home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!iWarp.intel.com|ichips!ichips!glew
- From: glew@pdx007.intel.com (Andy Glew)
- Subject: Re: No Branch Delay Slot(s)...
- In-Reply-To: phil@news.ccutah.edu's message of Wed, 9 Sep 92 04:42:31 GMT
- Message-ID: <GLEW.92Sep9201208@pdx007.intel.com>
- Sender: news@ichips.intel.com (News Account)
- Organization: Intel Corp., Hillsboro, Oregon
- References: <1992Sep9.044231.12217@fcom.cc.utah.edu>
- Date: Thu, 10 Sep 1992 04:12:08 GMT
- Lines: 40
-
- [phil@news.ccutah.edu (Phillip Neiswanger)]
-
- In the august issue of Byte there is a rather lightweight article on DEC's
- Alpha architecture and its current incarnate the 21064. In the article it
- states that DEC deemed the branch delay slot to be of little value in
- future multiple(read >2) instruction issue implementations of the Alpha
- architecture. If I remember correctly, the article states that the use of
- delayed branch slots could introduce incompatibilities from implementation
- to implementation. This does not seem very intuitive to me. Would anybody
- care to discuss how branch delay slots are going to affect future generation
- of RISC cpus as they enter the era of multiple(read >2) instruction issue
- implementations.
-
- Branch delay slots are an artifact of a "first implementation".
-
- E.g. the MIPS R[23]000 had one branch delay slot, so that was made
- part of the architecture.
-
- The MIPS R4000 really has 2 (maybe 3, I'm sure I'll be corrected)
- branch delay slots, but there is still only one made architecturally
- visible.
-
-
- I only know of one mechanism to expose multiple branch delay slots in
- a manner that varies cleanly between implementations, in that the same
- binary will run nor matter how many branch delay slots are implemented
- - Hwu and Chang's "Forward Semantic". But even this assumes an
- in-order machine implicitly.
-
-
- --
-
- Andy Glew, glew@ichips.intel.com
- Intel Corp., M/S JF1-19, 5200 NE Elam Young Pkwy,
- Hillsboro, Oregon 97124-6497
-
- This is a private posting; it does not indicate opinions or positions
- of Intel Corp.
-
- Intel Inside (tm)
-