This thesis presents an investigation into the optimization of Graph Neural Network (GNN) architectures using the Graph Neural Architecture Search (GraphNAS) framework, an automated system that employs reinforcement learning ...

Asynchronous circuit design has been conventionally regarded as a valid alternative to synchronous logic due to its potential for low consumption of resources, power and delay. This includes areas such as the communication ...

Current quantum processors must be scaled to match the computational power needed for quantum algorithms that present quantum advantage. It is not feasible to keep integrating more qubits on a single core because qubit ...

Communications within a chip or even groups of chips are typically carried out using physical connections to transmit signals between them. Nowadays, the use of wireless communications using antennas inside chip packages ...

El campo de investigación de las GNN ha experimentado un importante aumento en los últimos años. Esto es consecuencia de su capacidad única para de gestionar grafos, una estructura de datos muy común que se aplica ampliamente ...

Els centres de dades són elements fonamentals en el dia a dia d'un gran nombre de sectors (tràfic web, IA, banca, comunicacions etc). Aquests necessiten tenir la capacitat de gestionar tot el tràfic que circula en el seu ...

Multicore architectures cover a wide technologic spectrum and it is on the vanguard for an unlimited number of applications. As technology keeps growing and number of cores per chip increases (reaching what nowadays is ...

En los últimos años, un nuevo enfoque de modelado y aprendizaje a partir de datos relacionales ha conquistado el mundo del Deep Learning: las Graph Neural Networks (GNN). Los algoritmos distintivos de las GNN se encuentran ...

Els avenços en nanotecnologia i materials avançats anuncien una nova era per als dispositius a nano escala en medicina de precisió. Aquests dispositius, equipats amb capacitats integrades de detecció, computació, comunicació, ...