I somehow cannot believe that nobody had mentioned the first significant instance of similar crash at least for the DOS days: Borland's CRT.TPx/conio.a which co

I somehow cannot believe that nobody had mentioned the first significant instance of similar crash at least for the DOS days: Borland's CRT.TPx/conio.a which consistently caused division by zero on anything faster than Pentium MMX. At the time the wisdom was that anything over 200MHz is too fast, but in reality anything AMD with "Pentium Rating" over 200 was fast enough to cause consistent crashes and 266Mhz Pentium MMX was slow enough that it mostly worked (and anything i686 consistently crashes).