How to Stress-Test Graphics Cards (Like We Do)

Readers often ask us which applications are best for stability and temperature testing, and why we use some, but not others. In response, we generated some results to help quantify our choices for measuring power, heat, and the usefulness of overclocks.

Why Do We Need to Run Stability & Stress Tests?

As enthusiasts, we're always eager to figure out more about how new graphics hardware works at a technical level. Theoretical performance is easy to calculate based on the speeds and feeds AMD and Nvidia give us. But those numbers aren't always practical, particularly once we bolt our cards in and close up our cases. Even the best reviews don't reflect the exact conditions in your specific chassis. Perhaps even more important, when it comes to dialing in an overclock, you're left to your own devices. What we see from one sample rarely applies exactly to other cards. Last, but not least, new builds and upgrades often present unexpected issues that have to be diagnosed. Knowing how to run your own stability tests helps push through them to ensure everything operates the way it should.

Whether your goal is to uncover thermal or power problems during normal operation, or determine how far components can be pushed before causing crashes, you have to use the right software. In this way, weak spots can be identified, fixed, and optimized.

So, we're compiling a two-part software round-up designed to give you a detailed overview of necessary tools, each with its own list of pros and cons.

  • Part 1: How to Stress Test Graphics Cards (Like We Do) (This Part)
  • Part 2: CPU Stability Testing and System Stress Testing (Coming Soon)

Important Warning about All Stress Tests

Before we begin, we need to warn our readers about the inherent dangers of running stress tests. The software that we’re presenting might not just produce a full graphics load, but also potentially push beyond manufacturer-defined power limits. Using these so-called "power viruses" can result in damage to the system, especially if they're used for extended periods of time.

Those who want to use these applications assume all responsibility for the outcome. Be sure to continuously monitor relevant parameters, including temperatures, with accurate and up-to-date utilities. At least you'll be able to abort your testing immediately if it becomes necessary.

Choosing the Right Monitoring Application

GPU-Z is the established go-to application for graphics card monitoring. It’s free, doesn’t take up a lot of resources, and displays each board's most important parameters. It can also write a log of those parameters in real time. GPU-Z is continuously updated and improved, making it a great choice for beginners (even though it doesn't provide an in-game overlay). In addition to temperatures, a number of loads, limits, fan speeds, and power/voltage values can be displayed and logged. Download GPU-Z here.

MSI Afterburner has even more bells and whistles. It’s not just an overclocking tool, but also includes many monitoring functions. In addition, it provides an in-game overlay that displays whatever real-time measurements you choose.

AMD’s Radeon software can do this as well these days, but Afterburner is easier to configure and more flexible in the variables shown on-screen (not to mention their fonts and colors). Download MSI Afterburner here.

HWiNFO64 can do more than just monitor graphics cards, but it’s still a good choice even if that’s all you need it for. The software reads practically any sensor output, and can write it to a file in real time. However, the sensor loop tends to lag due to its sheer number of readings. Even one-second intervals don't always prevent a lag in the time stamps. 

Consequently, our recommendation is to not just hide unnecessary sensor readings (network, system, drives, etc.), but to exclude them from the loop altogether. This makes for a less cluttered display and gets rid of the aforementioned lag, even using 500ms intervals. Download HWiNFO64 here.

We’ll see in the second part of this article that HWiNFO64 is also a fantastic tool for reading motherboard sensors.

Choosing the Right Graphics Card

For this article, we're deliberately using a lower-end graphics card with a 100W power limit because we’re also interested in exploring its individual component temperatures. Asus' ROG Strix Radeon RX 560 does the job nicely. Neither its memory modules nor its voltage converters make contact with the cooler, meaning they can't affect each other. They are also placed apart on Asus' PCB, allowing us to draw more confident conclusions.

We did conduct some spot tests with significantly larger graphics cards (Sapphire's Radeon RX Vega 64 Nitro+ and the Galax/KFA² GTX 1080 Hall of Fame). These confirmed what we suspected: due to the issues mentioned above, high-end cards heat up completely after ~30 minutes, obscuring the most important details.

Test System & Methodology

We introduced our new test system and methodology in How We Test Graphics Cards. If you'd like more detail about our general approach, check that piece out. We've adjusted the CPU and the cooling system to better suit this article.

