Re: iNav causing hard reset - SOLVED

[Joe Pluta <joepluta@xxxxxxxxxxxxxxxxx>]

Buck wrote:
> I've seen odd things happen by bumping just the frequency.  A Pentium 
> that would run Prime95 for 48 straight hours and blue screen crash when 
> transferring video via Firewire.  Set the clock back to nominal and no 
> more crashes - only change to the system.
>
> There's no guarantee that bumping the frequency will do odd stuff; you 
> just need to test, and when you find a strange thing, dial it back as a 
> diagnostic.  That's easier to do on a home system than a server/work 
> system...
>   

However, there's also no guarantee that the exact same thing you 
described above wouldn't happen with a chip at its nominal frequency.  
There's nothing about changing the bus frequency to a number greater 
than that printed on the chip's box that causes bad things to happen to 
the chip.  Yes, there is a threshhold at which every chip fails; but 
what that number is can only be determined through testing.

From what I've seen and read, memory is more likely to go casters up 
with a smaller overclock; in my case I was able to bump my CPU frequency 
more than 50% but my memory barely 12%, and that's seems to be in line 
with what others see.

Now, the firewire is an additional joker in the deck.  Firewire by its 
nature is very timing dependent, and it may well be that your firewire 
port was less amenable to overclocking than your CPU.  As I noted 
earlier, my graphics card wouldn't handle any overclocking whatsoever.  
As soon as I raised the system bus speed, the graphics card refused to 
play nice and forced the motherboard to cold reboot to factory speed 
settings.  Not nice (and this is on a Gigabyte motherboard, a board 
designed for overclocking).

Whether it's easier or harder on a work system depends on your work 
system, but it seems to me that the more productive you are, the better, 
and we all know that a faster workstation makes Johnny a more productive 
boy.

Joe