Thanks for your comments. I too have my suspicions about MaxxMam. I actually had taken 8 consecutive runs of it within a span of 7 minutes with nothing else running. And I got a fair amount of spread on the data. Worst of all, if the test completes in lesser time, then better is the overall transfer rate. If it is pushing the same amount of data around each time, then there is over 500 milliseconds of difference between the fastest and the slowest time. That means it is suffering from delays it is not accounting for (were ints disabled during the move?) This is just speculation as I do not know what it is doing inside.
Here are a couple of screenshots about it.
But, I am happy. I noticed a few things during the OC :-
1. I could push the FX-6300 to over 5.2 GHz rather easily and it would pass MemTest86 at that speed for hours.
2. At 5.2 GHz, I needed close to 1.55v on the cpu but only about 1.4v on the CPU-NB to pass MemTest86.
3. I had to reduce it down to 3.5 MHz to install Win 7.
4. To stabilize with Prime95 stress test at 5 GHz, I needed 1.55v on Cpu with CPU-NB voltage asking for over 1.45 or more. Even with a Xigmatec tower and push pull setup and with 10F mountain air blowing directly over the open rig, I could not stop the temps hitting 70C within minutes. The temp ramp up was that fast! That's when I called it a day.
5. At 4.8 GHz, it is rock stable with normal 73F indoor air and I can still do all my chores with some transcoding as well with the temps staying below 55c.
6. I tried to OC the 2133 rated DRAM to 2190. Either the M5A97 R2.0 motherboard or the F3-17000CL9Q-16GBZH DDR3-2133 from GSkill prevented me from doing that despite my best efforts. In other words, there is almost no headroom in that combination above DDR3-2133.
Given that the rig is destined to my in-laws who like to do Skype and play some silly card games on the net, it should be plenty. They just wanted a faster PC than their old Celeron!