Newsgroups: comp.speech,comp.dsp
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!gatech!howland.reston.ans.net!math.ohio-state.edu!uwm.edu!lll-winken.llnl.gov!noc.near.net!analog.com!analog.com!chris
From: chris.cavigioli@analog.com
Subject: Re: Request: half-rate GSM req. MIPS
Message-ID: <1995Mar21.145547.3045@analog.com>
Originator: chris@thor
Sender: usenet@analog.com (News pseudo-user usenet)
Organization: Analog Devices, Inc.
References: <3k9ekt$5sv@newsun.netmbx.de>
Date: Tue, 21 Mar 1995 14:55:47 GMT
Lines: 16
Xref: glinda.oz.cs.cmu.edu comp.speech:5036 comp.dsp:17808

In article <3k9ekt$5sv@newsun.netmbx.de> Jens-Peter Haack <peter@netcs.com> writes:
>i'm just reading the ETSI GSM 06.20 standard from March 1995
>and finding me unable to estimate the required DSP performance
>to do this CoDec.
>Is someone in this group able to estimate the required MIPS
>for this algorithm ? Has anyone implemented this stuff ?
>It's a VSELP CoDec, that tastes hard, but my first impression
>estimates about 2M Mults per second, resulting in somewhere
>between 5..15 MIPS Fixpoint DSP. But is that correct ??????
>   Jens-Peter Haack                        peter@NetCS.Com

The MIPS and memory estimates do not appear in the standard.
However, three conference papers contain most of this info.
See my post elsewhere today within comp.dsp

-chris.cavigioli@analog.com
