Skip to content

Here we make an optimized code run even faster by harnessing computational power.

Notifications You must be signed in to change notification settings

prafful1/Parallel-Universe

Repository files navigation

Parallel-Universe

Here we make an optimized code run even faster by harnessing computational power.

What do we do here (For people who don't like to read) ?

alt text

(For people who like to read)?

We take problems that can be divided into subtasks which further can be approached individually. We solve these problems in parallel. Basically we solve problems with divide and conquer technique in parallel

Why should I use this approach?

Well why not. We use great algorithms to optimize our code. Why not harness our computing power to make our programs run even faster.

What tools do we use?

Java is one of the most popular languages used by developers. So for this repository all the programs will be written using Java 8. It provides a unique implementation of Fork/Join to obtain parallelism.

Why Java 8?

  • Fork/Join was introduced in Java 7 and has greatly improved in Java 8.
  • Thread specific task queues and work-stealing algorithm were the major enhancements introduced by Java 7.
  • This new implementation improves task processing time leading to faster execution. alt text

Other important links.

  • Sample addition program
  • web crawler

About

Here we make an optimized code run even faster by harnessing computational power.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages