Skip to content

FPGA routing and switching using evolutionary algorithms and simple heuristics

Notifications You must be signed in to change notification settings

as51340/Bachelor-s-thesis

Repository files navigation

Bachelor-s-thesis

Rješavanje problema smještanja i povezivanja kod sklopa FPGA

Sažetak

Genetski su algoritmi jedna od metoda evolucijskog računarstva koji imaju široku primjenu. Najčešce se upotrebljavaju za rješavanje kombinatoričkih problema i optimizacijskih problema čija je domena realno područje, no mogu se izuzetno kvalitetno koristiti i za treniranje neuronskih mreža kao alternativa algoritmu propagacije unatrag. Ovaj se rad bavi rješavanjem optimizacijskog problema mapiranja logičkog u fizički FPGA svijet pomocu genetskog algoritma. Pripremljeni su testovi na kojima je izmjerena kvaliteta rada genetskog algoritma.

Solving placement and routing problems in FPGA

Abstract

Genetic algorithms are one of the methods used in evolutionary computing that has many applications. They are used for combinatoric and in optimization problems whose domain is continuum but can also give very satisfying results when used for training neural networks as an alternative to backpropagation algorithm. This paper is dealing with the problem of mapping logical model into physical FPGA model with the help of genetic algorithm. Different tests are prepared for measuring algortihm’s quality.

About

FPGA routing and switching using evolutionary algorithms and simple heuristics

Resources

Stars

Watchers

Forks

Packages

No packages published