Path: csiph.com!xmission!news.snarked.org!border2.nntp.dca1.giganews.com!border1.nntp.dca1.giganews.com!nntp.giganews.com!news.iecc.com!.POSTED.news.iecc.com!nerds-end From: Christian Gollwitzer Newsgroups: comp.compilers Subject: Re: Parsing using a Graphics Processing Unit (GPU)? Date: Tue, 1 Sep 2020 09:22:11 +0200 Organization: A noiseless patient Spider Lines: 18 Sender: news@iecc.com Approved: comp.compilers@iecc.com Message-ID: <20-09-002@comp.compilers> References: <20-09-001@comp.compilers> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit Injection-Info: gal.iecc.com; posting-host="news.iecc.com:2001:470:1f07:1126:0:676f:7373:6970"; logging-data="82224"; mail-complaints-to="abuse@iecc.com" Keywords: parse, performance Posted-Date: 01 Sep 2020 12:03:24 EDT X-submission-address: compilers@iecc.com X-moderator-address: compilers-request@iecc.com X-FAQ-and-archives: http://compilers.iecc.com Xref: csiph.com comp.compilers:2573 Am 31.08.20 um 12:35 schrieb Roger L Costello: > Has the parsing community found a way to take advantage of GPUs? I don't think that you can speed up parsing a lot using GPUs. GPUs are generally good at parallel processing. A single thread is usually slower than a CPU thread, and there is overhead, because they are not self-contained - i.e. you can usually speed up only some part of a program, and it needs to be uploaded to the GPU and downloaded back. GPUs also have faster memory, *if* you access it either in big blocks or as a serial stream, in which case the compiler can transform it to block access. For random accesses, the memory is slower. I have done some basic GPU programming, and I think that parsing is not a parallel task in that sense. The parser reads the input as a stream of tokens; you can't split the C file at some arbitrary point in half and parse both parts independently. Christian