Assessing Defragmentation Strategies for FPGAs

• Autor(a) principal: Manuel G. Gericota
• Data de Publicação: 2006
• Outros Autores: Gustavo R. Alves, Luís F. Lemos, José M. Ferreira
• Tipo de documento: Livro
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://repositorio-aberto.up.pt/handle/10216/84666
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

Fragmentation on dynamically reconfigurable FPGAs is currently a major obstacle to the efficient management of its logic space. When resource allocation decisions have to be made at run-time a relocation of currently running functions may be necessary to release enough contiguous resources to implement incoming functions.Relocation should take into account any specifics of functions functionality and also those of the FPGAs architecture as to not affect systems performance. A simple and fast method to assess performance degradation of a function during relocation is proposed in this paper. This method is based on previous function labelling and on the concept of proximity vectors.