CPU and Memory Overclocking
We tested Asus P5QL-E mainboard in an open testbed with the following hardware and software components:
- Mainboard: Asus P5QL-E, rev. 2.05G, BIOS 0801 from 10.14.2008;
- Intel Core 2 Duo E8400 (3.0GHz, 333MHz FSB, 6MB L2 cache, Wolfdale, rev. E0, 1.25V Vcore);
- Intel Core 2 Quad Q9300 (2.5GHz, 333MHz FSB, 6MB L2 cache, Yorkfield, rev. M1, 1.175V Vcore);
- Memory: 2 x 1024MB DDR2 Corsair Dominator TWIN2X2048-9136C5D (1142MHz, 5-5-5-15, 2.1V Vcore);
- Graphics card: ATI Radeon HD 4870 512MB;
- HDD: Samsung SP2504C (250GB, SATA II, 7200 RPM, 8MB, rev. A);
- CPU cooler: Zalman CNPS9700 LED;
- Thermal compound: Noctua;
- PSU: Antec NeoPower HE 550 (550W);
- OS: Microsoft Windows Vista Ultimate SP1 (32-bit).
We have just seen from the previous chapter of our review that Asus P5QL-E mainboard ahs the entire range of features necessary for successful processors and memory overclocking. To find out the maximum bus frequency when the board remains stable with a dual-core Intel Core 2 Duo E8400 processor, we lowered its clock frequency multiplier to the minimal setting of x6, lowered the memory frequency, increased Vmem and Vmch.
We decided to start with relatively low frequency of only 450MHz, but the mainboard refused to boot. After a while we found out that the board would only boot at 430MHz FSB. Even without any voltage increase our processor can overclock to 450MHz bus frequency, so we returned its multiplier to the nominal x9. The board could boot the OS, but refused to reboot. We couldn’t reach stability even at 425MHz FSB, but at 420MHz the system was fully stable and passed all the tests.
An Intel P45 Express based mainboard with the extensive functionality of Asus P5QL-E could easily go past 500MHz FSB. Not the best model could probably fall a few MHz short of that number, however, 420MHz FSB is way too lot for an operational bus frequency. Maybe our sample was not the best one for overclocking experiments? However, after looking through some related forum posts I arrived at the conclusion that almost all mainboards on Intel P43 Express chipset from various manufacturers can only overclock to 420-430 MHz FSB. This could be the greatest difference between the two chipsets from the same family.
It was pretty disappointing, but we could only work with what we had at our disposal. Now we have to try increasing the memory frequency that so far has been working at very low frequency of 840MHz. I would like to remind you that our DDR2 Corsair Dominator TWIN2X2048-9136C5D memory modules can theoretically work at 1142MHz with 5-5-5-15 timings and 2.1V voltage setting. The actual values are usually a little lower, so I wasn’t surprised with instability at 1119MHz. The system refused to start at all at 1053MHz memory frequency, which could also happen because of “inconvenient” dividers. However, even when we lowered the memory frequency to 1009MHz – an absolutely real and attainable value – we still couldn’t achieve stability. So, we had to go back to where we came from: 840MHz. Unfortunately, the board works with the memory as poorly as it overclocks CPUs.
Let’s try optimizing the memory performance at least a little bit to achieve higher speeds. First let’s lower the timings from 5-5-5-15 to 4-4-4-12. It worked.
However, the mainboard set Performance Level 13 by default, which was way too high. Using Ai Transaction Booster parameter we lowered this setting to 8, which immediately affected the performance: read speed increased and latency reduced.
We have one more option to check out: DRAM Static Read Control, DRAM Read Training and MEM. OC Charger – the BIOS parameters that help improve the memory performance at high frequencies. We don’t need any help at ridiculously low 840MHz, so we set them to just the opposite: Disabled instead of Enabled and the other way around. Ai Clock Twister will be set to maximally aggressive setting: Stronger. These measures help improve the situation a little more.
Now looks like we have done everything we could and the system is ready for a competition. However, we discovered one very unpleasant issue: every “cold” start always failed. The system would refuse to boot, reset all parameters and report over-overclocking upon next restart. But then it would work just fine and pass all tests without any settings changes. I have to admit that this is hardly normal, so we had to lower the bus frequency to 415MHz. Only then we could get the system to boot fine, work stably, restart without any issues and Asus P5QL-E shut-downs.
Before we got down to performance tests, we decided to see how well the board could handle quad-core processors overclocking. Our system equipped with an Intel Core 2 Quad Q9300 processor would boot only at 430MHz FSB, but reported an error on restart. We have already experienced all that during our dual-core CPU overclocking experiments, so to save some time and effort we went on to 415MHz FSB right away. Tests confirmed system stability; the system would boot, work, restart and shut down flawlessly.
This is a very low result, of course, but it was good to see that the board could overclock quad-core processors the same way it could overclock dual-core ones, and not any worse.