The freedom to push your computer well beyond the stock specifications is the best part of PC gaming. If you have been facing niggling hardware troubles or otherwise want to learn how to use advanced overclocking tools to optimize your gaming setup without necessarily overclocking your components, we show you how to make use of RivaTuner Statistics Server (RTSS) to monitor your gaming hardware performance.
What Is RTSS?
RivaTuner Statistics Server was initially a companion to the RivaTuner open-source GPU overclocking tool designed for Nvidia GPUs, with limited support for ATI/AMD hardware. However, it has now become a standalone tool offering hardware and frame-rate monitoring capabilities, with additional support for frame limiting and video capture.
RivaTuner itself has since been assimilated into virtually all major GPU overclocking tools, such as MSI Afterburner, EVGA Precision X, and ASUS GPU Tweak. This guide is based around MSI Afterburner, which comes bundled with RTSS. Here’s what you need to do to get started.
Install and Configure RTSS
1. Download the latest stable release of MSI Afterburner bundled with RTSS.
2. Install the tool and be sure to select RivaTuner Statistics Server when you are prompted to choose components during the installation. It’s not possible to monitor critical hardware parameters without RTSS.
3. Open MSI Afterburner and dial in the overclock settings for the GPU. Alternatively, you can leave everything unchanged at the default settings and head straight to configuring RTSS by clicking on the Settings button represented by the gear icon.
4. Navigate to the Monitoring tab in the Settings window. The tab contains a list of hardware parameters relevant to the CPU and GPU.
5. Select the hardware parameter you wish to monitor. Make sure the checkmark beside it isn’t greyed out. Monitoring these parameters consumes CPU time, so choose what you need if you are concerned about the performance of RTSS.
6. With the desired parameter highlighted, tick the first checkbox below the list of hardware parameters named “Show in On-Screen Display.” The drop-down menu next to it allows you to choose whether you want this data displayed as text, a graph, or a combination of both. Repeat this procedure for all parameters you want to monitor.
7. Click on the OK button to save these settings. The next time you run a game, the selected parameters will be overlaid on the top left of the screen.
Whether you’ve opened RTSS with MSI Afterburner or separately, you can customize a lot of things with RTSS itself. Open the separate RTSS app either through the system tray or by navigating to the executable on your PC. (You can also search for “rivatuner” through the Start menu.)
You’ll see plenty of options in the Rivatuner app, such as starting it with Windows, the dimension of the OSD (on-screen display) and so on.
One of the more interesting features here is that if you click Add at the bottom-left corner, you can add RTSS profiles for specific programs and apps. I, for example, added the game Hades here.
With Hades selected, I can now change the different RTSS settings specific to that game. A good option for games, for example, is to enable Stealth mode, which will prevent it from interfering with anti-cheat software in online games. Other games, meanwhile, will require the “Custom Direct3D support” option to be enabled.
There’s a lot here, so have a play around and do some customizing!
Using RTSS to Monitor Hardware Performance
RTSS comes with a long list of features and capabilities, but we are primarily interested in the capability to monitor hardware parameters to improve the gaming experience. That involves keeping an eye on numerous statistics involving thermal, utilization, power, voltage, and clock speeds for the CPU, GPU, and memory subsystems. The following lays out the parameters you should take note of.
Keeping an eye on GPU temperature relative to the ambient temperature gives a good idea of the overall health of the cooling system of your graphics card. An overheating GPU is a sign that remedial measures such as GPU heatsink cleaning and thermal paste reapplication is required. Monitoring the Fan Speed to verify proper operation along with using aggressive fan profiles should also help.
Modern GPUs tend to take manual overclocking control away from users and instead base it on thermal headroom. Getting the most out of your GPU is no longer restricted to prevent thermal throttling, but to ensure that the GPU runs as cool as possible. Case in point: both Pascal and Turing Nvidia GPUs provide loads of extra boost clock speeds with every five-degree drop in temperature. The Temperature Limit parameter is also a great means to know if your GPU performance is hit hard on account of thermal throttling.
