From eeriksson@hanse.de Wed Sep 27 16:27:55 1995 Path: ashpoly!cascade!brgate!saturno!hanse!eeriksson From: eeriksson@hanse.de (Elaine Eriksson) Newsgroups: comp.dsp Subject: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep27.162755.3547@hanse.de> Date: Wed, 27 Sep 1995 16:27:55 GMT Organization: Hanse-Netzwerk Hamburg NNTP-Posting-Host: hanse.de X-Newsreader: NewsWatcher 2.0b24 Lines: 11 Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Elaine Eriksson "Extraordinary claims require extraordinary spectra." Atari Falcon 030 / NEXTSTEP 3.3 From bwatanabe@rwth-aachen.de Thu Sep 28 05:21:55 1995 Path: ashpoly!cascade!transat!ukc!fennimore!rwth-aachen!bwatanabe From: bwatanabe@rwth-aachen.de (Bastien Watanabe) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep28.052155.6126@rwth-aachen.de> Date: Thu, 28 Sep 1995 05:21:55 GMT References: <1995Sep27.162755.3547@hanse.de> Organization: RWTH Aachen Informatik NNTP-Posting-Host: rwth-aachen.de X-Newsreader: TIN [version 1.2 PL2] Lines: 10 In article <1995Sep27.162755.3547@hanse.de>, eeriksson@hanse.de (Elaine Eriksson) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Bastien Watanabe US Robotics Courier V.34 @ ISDN 64k bps "5 sigma is a discovery. 7 is a question." From tkarlsson@mindspring.com Sat Sep 30 01:18:55 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!mindspring!tkarlsson From: tkarlsson@mindspring.com (Tomas Karlsson) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep30.011855.9676@mindspring.com> Date: Sat, 30 Sep 1995 01:18:55 GMT References: <1995Sep27.162755.3547@hanse.de> <1995Sep28.052155.6126@rwth-aachen.de> Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: Netscape 1.1N Lines: 9 In article <1995Sep28.052155.6126@rwth-aachen.de>, bwatanabe@rwth-aachen.de (Bastien Watanabe) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Tomas Karlsson "Hold for relay." Pentium 90 (OS/2 Warp) / Linux 1.2.13 From tsuzuki@cam.ac.uk Mon Oct 02 01:08:55 1995 Path: ashpoly!cascade!mxgate!coyoacan!cam!tsuzuki From: tsuzuki@cam.ac.uk (Takeshi Suzuki) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Oct02.010855.9196@cam.ac.uk> Date: Mon, 02 Oct 1995 01:08:55 GMT References: <1995Sep27.162755.3547@hanse.de> <1995Sep28.052155.6126@rwth-aachen.de> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 10 In article <1995Sep28.052155.6126@rwth-aachen.de>, bwatanabe@rwth-aachen.de (Bastien Watanabe) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. My 14.4k connection dropped twice trying to fetch this thread. Keep articles concise! -- Takeshi Suzuki -- University of Cambridge Computer Laboratory "We don't sign stars. We build them." From esanchez@berkeley.edu Mon Oct 02 03:59:55 1995 Path: ashpoly!bigsur!pacrim!krgate!eunha!berkeley!esanchez From: esanchez@berkeley.edu (Élodie Sánchez) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Oct02.035955.8433@berkeley.edu> Date: Mon, 02 Oct 1995 03:59:55 GMT References: <1995Sep27.162755.3547@hanse.de> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 12 In article <1995Sep27.162755.3547@hanse.de>, eeriksson@hanse.de (Elaine Eriksson) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. If you are posting from AOL, please read RFC 1036 before posting again. -- Élodie Sánchez Hayes Optima 288 @ 9600 Telebit bps "In theory there is no difference between theory and practice."