Articles: Mainboards
 

Bookmark and Share

(1) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 ]

There is another way. We can disable Intel Turbo Boost technology, so the CPU clock frequency multiplier will stop increasing. And with the x24 multiplier we can get 4.2 GHz resulting clock if we raise the base clock frequency to 175 MHz. In this case the memory frequency will also be higher – 1400 MHz. However, we think that the third approach when we lower the processor clock frequency multiplier would be the best. We disable Intel Turbo Boost technology, lower the multiplier to 21x, but increase the base clock to 200 MHz. In this case, we can even leave the memory frequency at its nominal value of 1600 MHz. So, during our tests we used this particular approach and Gigabyte GA-X58A-OC did very well, although we did have to spend some time on finding the optimal voltage settings.

We do not approve of “screenshot overclocking”, when you can take a screenshot of the CPU working at a certain frequency, but there is no real long-term stability. This type of record-breaking hardly has any practical value. We always overclock our systems with intention to use them in this mode. Note that we never use any shortcuts like disabling additional controllers, for example. And we always do our best to keep all power-saving technologies up and running, if circumstances permit. So, in this case power-saving was in place and the processor core voltage and clock frequency multiplier dropped in idle mode.

We are going to compare Gigabyte mainboards against Intel DX58SO2. This board made a great impression during our test session. It was convenient to work with, was very energy-efficient, overclocked our quad-core processor very well. It did just as good with a six-core processor, and as it seemed to us back then, it repeated the results previously obtained on Gigabyte GA-X58A-OC.

Just like in the previous case all power-saving technologies worked normally. The only difference from the GA-X58A-OC mainboard was the fact that all these technologies were enabled by default.

However, the gaming Gigabyte G1.Sniper board turned out a disappointment. When we increased its base clock to 200 MHz, it couldn’t even load the operating system, not to mention any tests. When we overclocked it using a different method, at 175 MHz base clock and 24x multiplier, the OS loaded fine, but we couldn’t get our system to run stably in this mode. As a result, we had to work with the 168x25 overclocking method, which was the least efficient of all because of higher power consumption in idle mode and too low memory frequency.

The board still managed to overclock our test CPU to its maximum without losing processor power-saving technologies, but it is hardly a consolation in this situation.

By the way, they didn’t have to cover the O.C. Button on the mainboard back panel intended for automatic system overclocking with a protective plastic cap. This button turned out practically useless. You can press it without fear, because in this case the base clock is only increased from 133 MHz to 140 MHz. This 5% boost is hardly of any practical value and you won’t even notice a speed increase. So in the end we were certain that Gigabyte G1.Sniper mainboard would end up scoring really low because of all the little issues we pointed out. However, the results of our performance tests changed everything.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 ]

Discussion

Comments currently: 1
Discussion started: 08/04/11 08:06:07 PM
Latest comment: 08/10/11 05:12:01 AM

View comments

Add your Comment