| Date: | Mon, 24 Apr 2006 10:51:54 +0300 |
| From: | Diomidis Spinellis <dds@aueb.gr> |
| Organization: | Athens University of Economics and Business |
| User-Agent: | Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.0.2) Gecko/20060404 SeaMonkey/1.0.1 |
| MIME-Version: | 1.0 |
| Newsgroups: | comp.lang.java.programmer |
| Subject: | Re: strange performance behavior of a mathematical method: why? |
| References: | <1145823428.424885.264990@u72g2000cwu.googlegroups.com> <e2grdf$2bf$1@volcano1.grnet.gr> <1145828078.019089.116170@i39g2000cwa.googlegroups.com> |
| In-Reply-To: | <1145828078.019089.116170@i39g2000cwa.googlegroups.com> |
| Content-Type: | text/plain; charset=ISO-8859-1; format=flowed |
| Content-Transfer-Encoding: | 7bit |
Dimitri Ognibene wrote: > what do you tjink of this method implementation? > any tip? > must i pass to c++ & asm to get better performance? > note that the g method called inside the loop is final and the compiler > does inline it, or it looks to be so from the profiling data. From what I can see this particular code manipulates floating point numbers. This is one of the (not common) cases where a good Java compiler and JIT-based runtime system should be able to give you the same performance as the one you would achieve with (say) C. You are unlikely to gain anything from moving to assembly, unless you can utilize instructions that compilers don't typically support. For example, I see that your code calculates some type of dot product: if your data allows it, you could gain by using Intel's SSE SIMD extensions or AMD's 3DNow. Another possibility would be to move your code to be executed by a 3D graphic card's hardware. -- Diomidis Spinellis Code Quality: The Open Source Perspective (Addison-Wesley 2006) http://www.spinellis.gr/codequality?cljp
Unless otherwise expressly stated, all original material on this page created by Diomidis Spinellis is licensed under a Creative Commons Attribution-Share Alike 3.0 Greece License.