Skip to main content

Hi there!

The magical world of parallel computing awaits you!


We all know Serial computing using our favourite mainstream programming language( C, C++, Java, Python ) on a single faithful CPU with a module of RAM of a standard PC, one machine instruction after another. (Don't fret even if you don't! We'll discuss concepts gradually transitioning from serial processing to concurrency to parallelism and beyond!)

Now, suppose we have a 2.0 GHz single core CPU, so it is capable of about 2 billion instructions per second.
Compare that to Human calculation time, about 2 seconds per instruction!!! (assume simple addition or relocation instruction)

What if we want more than that? Just visit the closest computer hardware store and buy another PC with:

  • n processors with m cores, x.yz GHz
  • Superscalar architecture, Hyper threading, etc
  • r GB GDDR5 RAM and s GB SSD 
  • Integrated Graphics and Discrete Graphics card with abcd GPGPU cores

(Wauw, that is possibly going to hit your bank account...)

Now we're ready to go! We've got all necessary hardware for improving performance of our code! Right?

 NOPE.

Parallelization being the next step of serial computing is a misconception. There are many optimizations and critical algorithm changes which can lead to amazing performance gains without even entering parallelism. So, remember that parallelization is not the only solution for multitasking.

Once the program is most logical and optimal, we can dive into parallelizing/multitasking. I'll describe many common techniques in future posts, so let's save that for later.

In this blog, I'm going to help you understand the principles behind parallelization and improving computation performance.

There is a lot to convey in this vast subject, so there's something new to learn everyday, so I'll keep the blog updated with latest bits of general advice and code snippets! (in mainstream programming languages)

Here's my (rather silly) animation of serial and parallel computing!




Static Image Source: Lemon shooter by Balderek
Labels:

Comments

Post a Comment

Popular posts from this blog

Parallel computing jargon

Parallel is always better than Serial... right? NO Well, in the most general case, parallel computing is better than serial computing in terms of speed and throughput. Sometimes, we have to make other considerations too. As a comparison , consider computer networking where serial transmissions are straight-forward and faster than their parallel SCSI  counterparts! Some processes are inherently not parallelizable, due to presence of data dependency. (Two bank account withdrawals from different locations which may lead to negative account balance if done simultaneously! Anyway, such a pair of withdrawals with critical section management using semaphores/mutexes conceptually and momentarily reduces to serial execution...) On a lighter note, the process of Sneezing and keeping your eyes open is not parallelizable for example! Before jumping into the concepts and principles of parallelizing a given task, let us go through some interesting set of ( controversial ) ke
Post a Comment
Read more

EnchantingProgram: Spoiler alert

This is part-2 of the "EnchantingProgram" post, read this post first: http://magical-parallel-computing.blogspot.in/2017/04/a-simple-python-program-using.html So, let's see the actual reason of the speedup in the C and C++ programs. Lo and Behold, it is the effect of Branch Prediction ! Surprised? Well, at least my comments in the programs should have given you some direction!!! The if condition leads to a branch in the control flow. We know that branch predictions lead to pipeline flushes and create a delay in the piped execution scheme. Modern microprocessors utilize complex run-time systems for throughput, execution speed and memory efficiency. One example of such a technique is dynamic branch prediction. A long time ago, microprocessors used only a basic technique called static branch prediction, with two general rules: A forward branch is presumed to be not taken A backward branch is presumed to be taken Now, static branch p
Post a Comment
Read more

A practical comparision of multitasking libraries

The following code-base attempts to compare multiple libraries on a simple experiment of Algebraic operations. The results are in favour of certain libraries because they are more natural in he application of Compute bound tasks, whereas the others are better at I/O bound tasks. Threading in Python is considered broken... I've also refrained from using the MPI library of C. I've made a driver.py program which create 5 sets of test-cases of increasing size to test the excution times of each version of each program. The comparison is across different libraries and languages, with the following programs: C : optimal serial code openMP directive based parallelism pthread library Python : optimal serial pyMP Python multiprocessing module Note that the gcc compiler automatically vectorizes the addition of arrays x and y using SIMD compliant hardware and appropriate data-types. I'll be adding the GPU comparison as soon as I can! Currently I don't h
Post a Comment
Read more