If you’re a PC or laptop gamer, you’ll probably already understand the function and importance of the graphics processing unit (GPU). This is the specialised electronic circuit, in the form of a powerful graphics card, which rapidly manipulates and changes memory in order to accelerate the creation of images in a frame buffer. In other words, it’s what allows fancy modern gaming graphics to work the way they do, creating stunningly realistic visuals and/or ultrafast lag-free gameplay in the process.
GPUs are not only found in gaming computers but are also employed in smartphones and consoles; however, users of these types of devices rarely take them apart to upgrade constituent parts or in order to meet the system requirements of the next cutting-edge game.
Now, a PhD student at the University of Manchester, Alex Chow, has used GPU technology to create highly complex simulations of “violent fluid flows” that can be used in scientific research.
GPUs are a far cheaper alternative to supercomputers
These kind of complex simulations tend to have millions of data points and can require billions of calculations to achieve any sort of useful accuracy. This requires a lot of processing power, and so these sims are usually run on “supercomputers.” Despite the name, these are actually comprised of numerous linked central processing units (CPUs) and thousands of computer cores rather than being a single super-powered computer.
This can provide a huge amount of power that can be scaled up according to the demand of the simulation or project. They also tend to be prohibitively expensive and use vast amounts of energy, meaning that they tend to be limited to only a relatively small group of facilities, teams and individual researchers.
Chow has developed software from an open-source code that uses a GPU to simulate complex, violent hydrodynamic (water) flows, such as modelling the impact of powerful oceanic waves crashing against offshore turbines. The software, which could be adapted to other engineering simulations, can compute millions of data points at a fraction of the cost of a supercomputer.
Chow said: "Using this kind of technology reduces the costs of complex scientific simulations from hundreds of thousands of pounds to just a couple of thousand. An advantage is that most researchers and small engineering companies are able to afford a relatively powerful laptop/computer with a quality GPU, so it makes this kind of simulation and research even more accessible."
Simulations can be more practical than physical experiments
Talking specifically about his project and the value of using computer modelling, Chow added: "The amount of energy produced from offshore environments is increasing as the world tries to meet the world’s energy targets, but the ocean environment can be very violent and harsh, so efficiently designing structures for these environments is a difficult task.
“Using physical experiments can be extremely impractical and not representative of the problem. These simulations allow engineers and researchers to make important decisions about the design of a structure without having to invest in site visits and costly experiments."
