Excel 97 should take anything less than 10 to the negative 307th power and automatically round it to zero. Instead, somewhere around 10 to the negative 1 billion--that's a decimal point, 1 billion zeroes, and a one--the program starts returning an answer of "10."
The problem was brought to light by NEWS.COM reader Dale Anson of Omaha, Nebraska, who was doing computations related to his job. He acknowledged that not many Excel users are likely to be affected by the bug but surmised that these types of calculations in the scientific community "are not that uncommon." Others, however, are scratching their heads.
"I can't imagine a setting or physical description of the universe that remotely resembles that number," said Stanford University professor of mathematics Gregory Brumfiel.
Microsoft Office product manager Matthew Price confirmed that the problem exists and that the company is looking into the cause. He stressed that Microsoft takes any sort of calculation error very seriously but also couldn't help but wonder who the bug would affect.
"Excel is used for all kinds of calculations, but this is so incredibly small," Price said.
Following guidelines for scientific computation set by the Institute of Electrical and Electronics Engineers, Excel doesn't even do the math for exponents below -307, Price said. Instead, it automatically rounds to zero. For exponents above 307, Excel automatically returns the "#" symbol for infinitely large numbers.