The same reasons apply to CPU temperatures as well. However, processor clock speeds aren’t as temperature-dependent as that of GPUs. Thermal throttling is what we are looking for here. Be sure to use the CPU1 temperature-monitoring option. This is the accurate temperature reported directly from the CPU die. The option labelled CPU temperature is reported from software and isn’t quite accurate.
CPU and GPU Usage
These parameters are extremely important. Depending on how optimized a game is, CPU utilization will range from 40 to 80 percent for most gaming rigs. However, if your CPU is at 100 percent, that means it is the weak link and that your GPU is sitting idle, waiting for the CPU to finish its job. GPU utilization will be less than 100 percent in this case. You will get much better performance by upgrading to a more powerful CPU.
The converse holds true for GPU as well. However, your GPU should hit 100-percent usage under normal conditions. Otherwise, it is being bottlenecked by the CPU.
This refers to the video memory (VRAM) and shouldn’t be confused with the RAM Usage option which pertains to the main system memory. Certain game settings, such as shadow quality, ambient occlusion, multi-sampled AA, and anything that uses normal maps will require extra VRAM. Once VRAM usage exceeds the max installed memory on your GPU, the framerate will drop catastrophically.
Some rare games, such as the new installments in the Resident Evil series, warn of potential memory bottlenecks, but for most other games, keeping an eye out for memory usage is critical if you experience sudden framedrops and/or microstutters.
Consider keeping an eye on this parameter if your framerates drop precipitously without any explanation. BUS Usage is the utilization of the PCIe bus, which is the interface governing the communication link between the CPU and GPU. No current GPU is capable of saturating the modern PCIe 3.0 x16 bus, so this parameter should register low percentages under full load. If BUS Usage reads 100 percent, you’re looking at a problem with the PCIe bus.
This can be caused by a number of factors ranging from driver issues, incorrect BIOS setting, and even debris in the PCIe slot or an incorrectly-seated GPU preventing optimal electrical contact between the GPU and the motherboard. This problem is also caused by cheaply-made PC cases with misaligned rear I/O mounts that cause GPU seating issues. This can be further verified by running a GPU info/diagnostic tool such as GPU-Z.
Once you know the maximum clock speeds of your CPU/GPU core and memory, any discrepancies therein are signs that the components are either experiencing thermal throttling or facing underutilization due to other underlying problems. This parameter also helps diagnose memory issues. Memory cooling on most GPUs is weak, so overclockers should watch memory clock speeds closely for signs of thermal throttling and dial down memory overclocks to prevent damage to the GPU.
No Load/Power Limit
Extremely useful for Nvidia laptop users. This will tell you if there’s some problem with the power-saving feature of the GPU. If no load limit shows while you have poor framerate, there’s some issue with the GPU. It could be the PCI-E bus or some power-saving feature, but you at least know what to troubleshoot. Power limit is important for overclockers. This will tell you whether the CPU/GPU is being throttled because the power delivery subsystem cannot provide enough wattage to hold the overclocks.
No Load/Power/Voltage Limit
These nifty monitoring parameters are a godsend for diagnosing odd GPU behavior and performance issues. No Load Limit basically tells you when the GPU is idling due to lack of processing load. You know there’s something amiss with your GPU when the No Load Limit pops up during actual gameplay while causing a severe drop in framerate. For example, Nvidia laptops using hybrid GPU configuration will hit the No Load Limit if the onboard GPU is used instead of the discrete GPU.
Power and Voltage Limit indicators are relevant to overclockers. Hitting these limits is a sign that your GPU is being throttled because the power delivery subsystem cannot provide enough wattage to hold the overclocks.
Keeping an eye on your hardware performance gives you a fair idea of the overall health of your system, which is critical to diagnosing any performance issues and underlying hardware troubles in time and without much heartburn. For more tips to monitor hardware performance, see our guide to stress-testing your GPU using Furmark, as well as stress-testing your CPU using Cinebench. Also, here’s how to stress test a graphics card on Linux.