home *** CD-ROM | disk | FTP | other *** search
- Path: grafix.xs4all.nl!john.hendrikx
- Date: Sat, 24 Feb 96 20:36:33 GMT+1
- Newsgroups: comp.sys.amiga.misc
- Distribution: world
- Subject: Re: Speed: 68040 vs. 68060
- MIME-Version: 1.0
- Content-Type: text/plain; charset=iso-8859-1
- Content-Transfer-Encoding: 8bit
- From: john.hendrikx@grafix.xs4all.nl (John Hendrikx)
- Message-ID: <john.hendrikx.4h61@grafix.xs4all.nl>
- Organization: Private
-
- In a message of 22 Feb 96 Lee Huggett wrote to All:
-
- >> SysInfo is crap, that 40MIPS is really wrong. My 040/40 does that.
-
- LH> I was mearly stating a fact Sysinfo does give 40 MIPS for an '060
-
- That's because SysInfo's MIPS test isn't real world code, it doesn't make use
- of the 060's ability to execute more than 1 instruction at the time. Normal
- programs however (the ones not especially optimized for the 060) still have a
- good chance to execute more than 1 instruction at the time (I think that
- 'normal' programs would do about 1.5 instruction/cycle on average on the 060),
- so SysInfo is quite unrealistic.
-
- LH> I know the '060 is faster the 040/40 (I would have bought one of those
- LH> otherwise) but I think it is unlikely that you would get 81 MIPS out of
- LH> it at a sustained rate
-
- You can get even more out of it than that at a sustained rate (the maximum
- possible is 100 MIPS at 50 MHz), it just depends on what kind of code your
- running on it. To get optimal performance on 060 however you'll need programs
- to be recompiled or optimized for it.
-
- >> Most of the SAS/C compiled code does not have be specially 68060
- >> optimized, in a real world situation the 060 is usually able to run 2
- >> instructions at the same time.
-
- LH> If code is optimised for an '060 is has more chance of running 2 or
- LH> more instructions at once then unoptimised code.
-
- The 060 can't run *more* than 2 instructions at once.
-
- Grtz John
- -- Via Xenolink 1.985B5, XenolinkUUCP 1.1
-