4 ms·
All he did was compare a specific instance where Java's type erasure lead to bad performance. As one commenter has said, using a hashtable specialised for primi
by verroq 15y ago
All he did was compare a specific instance where Java's type erasure lead to bad performance. As one commenter has said, using a hashtable specialised for primitives would yield much better results.
Then he cherry picks some examples from the language speed shoot-out, in which Java beats F# in speed in 6/8 of the challenges.
- nickik 15y agoThe problem with the shoot-out it that it tests mono, right? I have not jet seen good CLR vs JVM benchmarks of the same code and algorithm (C#/Java would work well). Does anybody has any pointer. It would be good marketing of Microsoft to show that the CLR is faster then the JVM for some typical benchmarks.