The motherboard is arguably one of the most important parts of your PC. Some may say it’s the CPU or the GPU, but both would be useless without the motherboard to back them up. As part of the PC Guide Testing lab, we personally test every motherboard we review, to bring you the best opinions on the latest hardware and tech backed up by real-world, hands-on testing.
Our testing suite consists of the best and most up-to-date components around, but because there are a multitude of different sockets and chipsets out there, we need a component set in place for each.
AM5 component set
| CPU: | AMD Ryzen 9 7900X |
| GPU: | Nvidia RTX 4070 Ti |
| Memory: | Corsair Dominator Titanium @6600MHz |
| PSU: | ASUS Thor Platinum II 1000W |
LGA 1700 component set
| CPU: | Intel Core i5-14600K |
| GPU: | Nvidia RTX 4070 Ti |
| Memory: | Corsair Dominator Titanium @6600MHz |
| PSU: | ASUS Thor Platinum II 1000W |
AM4 component set
| CPU: | AMD Ryzen 7 5800X3D |
| GPU: | Nvidia RTX 4070 Ti |
| Memory: | Corsair Vengeance @3200MHz |
| PSU: | ASUS Thor Platinum II 1000W |
The reason we need three sets of testing components is that CPUs of different generations (sometimes) and different manufacturers require separate sockets. You can’t fit a CPU into a motherboard whose socket doesn’t match. This, of course, means that we cannot compare all motherboards directly, but we can compare their values.
For the motherboards we can compare directly, we divide the results up into the respective sockets. (AM5, AM4, LGA 1700, Etc) This can be done because the only non-constant is the motherboard, and any performance difference can be attributed to that alone.
Box contents and setup
Firstly we analyze everything that comes with the motherboard and lay it all out, so you can get a feel of exactly what you get when you purchase the motherboard. It’s then hooked up to our test bench, specifically on the Cooler Master Master Frame. This open-air test bench is easy to access and work on, especially when swapping parts in other benchmarks.
We always use the same hardware that is listed as it’s important for keeping the results fair. We use quite a mix of high TDP and low TDP CPUs, but we use a very capable 240mm Ryujin III cooler regardless. It’s enough to handle all of the CPUs we use and ensures that CPU temperature won’t contribute to any performance losses.
Synthetic benchmarks
We test the motherboard’s performance in 5 synthetic benchmark tests, well, more accurately, we measure the CPU’s performance and track the motherboard’s impact. We calculate a crude efficiency value by taking the PPT of the CPU (the amount of power that the motherboard is sending through the power rails to the CPU), and the Package Power (the amount of power that the entire CPU package is using). Subtract the Package power from the PPT and you have a current that corresponds to loss.
Throughout all this, we also closely monitor the motherboard and motherboard VRM temperatures, to see how well the VRMs are cooled. Temperature also decreases the efficiency of some electrical components, so we make sure to note any correlations there.
Here’s the list of the synthetic benchmarks we use to benchmark motherboards:
- Blender Monster
- Blender Junkshop
- Blender Classroom
- Cinebench r32 (Multi-core)
- Geekbench 6 (Multi-core)
The performance of the motherboard should be the same across the chipset, across the generation in fact. There might be small differences but nobody should have the performance of their CPU diminished because they are not on the best motherboard. That’s what we aim to find out.
While we do keep an eye on performance, an area of which we take great consideration is the usability of the motherboard and its connectivity.
How we score
We have carefully considered the weighting of each section when scoring motherboards, we consider the performance, efficiency, and more. All of our experts agree that this is the optimal scoring, and it is as follows:
Motherboard review scoring
- Performance / BIOS: 40%
- Connectivity / features: 30%
- Efficiency: 10%
- Build quality: 10%
- Aesthetics: 10%
Performance / BIOS 40%
Performance is considered as an average when weighted against other motherboards, if the performance drops below the average, then performance is negatively scored. The BIOS’s ease of use is also influenced by performance, as it governs overclocking and memory speeds. We carefully consider how easy to use the BIOS is, what features it has, and how easy it is to update the BIOS.
Connectivity and features 30%
Connectivity is a very important thing when it comes to motherboard considerations, as you might need a specific set of connections for your intended use. We take into account the type and speed of the connectivity that the motherboard offers, and how plentiful it is.
Efficiency 10%
Now we’re not engineers by any measure, so we have a very crude way of measuring power losses by the motherboard VRMs. We do this by measuring the CPU PPT (the amount of power the motherboard is sending to the CPU) and the CPU package power (The amount of power that the CPU is using). By taking these values and subtracting the Pacckage Power from the PPT, we get a rough efficiency value.
Build quality 10%
Build quality ties somewhat into features as the main feature of a strong motherboard is PCB armor. Some high-end motherboards come with armor either on the front or back to protect against shorts and sometimes serve as a method of heat dissipation. Other build-quality features encompass PCB layer count, quality of materials, and robustness. (not that we try to break anything)
Aesthetics 10%
The looks of a motherboard might be more important to those focused on the glory and showmanship of PC building, but we here at PC Guide are mainly focused on the performance of the components we use. We do understand that looks are a consideration, so we weighted it a good 10% towards our overall score.
