupled with the fact that you can easily add multiple GPUs to a motherboard, has made GPUs popular for rendering.

For simpler rendering scenes, GPUs make the most sense, but historically, the CPU has been the most capable of producing photo-realistic images and handling the most complex tasks. Due to recent breakthroughs, more and more tasks are being handled by GPUs including ray tracing, a method of rendering known for producing the most realistic images. 

And just like CPUs, GPUs have a speed associated with them as well. 

One thing that is unique to GPUs is the on-board RAM, called VRam (short for video ram). This is important to consider because when it comes to GPU rendering, many software requires the scene being rendered to be held in VRam, and if you’re using large textures, particles, or referencing many parts within your scene, it’s easy to run out of VRam. So, buying a GPU with plenty of VRam is important, but also affects the cost of the GPU quite a lot. 

So What?

Most computers prioritize the CPU or the GPU. The tasks you’ll be doing all the time will determine which one you want to prioritize. Taking a closer look at the specific software and the tools within it will allow you to decide on the specific model of CPU or GPU that your software will be able to take advantage of.

For example, if you’re creating 3D CAD models all day, most 3D modeling tools perform single-threaded operations only. You can test this by reading the software documentation, asking on a forum or watching your computer’s resource manager and watching how many CPU cores are being used as you use the software.

Simulations and 3D rendering are two of the tasks that often DO take advantage of as many cores as you can throw at them, but be sure to double-check, as once again, not every software is built to support this.

Step 3: Pull The Trigger (My New PC specs)

If it’s not painfully obvious, I decided to order my new computer from a company that specializes in making custom PCs for professionals. While there are tons of options I could have gone with, I chose to patronize Puget Systems. Here’s why I went with the smaller Seattle-based computer company:

• Incredibly knowledgeable – Puget’s blog is chock-full of some of the best testing and practical computer advice written by some expert employees (as opposed to outsourcing their blog content).
• Friendly and Personable – When you call Puget, you speak with a human who asks about how you work and what your goals are and they make recommendations. If I’m going to trust my business with a company, I want to know there’s a human on the other end I can speak with.
• Fair Pricing – While paying a company to build you a PC is not as cheap as building one yourself, Puget does not unnecessarily inflate the cost of their services. They also forgo unnecessary bells and whistles to pass savings onto their customers.
• Customization & Transparency – If you can’t configure exactly what you want on their website, Puget can tailor your hardware to meet your needs. And the entire process from placing your order to receiving your machine is updated in real-time for you to track once you log into Puget’s website. You’ll be able to see who’s picking out parts, who’s building the machine, all the tests and benchmark results as well as photos of your machine, its bios settings and thermal imaging under load.
• Legendary Customer Support – When you own a business or rely on your computer for income, having reliable customer support is worth its weight in gold. Puget has this in aces. They’re accessible, responsive and helpful no matter what you run into throughout the process of ordering and owning a Puget Systems PC.

Here’s the video version that covers my experience of working with Puget Systems and my new computer’s specs.

My Puget Systems PC Spes:

• Motherboard: Gigabyte TRX40 AORUS PRO Wifi
• CPU: AMD Ryzen Threadripper 3970X (3.7 GHz 32 Core 64 thread)
• RAM: 4x Crucial DDR4-2666 32GB (total 128 GB)
• GPU: EVGA GeForce RTX 2080 Ti 11GB Open Air
• Hard Drive: Samsung 860 EVO 1 TB SATA3 2.5 inch SSD
• Scratch Disk: Samsung 970 EVO Plus 2TB M.2 SSD
• Archive Drive: Western Digital Ultrastar 8TB SATA3
• Case: Fractal Design Define R6 USB-C Blackout
• Power Supply: EVGA SuperNOVA 1000W P2 Power Supply
• CPU Cooling: Noctua NH-U14S TR4-SP3
• Thermal Paste: Arctic Cooling MX-2 Thermal Compound
• Case Fans: Noctua NF-A14 PWM 140mm

If you’re at all curious about whether Puget Systems is a good match for you, please visit their site and check them out! They’re friendly and able to answer any questions you may have for them.

Visit Puget Systems

That concludes this hefty article! I’ve been using my computer from Puget Systems for a solid couple of months and can say it’s working like a charm. It’s nearly silent, very fast and snappy. I look forward to reporting back on how it’s boosted my productivity and removed bottlenecks from my workflow once I get through a few client projects. Stay tuned!

I hope this in-depth article was helpful! Have any questions for me? Feel free to join the Discord and shoot me a message!

Will Gibbons is an industrial designer-turned rendering specialist. He's trained and worked with some of today's most recognizable brands and now is focused on creating the best product visualization content on the internet.