This is the main idea about parallel computing. GPU computing is like having hundreds and thousands of skinny guys with old vans, which can be pretty effective.

Each of the men is much slower than the strong guy, but if they are able to transport your load concurrently they will have the time to complete the task, have a beer together and come back to their family before the strong guy has done half the work. If you need the work done quickly and have to move a lot of furniture, of course you will chose the second company. What would you prefer? Of course it depends of how much stuff you have to move and how much time you have to do it. The second one instead has 5 normal men available, each of them with a van of average speed. You have two choices: the first company can provide you with the strongest man alive driving the world's fastest van. You have to move to a new town and are looking to hire a furniture mover company. To better explain the key concept behind parallel computing, let me give you a more 'realistic' example. I will cover parallel computing in detail in later articles, but if I had to outline in a few words what parallel computing is I would say that it is the partition of repetitive (and therefore often time consuming) tasks into singular tasks, each of which is performed by a single core or processing unit. GPU computing is a new branch of computer science and, more specifically, of parallel computing. CUDA is a proprietary programming language developed by NVIDIA for GPU programming, and in the last few years it has become the standard for GPU computing. In this guide I will explain how to install CUDA 6.0 for Mac OS X. This is the first article in a series that I will write about on the topic of parallel programming and CUDA.

Nvidia cuda emulator for mac