The hardware used in our lab includes:

Test Equipment & Environment
System
Intel Core i7-8700 ($312.00 On Amazon)
MSI Z370 Gaming Pro Carbon AC ($199.99 On Amazon)
G.Skill Trident Z 16GB (2 x 8GB) ($163.95 On Amazon)
Crucial MX300 1TB ($189.99 On Amazon)
be quiet! Dark Power Pro 11 850W ($199.00 On -)
Cooling
Alphacool Eisblock XPX ($73.99 On Newegg)
Alphacool Eiszeit 2000 Chiller ($1076.29 On Newegg)
Thermal Grizzly Kryonaut ($11.99 On Amazon) (Used when Switching Coolers)
PC Case
Microcool Banchetto 101
Graphics Card
Asus RX 560 Strix OC ($139.99 On Amazon)
Monitor Eizo EV3237-BK ($1695.36 On Newegg)
Power Consumption MeasurementContact-free DC Measurement at PCIe Slot (Using a Riser Card)
Contact-free DC Measurement at External Auxiliary Power Supply Cable
Direct Voltage Measurement at Power Supply
2 x Rohde & Schwarz HMO 3054, 500 MHz Digital Multi-Channel Oscilloscope with Storage Function
4 x Rohde & Schwarz HZO50 Current Probe (1mA - 30A, 100 kHz, DC)
4 x Rohde & Schwarz HZ355 (10:1 Probes, 500 MHz)
1 x Rohde & Schwarz HMC 8012 Digital Multimeter with Storage Function
Thermal Measurement1 x Optris PI640 80 Hz Infrared Camera + PI Connect
Real-Time Infrared Monitoring and Recording
Operating System
Windows 10 Pro USB ($199.99 On Amazon) (1709, All Updates)

MORE: Best Graphics Cards

MORE: Desktop GPU Performance Hierarchy Table

MORE: All Graphics Content

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
12 comments
Comment from the forums
    Your comment
  • phobicsq
    Doesn't hwmonitor cost money?
  • FormatC
    HWiNFO is free :)
  • Th_Redman
    Great article of information Igor(and Tom's, of course). I use a number of these stress tests and your article listed some I've never heard of or read about, so thank you.
  • Jay E
    But did you eat the egg?
  • Unolocogringo
    Very nice article to point newbie overclockers towards. You have to have some basic understanding if you want to overclock successfully.
    I have overclocked everything possible since my first overclock.A pentium 75mhz I overclocked to 90mhz. This was mid to late 1996. I learned how to do it from This site. Toms Hardware (sysdoc.pair.com back then).
    Since I overclock every thing to stable 100% load 24/7/365 for Folding@Home and occasional gaming, It must be 100% stable for correct folding results. And of course gaming with my son and grandson.
    I use most of the tests and tools you do, except for the fancy thermal images, to achieve this. Nice to know my testing methods are the Same as yours, but mine last 36 to 48 hours on final overclock settings before being put into service.


    Overclocking is a serious affliction , even my non overclockable SuperMicro 2p server board is overclocked from 2.5 to 3.0 on all 8 cores and folding away for years. :)
    Enjoyed your article and testing methodology explained. Thanks
  • ddferrari
    For my uses (gaming, surfing) I see no reason to push a component to its power or thermal limit via synthetic tests. All that does is shorten its life span. I don't care if my OC fails during a multiple hour, unrealistic load. If it runs fine during real-world usage then I'm satisfied.

    All my components are overclocked, and I test them for stability the old fashioned way: I USE them. They key is to overclock only one component at a time and see if problems arise while gaming. I keep bumping up the OC until an issue pops up- then I know where the maximum lies.

    There seems to be a lot of monkey-see-monkey-do going on around the internet these days.
  • stonedwookie
    We dont care about the stats what we want to know is did the eggs taste good?
    what would you rate the eggs ?
  • FormatC
    The egg got only three-stars rating (3/5).

    The reason why:
    It was simply too long for my taste and it is a real pain to look over such a long time at this egg if you are hungry :P
  • Co BIY
    What is the best thermal paste to use for a mining rig omelet pan?
  • FormatC
    Olive Oil. The best taste :)
  • captaincharisma
    for best results you must play at least 30 mins of final fantasy XV with all settings on ultra
  • themartian
    DiRT Rally is also very good at detecting OC issues of GPU's and CPU's. It has built benchmark with loop functionality. Althrough it doesn't place much load for current hw for the stability testing it works very well.