Well, a bit of info about those PQI sticks on P4C800-E:
Those are PQI PowerMemory 1GB Sticks (aka MD441GUOE - no heatspreaders). I used DRAM voltage of 2.85v, since at lower voltages the RAM didn't always want to startup at all, when overclocked.
In single channel I could get one stick of four to pass MemTest at 225MHz 3-3-3-6, I didn't try to load Windows and check the stability, since Single Channel isn't an option for me.
At 230MHz both 3-3-3-6 and 3-4-4-8 gave me errors at the 98%-99% of Test #5 - about 600 errors.
The 2nd and the 3rd sticks failed the Test #5 the same way the 1st stick did, but at 225Mhz, and passed at 220MHz.
The last stick didn't want to OC around 220MHz at all, the system got stuck on the POST screen, so I didn't test it anymore .
The strange thing is, three sticks of four fail at the 2-3 last percents of the Test #5 (they don't fail the tests #1-#4, but I didn't go further than #5, probably they fail some following tests too, I really have no much time for this), what is that? What could be the reason of systematic failure at the end of the Test #5? They don't overheat - that's for sure.
Ok, that's not all, there're 4 more sticks left in the store, I'm going now to exchange the 3 lazy sticks for the new ones, and maybe I'll be lucky to find another stick that will do 225MHz, since I don't need more: 282MHz / 5 * 4 = 225.6MHz.
Another strange fact is, in Dual Channel the sticks don't like the 2:3 divider - 282MHz / 3 * 2 = 188MHz - underclocking actually - but the sticks didn't pass the Test #5 this way... Why?
Still, the pairs of 1st + 2nd or 1st + 3rd sticks are SuperPI stable at the 250MHz with the 5:4 divider (DDR400) and 3-3-3-6.
By the way, Crucial's Z503 didn't start at 282MHz neither with 5:4 divider, nor with 3:2.
But a pair of Kingston HyperX PC3200 256MB (BH-5 - my 1st RAM) worked perfectly 2-2-2-5 with any divider at any frequency around 200MHz.
What do you say, guys, anything you can advise? Thanks anyway for reading all this.
Lood007 - thanks, man .
Bookmarks