You are comparing a dev version of safari to the latest stable of chrome, you might want to compare the dreadful current iteration of safari? And where's the memory usage, as well as the memory leaks in say 1,2, or 3 hours? Plus the sample of webpages is way to small AND chrome has java enabled. The number of extensions on each one is also not factored in. Is there any way you can make this comparison less valid?