ed@plx.UUCP (Ed Chaban) (11/18/86)
There has been much screaming about which chip is faster 80386 or 68020. Recently, I received benchmark data from Neal Nelson & Associates. This benchmark showed results from a Compaq 386. I compared this data to the results from a Tower32 The 386 was SIGNIFICANTLY *SLOWER* than the '020 in LONG INTEGER MATH while it was virtually *EQUAL* to the '020 in SHORT INTEGER MATH. Don't say it was the memory because the "String Copy and Compare" test was identical too. So, What's the story intel? What do I have to do to make the 386 beat the '020 Hmmmmm??? Does anyone out there have some 386 data they care to share? Whetstones/Dhrystones perhaps? -ed- sun!plx!ed
mark@cogent.UUCP (Mark Steven Jeghers) (11/18/86)
In article <324@plx.UUCP> ed@plx.UUCP (Ed Chaban) writes: >There has been much screaming about which chip is faster 80386 or >68020. Recently, I received benchmark data from Neal Nelson & Associates. >This benchmark showed results from a Compaq 386. I compared this data to >the results from a Tower32 > >The 386 was SIGNIFICANTLY *SLOWER* than the '020 in LONG INTEGER MATH >while it was virtually *EQUAL* to the '020 in SHORT INTEGER MATH. >Don't say it was the memory because the "String Copy and Compare" >test was identical too. > >So, What's the story intel? What do I have to do to make the >386 beat the '020 Hmmmmm??? Does anyone out there have some 386 data >they care to share? Whetstones/Dhrystones perhaps? Is disinformation a possibility here? You know, get the rumor out before anyone knows for sure which is faster, thus grabbing a little head start in the race? Yes? No? -- +----------------------------------------------------------------------------+ | Mark Steven Jeghers - the living incarnation of "Deep-Thought" | | ("You won't like the answer ... you didn't ask it very well.") | | | | {ihnp4,cbosgd,lll-lcc,lll-crg}|{dual,ptsfa}!cogent!mark | | ^^^^^^-------recommended------^^^^^ | | | | Cogent Software Solutions can not be held responsible for anything said | | by the above person since they have no control over him in the first place | +----------------------------------------------------------------------------+
tomk@intsc.UUCP (Tom Kohrs) (11/18/86)
> Recently, I received benchmark data from Neal Nelson & Associates. > This benchmark showed results from a Compaq 386. I compared this data to > the results from a Tower32 > > The 386 was SIGNIFICANTLY *SLOWER* than the '020 in LONG INTEGER MATH > while it was virtually *EQUAL* to the '020 in SHORT INTEGER MATH. > Don't say it was the memory because the "String Copy and Compare" > test was identical too. What was the sofware environment? As far as I know at this point Compaq is shipping the Compaq 386 only with MS-DOS. This uses 16 bit compilers and 286 programming models. In other words it treats the 386 as a fast 8088 with all arithmetic being done 16 bits at a time. If the OS environment was UNIX V.3 for the 386 then I would want to know more about what the benchmarks are. If you have the environment information I would appreciate seeing it. -- ------ "Ever notice how your mental image of someone you've known only by phone turns out to be wrong? And on a computer net you don't even have a voice..." tomk@intsc.UUCP Tom Kohrs Regional Architecture Specialist Intel - Santa Clara
ed@plx.UUCP (Ed Chaban) (11/20/86)
