4 ms·

They used to have a website talking about the FLEET architecture, but it seems to have been taken down. Here it is on archive.org: https://web.archive.org/web/

They used to have a website talking about the FLEET architecture, but it seems to have been taken down. Here it is on archive.org: https://web.archive.org/web/20120227072220/http://fleet.cs.berkeley.edu/ https://web.archive.org/web/20120227072220/http://fleet.cs.b... Edit: the page cited above has these links, but I should explicitly call the slides they call the best introduction to Fleet [1], and a page full of memos [2] [1] https://web.archive.org/web/20120227072220/http://fleet.cs.berkeley.edu/docs/slides.pdf https://web.archive.org/web/20120227072220/http://fleet.cs.b... [2] https://web.archive.org/web/20120227072220/http://fleet.cs.berkeley.edu/docs https://web.archive.org/web/20120227072220/http://fleet.cs.b...

ChuckMcM 11y ago
FWIW, as I recall this was FLEET's fatal flaw (part of the communication discussion); * This can cause deadlock * Programmer must keep input dock fifos from overflowing Sun did a lot of work with async logic in the SPARC 10, it was written up in IEEE Spectrum I believe, and one of the things that always is a problem are that fabrics without flow control (back pressure or emission control) are subject to failure at the worst possible time.