3 ms·

NAND flash gets baked, lives longer

• PakG1 14y ago
  Did I read 800C properly? Even if it's only for a minuscule moment, isn't that dangerous? Especially if there are environmental factors that can unpredictably affect the process? After years of being conditioned to try to cool my hardware as much as possible, it's a bit unsettling that I'd want to heat it up, especially at such a high temperature, even if it's for only a tiny moment. Excuse me for my ignorance, but I'd like someone to explain to me why this is OK.
  • ars 14y ago
    It's deep inside the chip - there is probably insufficient energy for it to still be hot once it makes its way out of the package (even if it were to get stuck in the on state). However, 800C is really hot - there's probably a reason this is not a shipping product yet. If I were making one of these I'd charge a capacitor, then use the energy in a burst to get really hot. This smooths out the power draw and makes it completely impossible for it to stay stuck on.
  • ghshephard 14y ago
    Temperature isn't as important as heat energy in this scenario. We are likely talking about an absolutely minuscule amount of energy that is easily dissipated.
• richardjordan 14y ago
  Interesting, though years since I did any materials science so just reading as a layman. Still, I'm surprised this wasn't tried before. It's been years since I read the book Inside Intel but I do seem to recall that an extra dose of cooking led to dramatic drops in failure rates for their chips. I suppose I'd have thought that'd lead folks in the space to try all kinds of things like this all the time.