1.   A floating point accelerator, larger disk and memory expansion are options.

2.   This is a project aimed at building a computer which can process one million floating point operations per second.

3.   Floating point variables are stored in binary floating point format with their least significant byte first.

4.   If the exponent of a floating point number is zero, the number is stored in integer format in the mantissa.

5.   If the exponent is not zero, then the variable has a floating point value.

6.   And given the public relations nightmare for Intel, it is unlikely that the floating point unit problem, at least, will occur again.

7.   Both bugs affected so-called floating point calculations, which improve the accuracy of complex computations.

8.   Intel said its new supercomputer would be the first to achieve the goal of calculating more than a trillion floating point operations a second, known as a teraflop.