{"id":7834,"date":"2010-09-29T14:26:56","date_gmt":"2010-09-29T21:26:56","guid":{"rendered":"http:\/\/wapreview.com\/?p=7834"},"modified":"2020-10-06T16:40:53","modified_gmt":"2020-10-06T23:40:53","slug":"opera-mobile-10-1-symbian-beta-2-much-improved-javascript-performance","status":"publish","type":"post","link":"https:\/\/wapreview.com\/7834\/","title":{"rendered":"Opera Mobile 10.1 Symbian Beta 2 – Much Improved JavaScript Performance"},"content":{"rendered":"

\"Opera \"Opera<\/p>\n

Today Opera Software released<\/a> a new Beta of Opera Mobile 10.1 for Symbian^1, Symbian^3 and Symbian 3rd Edition phones. Download Beta 2 at m.opera.com\/next<\/a>.  The main feature of the new release is the addition of the Carakan JavaScript engine’s Just In Time (JIT) compiler. Opera claims the new compiler makes JavaScript execution up to four times faster<\/strong>.<\/p>\n

I tested that claim by running the popular Sunspider<\/a> JavaScript benchmark on a Nokia N95-3, first with Opera Mini 10.1 Beta 1 and then with the new Beta. I think Opera is being too modest. Opera 10.1 Beta 2 was over 7 times faster on Sunspider <\/strong>than Beta 1; 24,516.8<\/strong> ms vs 181,405.8<\/strong> ms. Here are links to detailed results on the Sunspider benchmark tests for Beta 1 <\/a>and Beta 2<\/a><\/p>\n

I also tried Google’s V8<\/a>, another JavaScript benchmark.  Both Beta 1 and Beta 2 ran out of memory running the last test of the V8 suite. Google says that test, called Splay<\/em>, “…exercises the automatic memory management subsystem”.  Higher memory usage is a known issue with both Betas so presumably that is something Opera is working on.  On the six V8 test components that the Opera Browsers did complete, Beta 2 again did much better scoring 289.5 vs. 33.8 or 8.5 times better!<\/strong><\/p>\n

The table below summarizes the two Betas performance and their memory footprint after each test.  Memory was measured with Phonetinfo<\/a> 1.2 and is the difference in free RAM before starting Opera and after each test.<\/p>\n\n\n\n\n\n
Test<\/td>\nBeta 1<\/td>\nBeta 2<\/td>\nBeta 1 Memory<\/td>\nBeta2 Memory<\/td>\n<\/tr>\n
Sunspider<\/td>\n181,405.8 ms<\/td>\n24,516.8 ms<\/td>\n18,168 KB<\/td>\n27,248 KB<\/td>\n<\/tr>\n
V8<\/td>\n33.8<\/td>\n289.5<\/td>\n0 (DNF)<\/td>\n0 (DNF)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

In real world tests visiting pages that use Javascript extensively, Beta 2’s performance improvement was noticeable However don’t expect to see a huge difference with most pages.<\/p>\n