In article <64@cogent.UUCP> Mark Steven Jeghers writes: > In article <324@plx.UUCP> ed@plx.UUCP (Ed Chaban) writes: > >There has been much screaming about which chip is faster 80386 or > >68020. Recently, I received benchmark data from Neal Nelson & Associates. > >This benchmark showed results from a Compaq 386. I compared this data to > >the results from a Tower32 > > > >The 386 was SIGNIFICANTLY *SLOWER* than the '020 in LONG INTEGER MATH > >while it was virtually *EQUAL* to the '020 in SHORT INTEGER MATH. > >Don't say it was the memory because the "String Copy and Compare" > >test was identical too. > > > > Is disinformation a possibility here? You know, get the rumor out > before anyone knows for sure which is faster, thus grabbing a little > head start in the race? Yes? No? Only if you think Neal Nelson has an axe to grind. So far as I can tell, Nelson is very objective in his benchmarking. If you are accusing *ME* Mark, I'd have to be INSANE to misquote published data. As for "Head Start" it is obvious that Motorola *ALWAYS* had it. I personally don't give a sh*t if Intel's chip is faster or slower, all I wanna know is what is the reason for the long/short integer differances. -ed- Ed Chaban Plexus Computers Inc. Phone: (408) 943-2226 Net: sun!plx!ed
jsgray@watmath.UUCP (Jan Gray) (11/20/86)
In article <404@intsc.UUCP> tomk@intsc.UUCP (Tom Kohrs) writes: >> The 386 was SIGNIFICANTLY *SLOWER* than the '020 in LONG INTEGER MATH >> while it was virtually *EQUAL* to the '020 in SHORT INTEGER MATH. >What was the sofware environment? As far as I know at this point Compaq >is shipping the Compaq 386 only with MS-DOS. This uses 16 bit compilers >and 286 programming models. In other words it treats the 386 as a fast 8088 >with all arithmetic being done 16 bits at a time. If the OS environment >was UNIX V.3 for the 386 then I would want to know more about what the >benchmarks are. Who is selling UNIX V.3 for the 386? Can you even buy a C compiler which supports 32 bit mode? Jan Gray jsgray@watmath University of Waterloo 519-885-5921
tomk@intsc.UUCP (Tom Kohrs) (11/21/86)
in article 3428@watmath.UUCP, Jan Gray <jsgray@watmath> writes: > Who is selling UNIX V.3 for the 386? Can you even buy a C compiler > which supports 32 bit mode? > Unix V.3 for the 386 is currently under development at Intel and Interactive Systems. It is expected that it will be available on the general market sometime early Q1'87. Compilers that will support the 32 bit native mode of the 386 will be available at the same time. These will include C compilers from Greenhills, LPI, Gold Hill and Metaware. -- ------ "Ever notice how your mental image of someone you've known only by phone turns out to be wrong? And on a computer net you don't even have a voice..." tomk@intsc.UUCP Tom Kohrs Regional Architecture Specialist Intel - Santa Clara
ben@catnip.UUCP (11/27/86)
In article <3428@watmath.UUCP> jsgray@watmath.UUCP (Jan Gray) writes: >Who is selling UNIX V.3 for the 386? Can you even buy a C compiler >which supports 32 bit mode? The Santa Cruz Operation is selling its 386 toolkit to people who want to develop software to run on Xenix/386. The package supposedly includes a C compiler which supports the i80386 modes, and a minimal 386 kernel to test the software. The cost is $395, but the package requires that you already own Xenix/286 v 2.1.3. -- Ben Broder {ihnp4,decvax} !hjuxa!catnip!ben {houxm,topaz}/
ron@brl-sem.ARPA (Ron Natalie <ron>) (11/30/86)
In article <3428@watmath.UUCP>, jsgray@watmath.UUCP (Jan Gray) writes: > > Who is selling UNIX V.3 for the 386? Can you even buy a C compiler > which supports 32 bit mode? > Interactive Systems (argh, what a pain) should be finishing up the System V port for the 386 any day now. Last I heard they had finished it for the MULTIBUS I system and were working on the Multibus II 386 card. -Ron
vance@sci.UUCP (Vance Turner) (12/06/86)
In article <499@brl-sem.ARPA>, ron@brl-sem.ARPA (Ron Natalie <ron>) writes: > In article <3428@watmath.UUCP>, jsgray@watmath.UUCP (Jan Gray) writes: > > Who is selling UNIX V.3 for the 386?