Nice work Scotty!the OCZ passed stress test at 2.3v with 2:1 ratio (1066)

Can they go any lower on the vDimm or?
Nice work Scotty!the OCZ passed stress test at 2.3v with 2:1 ratio (1066)
Can they go any lower on the vDimm or?
Ok well it's good that you tried, I guess we need to accept the OCZ sticks are very thirsty for voltage, I've not ever had DDR2 memory that needs so much voltage but I guess all sticks/batches are very different.I cant lower the voltage at all
So far mem test is fine for 20 mins or so....needs a little longer!!
Ok no problem, were not quite done testing the memory at DDR2-1066 using the [2:1] multipler with a 266MHz-FSB but I can tell you that if the sticks pass fine with the 266MHz-FSB and [2:1] memory divider @ DDR2-1066 but they fail when using the 333MHz-FSB [8:5] @ DDR2-1066 then the problem is not the memory . . . . it's the Northbridge, but we will get to that soon enough!Tried FSB at 333 and 8:5 ratio but errors again even at 5-5-5-15
Hello nubonamission,I love O/Cing and would love to learn a lot more.
Ok well it's good that you tried, I guess we need to accept the OCZ sticks are very thirsty for voltage, I've not ever had DDR2 memory that needs so much voltage but I guess all sticks/batches are very different.
Hey nubonamission,Sure man, that sounds great. Appreciate it
Heh no problem scotty!I feel that the OCZ mem isnt performing anywhere like it should (2.1v 5-5-5-15 @ 1066) and as its only 3-4 weeks old I want to send it back.
This Crucial memory above was sold to run at a rated speed of 333MHz (DDR2-666) with 3-3-3-x timings @ approx 2.1 volts, but with a little reading up and tinkering in the BIOS I managed to overclock it to 600MHz (DDR2-1200) with 5-5-5-15 timings @ approx 2.06 volts, it was left testing in DOS based Memtest for over 20 hours and passed with zero issues. I am not suggesting you leave yours testing for so long but I prefere to be really really sure a given frequency/timings/voltage is rock solid stable, as soon as the Memtesting is complete I go straight in Real-World usage (gaming, crunching etc)CRUCIAL ANNIVERSARY DDR2-666
![]()
I assume you have taken the memory of the [1:2] divider before trying for 333MHz-FSB right? . . . otherwise the RAM is being asked to run at 666MHz (DDR2-1333)I get no joy booting at FSB333
I've not ever experimented with adjusting Windows virtual memory but are you sure that it isn't best to leave that alone for the moment? i.e let Windows handle it?I dont usually have virtual mem turned on
Sounds like either the memory or the Northbridge overheated . . .I tested like a man posessed at 333FSB and got an error at about 5 hours in last night!!!
You have wandered into an advanced area again heh . . . Performance Level (aka tRD) is part of the Northbridge, as you found out lower is better (reduces latency), the motherboard will set tRD itself according to memory frequency and Cas timing i.e lower the CAS and raise the DDR2-Frequency and the tRD will set itself lower. It's definitely something thats worth looking into and tweaking but it's a separate testing session from the memory really! . . . . tRD is fed by vNB (NB volts) . . more voltage helps the northbridge cope with low tRD however the extra voltage also brings extra heat which can also cause errors . . .I have brought the Performance level [down] to 6T from 7T to get much better latency but I may have to [raise] it again.
Performance Level (tRD) is unaffected by vDimm (Memory Voltage)Gonna tweak the NB [voltage] up a tad and try again. Failing that I might up the Vdimm up to 2.15v. I can see this is gonna take a while![]()
It is although the sticks I have are supposedly hand picked by Crucial for a limited edition they released in 2006. The sticks you are looking at may well perform the same but no guarantees.I can get hold of some Crucial Ballistix 2x1GB PC2-5300 DDR2(BL12864AA663) on the cheap, is that the similar to what you used?