Hi Stuart,
On Tuesday, June 18, 2013 11:12:28 AM UTC-7, Stuart Sierra wrote: > > One has to be very careful with this kind of > micro-benchmarking on the JVM. Dead-code elimination can > easily make something seem fast simply because it's not > doing anything. For example, in Java: > https://gist.github.com/stuartsierra/5807356 > > Being careful not to have dead code, I get about the same > results: 9 microseconds to sum an array of 10000 doubles in > Java. > > Getting back to Clojure, I notice a large difference between > Clojure 1.2.1 and 1.5.1 to sum an array. But looking more > closely, that difference is entirely due to dead-code > elimination: > https://gist.github.com/stuartsierra/5807676 > > With a loop doing real work, my tests show Clojure's > performance on this task is virtually indistinguishable from > Java. The performance hasn't changed since 1.2, although the > syntax has. (I get the same result with `areduce`.) > We thought we were being very careful -- taking into account warmup, dead code elimination, and so on (our actual benchmark suite runs the code many times and reports the sum of the results). > Also beware of this "helpful" Leiningen 2.1+ feature: it > disables some JVM optimizations to get faster start-up time. > https://github.com/technomancy/leiningen/wiki/Faster#tiered-compilation > > To avoid this, you need to add the following to your > `project.clj` file: > > :jvm-opts ^:replace [] > > The `^:replace` is critical. > > Whenever I'm uncertain about a benchmark, I run it without > Leiningen to make sure I know what's going on. > This was our issue. We had no idea that Leiningen added extra stuff to the host jvm-opts, thanks so much for your help with this! -Jason > > -S > > > On Thursday, June 13, 2013 4:50:48 PM UTC-4, Jason Wolfe wrote: >> >> Taking a step back, the core problem we're trying to solve is just to sum >> an array's values as quickly as in Java. (We really want to write a fancier >> macro that allows arbitrary computations beyond summing that can't be >> achieved by just calling into Java, but this simpler task gets at the crux >> of our performance issues). >> >> This Java code: >> >> public static double asum_noop_indexed(double[] arr) { >> double s = 0; >> for (int i = 0; i < arr.length; i++) { >> s += arr[i]; >> } >> return s; >> } >> >> can run on an array with 10k elements in about 8 microseconds. In >> contrast, this Clojure code (which I believe used to be as fast as the Java >> in a previous Clojure version): >> >> (defn asum-identity [^doubles a] >> (let [len (long (alength a))] >> (loop [sum 0.0 >> idx 0] >> (if (< idx len) >> (let [ai (aget a idx)] >> (recur (+ sum ai) (unchecked-inc idx))) >> sum)))) >> >> executes on the same array in about 40 microseconds normally, or 14 >> microseconds with *unchecked-math* set to true. (We weren't using >> unchecked-math properly until today, which is why we were doing the hacky >> interface stuff above, please disregard that -- but I think the core point >> about an extra cast is still correct). >> >> For reference, (areduce a1 i r 0.0 (+ (aget a1 i) r)) takes about 23 ms >> to do the same computation (with unchecked-math true). >> >> Does anyone have ideas for how to achieve parity with Java on this task? >> They'd be much appreciated! >> > -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
