On Nov 23, 2013, at 3:00 AM, Andrew Haley wrote: >> We've been doing fairly extensive micro-benchmark testing of the >> effect of adjusting the various algorithm crossover thresholds in >> BigInteger (cf. https://bugs.openjdk.java.net/browse/JDK-8022181). > > Out of interest, is this x86, sparc, or both? It probably doesn't > matter, but it might.
Various flavors of x86 and AMD Opteron but no SPARC as yet. Brian