Xref: utzoo comp.arch:18659 comp.compilers:1403 Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!world!esegue!compilers-sender From: larus@primost.cs.wisc.edu (James Larus) Newsgroups: comp.arch,comp.compilers Subject: Re: Compilers taking advantage of architectural enhancements Keywords: design, optimize Message-ID: <11513@spool.cs.wisc.edu> Date: 17 Oct 90 13:26:31 GMT References: <1990Oct9> <3300194@m.cs.uiuc.edu> Sender: compilers-sender@esegue.segue.boston.ma.us Reply-To: larus@primost.cs.wisc.edu (James Larus) Followup-To: comp.arch Organization: University of Wisconsin--Madison Lines: 13 Approved: compilers@esegue.segue.boston.ma.us Having implemented a few `state of the art' compiler algorithms (and a few beyond the state of the art), I can say that most competent graduate students could code them, but most companies wouldn't implement them. The key problem is that most papers present enough details, but don't present enough evidence that a technique is effective. A rational project manager would not commit the resources to try out these algorithms. There are exceptions, of course, but compiler papers frequently use towers of hanoi or puzzle-sized benchmarks. /Jim -- Send compilers articles to compilers@esegue.segue.boston.ma.us {ima | spdcc | world}!esegue. Meta-mail to compilers-request@esegue.