How to optimize C++ code?

Question

0.00/5 (No votes)

See more:

For illustration purposes, I have a data structure as shown below. What is the fastest Code 1 or Code 2?

C++

struct Array{

// length of array
public: int length;

// pointer to int array
public: int *pData;

// allocate array
public: bool allocate(int size);

// constructor
public: Array(int size){
allocate(size);
}

// default constructor
public: Array(){
length = 0;
pData = 0;
}

public: ~Array(){
if(pData) delete []pData;
}

};

Code 1:

C++

Array array(10000);

for(int i = 0;i < array.length;i++) array.pData[i] = i*2;

Code 2:

C++

Array array(10000);

int l = array.length;

int *pData = array.pData;

for(int i = 0;i < l;i++) pData[i] = i*2;

Posted 14-Jun-14 2:24am

BupeChombaDerrick

Updated 14-Jun-14 2:37am

v2

Add a Solution

Comments

[no name] 14-Jun-14 8:31am

Benchmark your code and find out for yourself.

BupeChombaDerrick 14-Jun-14 8:42am

Does accessing the member variable within a loop affect the code performance? I'am writing computer vision code and I want to fully optimize it.

Philippe Mori 14-Jun-14 9:05am

Such optimizations almost never matters and when it matters, you should probably use parallel loops anyway so that all cores would works.

Usually, you select appropriate algorithms (for ex. O(n log n) instead of O(n²)) and then you use profiling tools once you have working code to tune code where it does matters.

Usually it not low-lvel optimizations that matters much but high level optimizations.

BupeChombaDerrick 14-Jun-14 14:52pm

I have already created my own computer vision algorithms for mobile and I have this app using my vision algorithms https://play.google.com/store/apps/details?id=com.facebook.rethinkvision.Bimostitch, I want to optimize the code on low and high level. You are right I guess I need to do parallel processing, I'am currently doing some research on that topic, any links to good literature on parallel processing can be helpful.

Jochen Arndt 14-Jun-14 9:13am

There should be no need to optimize the code because this will be done by the compiler. You may let the compiler generate assembly output for both versions and compare them to see which might be faster.

BupeChombaDerrick 14-Jun-14 14:56pm

Just thought I had to tune things up myself, maybe some compilers ain't that intelligent :)

Joren Heit 16-Jun-14 8:45am

Completely off-topic, but still useful for you I hope. In C++, you don't need to declare access rights on every variable/function (unlike C#/Java). A class can have multiple private/public/protected regions, delimited by e.g. public:. Everything following this line will be publicly accessible. Furthermore, the members of a class are public by default, whereas the members of a struct are private by default. Therefore, if all your members are public anyway, you might as well put them in a struct (good habit arguments put aside).

BupeChombaDerrick 16-Jun-14 9:07am

Well the question is not about the way variables are declared, but how the processing is done, Code 1 is accessing a pointer variable every time in the loop, while Code 2 is only doing it once. It turns out that both codes will run equally fast when compiled in release mode since the compiler will optimize it for me, thus to further speed things up, parallel processing is the only answer :)

Joren Heit 17-Jun-14 9:24am

I thought it was clear that I'm aware of my comment being off-topic (that's why it's a comment). Just a tip to help you improve on your C++ vocabulary, take it or leave it! ;-)

3 solutions

Solution 2

You could try writing C++ and use std::vector and std::generate(). Only if that's too slow consider managing memory yourself. Even if you need to manage memory yourself then use std::unique_ptr to automagically sort out releasing the memory.

Posted 14-Jun-14 21:41pm

Aescleal

Updated 14-Jun-14 21:44pm

v3

Solution 3

in a release build both solution should have the same speed, because optimization.

Code 3:

C#

Array array(10000);

int l = array.length;

int *pData = array.pData;

for(int i = 0;i < l;i++) pData++ = i*2;

with offsetting the accessed array

pData++

:-)

Posted 15-Jun-14 23:57pm

KarstenK

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

CPallini · Accepted Answer · 2014-06-14T04:39:00

Solution 1

As already suggested, if in doubt, make the compiler generate the assembly codes for you and then compare them. With a decent compiler there should be no difference.

Posted 14-Jun-14 4:39am

CPallini

Comments

BupeChombaDerrick 14-Jun-14 14:58pm

Thank you. I guess I have to focus more on hardware acceleration.