caches

jeffanderson.us|Blog: "Among these is the availability of twice as many CPU registers as 32-bit hardware. This allows twice as much data to be processed without having to hit the caches or RAM."

BeginLinux: "Squid acts both as a proxy, working in behalf of a user, and as a cache. When squid works as a proxy and a user makes a request for a web site, squid retrieves the web page and then provides it to the user.

The user, in reality never reaches the Internet as the proxy server retrieves and caches all the sites the user makes requests for."

Derek recently blogged that “phar is cool!”  Cool is great … but it doesn’t answer important questions: how does loading your application from a .phar file affect overall performance and scalability?

 How well does it work with leading bytecode caches?

Where are the benchmarks for phar?

On the PHP 10.0 blog, Stas does a little benchmarking of a big part of each request to a Zend Framework application - the Loader.