{"id":3736,"date":"2011-07-27T22:02:43","date_gmt":"2011-07-27T21:02:43","guid":{"rendered":"http:\/\/www.walkingrandomly.com\/?p=3736"},"modified":"2011-08-16T15:12:54","modified_gmt":"2011-08-16T14:12:54","slug":"matlab-gpu-cuda-experiences-on-my-laptop-elementwise-operations-on-the-gpu-1","status":"publish","type":"post","link":"https:\/\/walkingrandomly.com\/?p=3736","title":{"rendered":"MATLAB GPU \/ CUDA experiences on my laptop – Elementwise operations on the GPU #1"},"content":{"rendered":"

This is part 1 of an ongoing series of articles about MATLAB programming for GPUs using the Parallel Computing Toolbox.\u00a0 The introduction and index to the series is at https:\/\/www.walkingrandomly.com\/?p=3730<\/a>.<\/p>\n

Have you ever needed to take the sine of 100 million random numbers?\u00a0 Me either, but such an operation gives us an excuse to look at the basic concepts of GPU computing with MATLAB and get an idea of the timings we can expect for simple elementwise calculations.<\/p>\n

Taking the sine of 100 million numbers on the CPU<\/strong><\/p>\n

Let’s forget about GPUs for a second and look at how this would be done on the CPU using MATLAB.\u00a0 First, I create 100 million random numbers over a range from 0 to 10*pi and store them in the variable cpu_x;<\/p>\n

cpu_x = rand(1,100000000)*10*pi;<\/pre>\n
Now I take the sine of all 100 million elements of cpu_x using a single command.<\/p>\n
cpu_y = sin(cpu_x)<\/pre>\n
I have to confess that I find the above command very cool.  Not only are we looping over a massive array using just a single line of code but MATLAB will also be performing the operation in parallel<\/strong>.  So, if you have a multicore machine (and pretty much everyone does these days) then the above command will make good use of many of those cores.  Furthermore, this kind of parallelisation is built into the core of MATLAB….no parallel computing toolbox necessary.  As an aside, if you’d like to see a list of functions that automatically run in parallel on the CPU then check out my blog post on the issue<\/a>.<\/p>\n
So, how quickly does my 4 core laptop get through this 100 million element array?\u00a0 We can find out using the MATLAB functions tic<\/strong> and toc<\/strong>.  I ran it three times on my laptop and got the following<\/p>\n
>> tic;cpu_y = sin(cpu_x);toc\r\nElapsed time is 0.833626 seconds.\r\n>> tic;cpu_y = sin(cpu_x);toc\r\nElapsed time is 0.899769 seconds.\r\n>> tic;cpu_y = sin(cpu_x);toc\r\nElapsed time is 0.916969 seconds.<\/pre>\n
So the first thing you’ll notice is that the timings vary and I’m not going to go into the reasons why here.  What I am going to say is that because of this variation it makes sense to time the calculation a number of times (20 say) and take an average.  Let’s do that<\/p>\n
sintimes=zeros(1,20);\r\nfor i=1:20;tic;cpu_y = sin(cpu_x);sintimes(i)=toc;end\r\naverage_time = sum(sintimes)\/20\r\n\r\naverage_time =\r\n    0.8011<\/pre>\n
So, on average, it takes my quad core laptop just over 0.8 seconds to take the sine of 100 million elements using the CPU.\u00a0 A couple of points:<\/p>\n