Lecture date: 13 December 2018
Graphics processing units (GPUs) have recently evolved from mere video cards to fully-fledged computing devices, no longer connected to visual peripherals. Leveraging the technological progress initiated by ever faster and richer visuals rendering, general-purpose GPUs contain hundreds of processors capable of crunching numbers concurrently. The GPU architectures offers recognized opportunities for scientific computing, but also pose constraints on how parallelism can be implemented efficiently. In other words, it is important to be aware of a continuum of situations linking hardware features, software coding requirements, and achievable improvements in computing performance. This non-technical talk will offer a little orienting knowledge on where the potency of GPUs lies, and present some notions useful to find your way in a rapidly evolving technological niche. It will hopefully help develop a sound intuition of what the so-called ´hardware acceleration' with GPUs is and invite reasoned interest in it.