Pentium FDIV bug
|
|
On October 30, 1994, Professor Thomas Nicely of Lynchburg College reported a bug in the Pentium floating point unit. He reported that certain division operations returned a value which was wrong by a very small amount. This result was quickly verified by other people around the Internet, and became known as the Pentium FDIV bug (FDIV is the x86 assembly language instruction for floating point division). Other people found division problems where the result returned by the Pentium was off by up to 61 parts per million. Note that this problem occurred only on some models of the original Pentium processor. Any Pentium family processor with a clock speed of at least 120 MHz is new enough not to have this bug.
This report stirred up a huge controversy. Intel at first denied that the problem existed. Later, Intel claimed that it was not serious and would not affect most users; however, people who could prove that they were affected would get their processor replaced by Intel. However, although most independent estimates found the bug to be of little importance and have negligible effect on most uses, it has caused a great public outcry. Companies like IBM (whose "586" microprocessor competed at that time with the Intel Pentium line) joined the condemnation. Finally, Intel was forced to offer to replace all flawed Pentium processors, at huge potential cost to the company (however, it turned out that only a small fraction of Pentium owners actually bothered to get their chips replaced).
References
- A page with precise information, also about the cause (http://www.cs.earlham.edu/~dusko/cs63/fdiv.html)
- A Tale of Two Numbers (http://www.mathworks.com/company/newsletters/news_notes/pdf/win95cleve.pdf), by Cleve Moler of The MathWorks
- ZIP-file containing more details (http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=1666&objectType=file) (See ZIP file format for details on the file)
- Intel's official site (http://support.intel.com/support/processors/pentium/fdiv/)