Labs
Motherboards
[PC Pro]
This month, we put 11 motherboards to the test. It's a complex task, as every one of these boards offers a unique combination of features. Measurements and benchmarks can therefore tell only part of the story: choosing a motherboard is often less about performance than capability.
But that's not to say there's no room for empirical analysis. Over the following pages, you'll find a detailed discussion of each board's various strengths and weaknesses, coupled with scores allocated to each based on a combination of subjective judgments and objective measurements. And, as usual, at the bottom of each review you'll find marks in four categories: Performance, Features & Design, Value for Money and a final Overall score.
Performance
Each board's Performance score is calculated from three factors. The first is a measurement of actual system performance. You might imagine that this is wholly down to the CPU, RAM, chipset and graphics card, but it's the motherboard chipset that handles communications between all of these components. And the various boards in this month's test use a total of nine chipsets from three different manufacturers, so it's to be expected that some boards will manage to squeeze a little more performance out of a system than others.
To measure the real-world performance gap, we build a standard system based on each motherboard; depending on the board's requirements, we install either an Intel Core 2 Duo E8500 or an AMD Athlon 6400+, with 2GB of either DDR2 or DDR3 RAM. We then run our standard benchmarks and record the difference in scores achieved.
The results are expressed in the graph above as a proportion of the average score for that CPU. As you can see, we observed a small but real variation of around 2% between boards. Since the difference is so small, it has only a minor effect on each board's Performance score. However, it's interesting to note that the slowest and fastest boards are both DDR3-based, indicating that, while the newer standard can achieve higher speeds than DDR2, simply using DDR3 won't magically improve performance.
The Performance score is also affected by power consumption. To measure this, we use each board as the basis of a minimal system, using a 65W TDP processor (either a Core 2 Duo E6700 processor or an AMD Athlon 64 X2 4200+), two 1GB DIMMs and a passively cooled Nvidia 8500 GT graphics card.
We don't take our readings while running Windows, for the simple reason that Windows' many background services can cause a seemingly idle computer's power draw to fluctuate by more than 5W - enough to obscure the differences between several boards. So, instead, we take our reading from the main BIOS screen, after waiting five minutes to ensure the components are completely stable and idle.
The results are shown above, but when reading the graph remember that some of this month's boards feature integrated graphics, meaning they can run without a graphics card. Real-world power consumption could thus be around 30W lower than the figure quoted above, although, of course, you won't be able to get the best out of modern games.
Finally, the Performance score also takes into account how far the BIOS lets you tune your PC's performance. Top marks go to boards that make it easy to monitor and tweak settings such as the RAM frequency and the CPU multiplier. Lower marks go to less user-friendly or fully featured BIOS implementations. Our scores are based on the current BIOS version as we went to press, but all manufacturers offer downloadable updates. So, if a board you like is missing a particular feature, check the website in case it's since been added.
Features & Design
Our score for Features & Design is self-explanatory. As you'll see from the table opposite, no two boards offer quite the same range of features and expansion options. Our rating represents how well featured each board is, taking into account issues such as the number of USB connectors, PCI slots and SATA channels, and whether the board has multi-GPU support. We also consider any thoughtful design aspects - clear labelling, internal power switches and so on. These aren't so heavily weighted, though, as once your system is up and running you may never need them again.
Value for Money
The Value for Money score reflects each board's price, but top marks don't just go to the cheapest contender. We take each board's score for Performance and Features into account, so a high score here means a board is particularly good for its price.
Overall
Finally, our overall score is an average of the other three, although it may be higher or lower than expected due to rounding.





