Opteron Floating point bug

Seems there are a number of dodgy Opterons loose which develop floating point errors if subjected to tight code loops which are repeated for hours on end. Some sort of thermal instability causes corruption.

Wonder if that would affect Distributed computing at all ?

The glitch is not thought to affect real-world applications. “It’s very hard to imagine this type of code in our [financial services] environment,” the source said. “The only thing I could think that would be coded this way would be some type of strange cipher code. For example, any type of ‘for’ loop that uses a compare operation would not have the problem.”

Full article Here

Glad I got Dual cores! :wink: