The Official OCZ VX Thread - Strictly Test Data Only!!!

[conrad.maranan]

Here is my list of parts that will have to be RMA'd sometime in the coming week:

- AMD Athlon 64 3000+
- OCZ DDR Booster (v2.0)
- OCZ PC 3200 Gold Edition VX (2x512MB)

My CPU has had trouble pushing two modules since day one. I've been holding onto it because it's such a good clocker, capable of 2610MHz 24/7 stable with the retail HSF. Today brings me sad news of a completely useless memory controller. Depsite the fact that two DIMMS are populated with memory, my system can only recognize one channel. Here's a look at 2x512MB. Notice how only 512MB and one channel is recognized. Trust me. I've got two modules in place and the motherboard is not at fault. I picked up another K8N Neo2 Platinum this morning only to discover the same results.

One VX module clocks approximately 10MHz higher than the other. The slower module will always get the same exact error at the same exact memory address.



I started out my tests running with a CPU multiplier of 9. As I continued to increase the frequency for 1:1 operation, I had to keep dropping the multiplier down.

My DDR Booster is also very unreliable. Why? I do not know. I have two 92mm Sunon fans that harness a more-than-sufficient amount of power from the Wind Gods, directly forcing air-conditioned winds onto the Booster and memory modules. The Booster will not hold any voltage setting without me having to fiddle with it every 30 minutes.

On a final note, at the very bottom of the back page of the Booster installation guide, you may read the following:

7. AMD recommends that Athlon64 CPUs not be subjected to VDIMM voltages above 2.65V. Operation above this voltage may shorten the life of the Athlon64 and/or cause permanent damage, as well as possibly voiding your warranty.

I'm pushing 3.5V to 3.6V, so go figure. Running tight 2-2-2 timings at high frequencies is extremely hard on the memory controller. Tight timings + High frequencies + High voltage = Tough luck, kid.