Explotación de la potencia de procesamiento mediante paralelismo: un recorrido histórico hasta la GPGPU

Due to the improvement of semiconductor manufacturing technologies, and higher integration scales in the last decades, the power of computing devices has experienced an impressive growth. However, some obstacles have been also found along the way. As a consequence, the battle for reaching higher clock frequencies almost ended a few years ago. Nowadays, the power of processors is not measured exclusively using GHz. Other factors, as the number of cores and their inner design, also have a large impact. This paper provides an historical review on how parallelism techniques have been adapted over time to overcome these changes aiming to better exploit the available hardware.