Joe Darcy <joe.darcy@...> writes: > One of my long-term goals is to finish the > small matter of programming [2] to port FDLIBM from C to Java, which > would immunize the JDK from this class of problem.
In case people are not aware, FastMath from Apache commons-math is a pure Java implementation: http://svn.apache.org/viewvc/commons/proper/math/trunk/src/main/java/org/apache/c ommons/math/util/FastMath.java?view=markup I ran the performance test in their test suite a few months ago and published the results (note that I didn't write the performance test, but I fixed some obvious micro-benchmark issues after they were mentioned in the comments): http://blog.juma.me.uk/2011/02/23/performance-of-fastmath-from-commons-math/ Best, Ismael
