I have managed to kill ram by running the frequency too high without giving it extra voltage, so thats the experience I was drawing on with my above statement. Not sure that you're 'incorrect' tho.
Not wanting to detract from the OP's post, but I find that to be an interesting statement, but one that I have a hard time accepting at face value.
If ram is run outside of specs (overclocked), without giving it add'l voltage, then as a general rule it will simply fail and produce errors, errors which can easily be identified by proggies such as Memtest. When the clock speed is reduced to a level that the modules can operate at (within the voltages provided), it will return to proper operations. I can't thing of anything that would happen to it when running at higher frequencies that would "kill" it, except perhaps a poorly ventilated system that would overheat. Why would you draw the conclusion that the sticks died as a result of being run at frequencies over stock and not something else?
Take, as an example, a module of DDR1 manufactured with Samsung TCCD ICs. These ICs were rated by Samsung to be a DDR466 part (PC3700), yet all of the major memory players produced modules using these same ICs in sticks rated as (1) PC3200 (underclocked, but tighter timings), (2) PC3700 (as spec'ed), and (3) PC4000, PC4200, PC4400 and PC4800 (all of these were "overclocked"). All used the same ICs and in many cases, the same PCB (BrainPower).
If simply running them at higher speeds than the manufactured rated them at would cause their death, then none of those other binned modules would have survived in the market, yet, they did and are still out there running in many rigs. The same story could be told for Hynix and Winbond based chips.
Similar DDR2 examples exist based on Micron D9xxx chips.
Keep in mind that sticks fail for a variety of reasons that may not have had anything to do with the speed at which they were running...they may have died within the same timeframe even if never overclocked a single MHz.
Add'l voltages, outside of specs and beyond might cause it die, but I can't quite buy that higher frequencies alone would cause their demise.
Sure would appreciate if anyone else can chime in on this because I'm curious.
Has anyone else had ram die from running it faster than what it was spec'ed for, but without giving it add'l voltage? I don't mean fail Memtest...I mean flat out die.
To the opening poster...what motherboard and ram are you using?
Are you wanting to overclock your processor or will you keep it at stock? If you are not going to overclock, then as was already posted, with the 1:2 divider you could run your processor at 200MHz (x4 = 800 FSB) and your memory at 400MHz (x2 = DDR2-800).
If you share your intents and desires, along with your system specs, I'm sure that you'll get all the help that you need.