Finding max value in CUDA

parallel-processing cuda reduction
kar picture kar · Mar 10, 2011 · Viewed 16.4k times · Source

I am trying to write a code in CUDA for finding the max value for the given set of numbers.

Assume you have 20 numbers, and the kernel is running on 2 blocks of 5 threads. Now assume the 10 threads compare the first 10 values at the same time, and thread 2 finds a max value, so thread 2 is updating the max value variable in global memory. While thread 2 is updating, what will happen to the remaining threads (1,3-10) that will be comparing using the old value?

If I lock the global variable using atomicCAS(), will the threads (1,3-10) compare using the old max value? How can I overcome this problem?

Answer

jwdmsd picture jwdmsd · Mar 10, 2011

This is a purely a reduction problem. Here's a good presentation by NVIDIA for optimizing reduction on GPUs. You can use the same technique to either find the minimum, maximum or sum of all elements.

Related questions

Using CUDA with Visual Studio 2017

I'm trying to install CUDA, but I get a message saying "No supported version of visual studio was found". I think that this is because I am using Visual Studio 2017 (Community), and CUDA currently only supports up to Visual Studio 2015. …

Read More
Can/Should I run this code on a GPU?

I'm working on a statistical application containing approximately 10 - 30 million floating point values in an array. Several methods performing different, but independent, calculations on the array in nested loops, for example: Dictionary<float, int> noOfNumbers = new Dictionary<…

Read More
Fast algorithm to calculate Pi in parallel

I am starting to learn CUDA and I think calculating long digits of pi would be a nice, introductory project. I have already implemented the simple Monte Carlo method which is easily parallelize-able. I simply have each thread randomly generate …

Read More