Peter Gutmann writes in an article quoted by Christian Wettergren

The greater the amount of time that new data has existed in the cell, the more the old stress is "diluted", and the less reliable the information extraction will be. Generally, the rates of change due to stress and relaxation are in the same order of magnitude. Thus, a few microseconds of storing the opposite data to the ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ currently stored value will have little effect on the oxide. ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^

Phill Hallam writes:

If the power is cycled as opposed to turned off only then a memory self test program will probably erase the data.

Assuming Peter Gutmann is correct, a memory test program "probably" won't do much. Of course, you data must be worth quite a pretty penny for an attacker to attempt to recover data from the oxides on the cells in your RAM. andrew