Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Benchmarks that I see around don't help to choose the optimal processor for work in java/eclipse. The only java benchmark that I have found is SciMark 2.0, but it is old and compares PCs, not processors. How could I compare the javaic effectivity of processors? Maybe some usual benchmark could be used as such? Maybe somebody is being testing CPUs in some usual java software and publishing results?

Any way except buying and comparing CPUs, please.

share|improve this question
Are you sure that focusing on CPUs is a good idea? Tools like Eclipse, which consist of many small components, but even my Glassfish instances are usually constrained by the I/O speed of the system. Comparing PCs instead of CPUs seems perfectly reasonable for that task. If you ask on order to prepare a purchase, invest in an SSD first, then in RAM, and only if you find out that some arbitrary midrange CPU doesn't suffice for your workload, revisit the CPU benchmarking. – jstarek May 15 '12 at 8:26
Java is not much effective language. Its progs work slowly. For example, Android emulator, as every emulator, is limited by CPU speed only. (with reasonable memory size, of course). And it is very slow on "arbitrary midrange CPU". RAM speed is relevant only for graphics, if you use integrated graphics and you use AMD. – Gangnus May 15 '12 at 8:35
As for SSD, it IS useful while building, that could be conducted anytime in Java. But it is simply the other bottleneck. – Gangnus May 15 '12 at 8:37
In my experience in Java development (and C dev too) on recent hardware, the only bottleneck is really the I/O. The io wait percentage is what make the CPU computation slow. As for pure computation power, Java is not ineffective. I've seen a space algorithms library that has been heavily optimised in C++ ported to Java, the Java port was performing on par with C++ or slightly under performed but with still good enough performances. More than the choice of CPU, it was the choice of containers/generics that was impacting the performances of the algorithms. – Huygens May 15 '12 at 9:23
@Huygens Thank you. Of course, I do optimize my soft. But 90% of time I am calling functions written by others. BTW, often they ARE optimized already. Good style often costs much, too. Anyway, I am asking about something else. – Gangnus May 15 '12 at 9:28

You must log in to answer this question.

Browse other questions tagged .