On the symfony blog today, there's a new article looking to dispel some misunderstanding that's come up surrounding the speed of symfony applications and what effect it can have in a real-world situation.

We regularly see framework benchmarks pop up somewhere on the net, comparing symfony to other frameworks, either in PHP or in other languages. The conclusion of some of them is "symfony takes more time than other frameworks to display a 'hello, world', which makes it unsuitable for real-world web applications".

If some benchmarks show that symfony is slower, jumping to the conclusion that symfony is not optimized is a big mistake. [...] If symfony is well implemented (or so we think), then its speed is just the sum of the individual speeds of all the features it includes. Or, to put it differently, Symfony's speed is the consequence of a series of choices that we made to decide whether a feature should or should not make it to the core.

They suggest that the speed of symfony applications does not come as a result of the main portion of the framework, but rather as a choice of the development team as to what parts they wanted to include in the core and what not to. They also point out that, while speed does matter somewhat, especially in larger applications, it shouldn't be highest on the list.