home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.amiga.programmer:17402 comp.sys.amiga.hardware:21588
- Newsgroups: comp.sys.amiga.programmer,comp.sys.amiga.hardware
- Path: sparky!uunet!usc!cs.utexas.edu!torn!utgpu!engb
- From: engb@gpu.utcs.utoronto.ca (Ben Eng)
- Subject: Re: CISC and RISC
- Message-ID: <BzBtx4.DLE@gpu.utcs.utoronto.ca>
- Organization: Jet Penguin Lavatories
- References: <70436@cup.portal.com> <amipb.04wr@amipb.gna.org> <37844@cbmvax.commodore.com> <Bz8FD1.Dxt@ns1.nodak.edu> <1992Dec14.155039.7747@ugle.unit.no> <BzAxFw.Is6@dcs.ed.ac.uk>
- Date: Wed, 16 Dec 1992 00:41:27 GMT
- Lines: 61
-
- In <BzAxFw.Is6@dcs.ed.ac.uk> jxp@dcs.ed.ac.uk (Joe Potter) writes:
-
- >In article <1992Dec14.155039.7747@ugle.unit.no> skogaas@solan.unit.no (John Olav Skog}s) writes:
-
- >>Optimizing compilers (it is easy to generate good code for the RISC
- >>architecture)
-
- > Actually, I'm of the understanding that current compiler
- >technology is rather left behind by the sheer complexity of RISC coding.
- >It's easy to generate code for RISC machines, but getting turbopowered
- >optimisations is hard. CISC is easier to optimise.
-
- I would like to see that statement qualified.
-
- >>Hard-wired control unit
- >>Optimized instruction pipeline
- >>LOAD/STORE - architecture
-
- > What's actually so good about this, by itself? Given all the
- >other parts of RISC philosophy, great, but as a point by itself?
-
- The features are good because they allow for a "better", faster
- microprocessor. The criteria for better are probably different for
- different people, but clock speed, pipelining, scalability,
- cost, the ability to be able to use new fab processes quickly,
- and such are considerations.
-
- >>One clock-cycle pr. instruction
-
- > Ideally. More usually four cycles, but with a four stage pipeline.
-
- Nominal throughput of 1 inst/clock, with a latency of 4 clock cycles.
-
- > It may be easier to make op.sys.-software on a CISC processor.
-
- I consider that to be an incorrect statement. The operating system
- is almost completely written in a language such as C or C++, just like
- most other applications. If the compiler handles the translation
- well, where is the difficulty in generating code for RISC processors?
- I don't remember seeing volumes of articles, or hearing many complaints
- from RISC OS implementors about how difficult it is to generate code
- for RISC as opposed to CISC.
-
-
- It seems to me like your arguments against RISC are that it is more
- difficult to program from a software point of view. Well the benefits
- you get are mostly embodied in the hardware improvements (speed,
- design, fab processes, scalability, and whatever else RISC chip
- companies deem important). RISC vendors work closely with compiler
- implementors when designing the instruction sets. Essentially, the
- machine interface for most software people becomes the compiled
- language. I mean, as an asm programmer, would you really want to
- keep track of what values you have in 64 or more registers? :-)
- So get rid of the asm programmers (or lock them up in a compiler
- vendor's lab) and everyone will be pretty happy.
-
- Ben
- --
- e-mail: engb@gpu.utcs.utoronto.ca or ben@jetpen.gts.org (Ben Eng)
- UofT EngSci 9T2 ``We are all masochists here.''
- Home: (416)-979-8761, (416)-979-7885
-