home *** CD-ROM | disk | FTP | other *** search
- Path: hell.team17.com!news
- From: boberg@team17.com (Stefan Boberg)
- Newsgroups: comp.sys.amiga.programmer,comp.sys.amiga.games,alt.sys.amiga.demos,comp.sys.amiga.misc
- Subject: Re: AB3D II beats Quake....
- Date: Sat, 30 Mar 1996 00:35:30 GMT
- Organization: Team 17 Software Ltd.
- Message-ID: <4jhvll$7ih@hell.team17.com>
- References: <74000105753944194756@BIRDLAND> <10017.6659T1424T209@mbox.vol.it> <4jbcno$7m9@soleil.uvsq.fr>
- NNTP-Posting-Host: zonk.team17.com
- X-Newsreader: Forte Free Agent 1.0.82
-
- Nicolas Pomarede <pomarede@isty-info.uvsq.fr> wrote:
-
- >The problem is not the size of the opcode, it's the time needed to read it.
- >All modern CPU read opcode in one cycle, and since buses are 32 or 64 bits,
- >reading 8 bit is usually a waste of power.
-
- Yes. But modern CPU's don't read instructions one by one. The
- Pentium for example has 128-byte prefetch queues, filled by reading
- 32-byte (I think) cache lines using 64-bit wide reads.
-
- >One of the big problem of the Pentium and 680x0 is that the opcode are not
- >of the same size. 680x0 can have opcode of 2,4,8,10 bytes and x86 can even
- >have opcode of one byte (those compatible with 286).
-
- Actually, the 80x86 can have VERY long instructions. Somewhere
- around 14 bytes I think. And BTW: the one-byteness of certain
- instructions has nothing to do with 286 or it's predecessors - even
- the 80x86 had multi-byte instructions.
-
- >RISC CPU can predecode up to 8 instructions in parallel,
- >because they know that every 32 bits there's a new intruction.
-
- I seem to recall that the P6 handles this by adding a classification
- stage in the pipeline, marking instruction boundaries. I think this is
- cached as well.
-
- >Also, you said PPC was the worst RISC, but on the other hand you would like
- >to see the 3DO M2 in an Amiga : do you know that the M2 is based on PPC603
- >and has its speed from it ?
-
- Well, the M2 is really fast only because of the custom graphics
- hardware. You could potentially use this in a computer, although I
- don't know what Matsushita would make of that. The CPU is of course a
- welcome improvement over the PlayStation's snail-like R3000A, though
- :-)
-
- >Nicolas Pomarede
- >e-mail: pomarede@isty-info.uvsq.fr
-
-
-
- ===============================================================
- Stefan Boberg boberg@team17.com
- "This, like, sucks in ways that we've never seen stuff suck
- before, so it's kinda cool!" - B&B
-
-