Article View: comp.sys.next.software
Article #1167Re: 3.0 Number Crunching Speed - a Reason to Relax
From: Matthew Dillon
Date: Fri, 25 Sep 1992 19:05
Date: Fri, 25 Sep 1992 19:05
28 lines
1551 bytes
1551 bytes
>In article <1992Sep24.070405.15244@udel.edu> pyrros@braindamaged.cis.udel.edu (Christos T. Pyrros) writes:
>>In article <1992Sep22.024416.19581@wam.umd.edu> pflikk@wam.umd.edu (Paul Glen Flikkema) writes:
>
>Correct me if I'm wrong, but when the 68881 and 68030 were "combined"
>to make the 68040, certain FPU functions such as trig functions were left
>out.
>
>It was my impression, however, that those calls were trapped by the CPU
>itself and executed in micro code... anyone know for sure?
>
>Chris
Transcendentals were left out of the 040. Motorola provides the trap
code and states that running transcendentals on an 040 through the trap
code is *FASTER* then running transcendentals inline on an 030/882
combination.
Non-transcendentals, the majority of what NeXTSTEP uses, run much, much
faster on an 040 then an 030/882 combination.
-Matt
--
Matthew Dillon dillon@Overload.Berkeley.CA.US
1005 Apollo Way uunet.uu.net!overload!dillon
Incline Village, NV. 89451 ham: KC6LVW (no mail drop)
USA Sandel-Avery Engineering (702)831-8000
Technical Details
Message-ID:
<dillon.0ocr@overload.Berkeley.CA.US>Path:
arc1-ca.pugleaf.net!sparky!uunet!overload!dillonReferences:
<1992Sep22.024416.19581@wam.umd.edu> <1992Sep24.070405.15244@udel.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!overload!dillon From: dillon@overload.Berkeley.CA.US (Matthew Dillon) Newsgroups: comp.sys.next.software Subject: Re: 3.0 Number Crunching Speed - a Reason to Relax Distribution: world Message-ID: <dillon.0ocr@overload.Berkeley.CA.US> References: <1992Sep22.024416.19581@wam.umd.edu> <1992Sep24.070405.15244@udel.edu> Date: 25 Sep 92 11:05:16 PST Organization: Not an Organization Lines: 29