Code optimisation is rarely worth it but if you can make an OS run faster then everyone will notice. With just over 200 lines of new code the improvements in Linux are described as dramatic!
Never believe a headline, the truth is always more complicated. However if you believe that operating systems and the people who create them are better than the rest of us then think again. How can it be that a simple change in a well developed operating system can create such an improvement in performance that it is regarded as remarkable? Surely the principle of operating system design are well known by now? It seems not and the operating systems are just as much "suck it and see" as apps.
The patch is to the scheduler and its effect is so remarkable that it can clearly be seen in the way the desktop behaves when under load. Even Linus Torvalds is impressed - he wrote:
It's an improvement for things like smooth scrolling around, but what I found more interesting was how it seems to really make web pages load a lot faster. Maybe it shouldn't have been surprising, but I always associated that with network performance. But there's clearly enough of a CPU load when loading a new web page that if you have a load average of 50+ at the same time, you _will_ be starved for CPU in the loading process, and probably won't get all the http requests out quickly enough.
So I think this is firmly one of those "real improvement" patches. Good job. Group scheduling goes from "useful for some specific server loads" to "that's a killer feature".
Other users have referred to the effect of the patch as amazing and dramatic. What is even more remarkable is that the kernel patch is just 233 lines of code - usually performance increases of this sort require major re-writes and deep architecture changes. For more technical details see patch.
The patch is currently not in any Linux distributions and to get it you will have to compile the kernel 2.6.38 or wait for the next release of Ubuntu (Natty Narwhal) whatever distro you prefer. Also given that the improvement is to the scheduler it will be effective on all architectures.