Hmmm...
Looking at the code I see
(defn sum-trees [iterations depth]
(let [sum #(+ (check-tree (make-tree % depth))
(check-tree (make-tree (- %) depth)))]
(reduce + (map sum (range 1 (inc iterations))))))
Shouldn't expressing the algorithm as a REDUCE and MAP instead of a LOOP
do the trick? I would expect that to compile into parallel code.
Otherwise, why go through all the pain of learning functional
programming (and convincing management)?
Randall R Schulz wrote:
> On Sunday 07 December 2008 07:11, Peter Wolf wrote:
>
>> I'm a n00b, but isn't the point of this language to be *faster* than
>> Java?... at least on a multiprocessor machine.
>>
>
> I don't think performance is a particular criterion for the design of
> this language. It's not unimportant, but the quality of the code it
> engenders, especially w.r.t. to concurrency and the difficulty of
> writing correct code using conventional thread-aware mechanisms,
> _is_ a key design goal.
>
>
>
>> Shouldn't the number of processors on the test machine make a big
>> difference to how fast it runs? Whereas, the Java version is only
>> dependent on the clock rate of the individual processors.
>>
>
> Only for algorithms that are both parallelizable and which have actually
> been written in parallel form. There is not yet (and may never be) any
> ability to automatically parallelize arbitrary algorithms or code.
>
>
>
>> What happens if we run this benchmark on a nice 4 core core machine?
>>
>
> And nothing else is running? One core will be used for the thread
> running this code. Another will do any I/O, though in this case there
> is virtually none, and another will do GC.
>
> This test is a sequential algorithm. I'm not familiar with the Alioth
> benchmarks / shootout, but maybe there are some parallel tests in
> there.
>
>
> Randall Schulz
>
> >
>
>
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Clojure" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
-~----------~----~----~----~------~----~------~--~---
