From tthorn@daimi.aau.dk Fri Jan 7 16:32:46 1994 Received: from avignon.daimi.aau.dk by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA28801; Fri, 7 Jan 1994 09:32:49 -0500 Received: by avignon.daimi.aau.dk id AA14023 (5.65c8/IDA-1.4.4 for Multiple recipients of list ); Fri, 7 Jan 1994 15:32:46 +0100 Date: Fri, 7 Jan 1994 15:32:46 +0100 From: Tommy Thorn Message-Id: <199401071432.AA14023@avignon.daimi.aau.dk> To: Multiple recipients of list Subject: ping Reply-To: Tommy.Thorn@daimi.aau.dk Just generating noice, to verify proper working riscy list, please ignore. (I haven't head a word since 27th of December.) /Tommy From VIGNANI%MSIE03%CRFV2@CSPCLU.CSP.IT Wed Jan 12 09:44:00 1994 Received: from cspsrv.csp.it by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA06657; Wed, 12 Jan 1994 04:44:14 -0500 Received: from DECNET-MAIL by CSP.IT (PMDF #12306) id <01H7LA0A0DUO000272@CSP.IT>; Wed, 12 Jan 1994 09:44 GMT Date: Wed, 12 Jan 1994 09:44 GMT From: VIGNANI%MSIE03%CRFV2@CSPCLU.CSP.IT Subject: Hi folks. May I help? To: riscy@sunsite.unc.edu Message-Id: <01H7LA0A0DUO000272@CSP.IT> X-Envelope-To: riscy@sunsite.unc.edu X-Vms-To: inet"riscy@sunsite.unc.edu" Hi there. I'm new to this list, and I'm curious about what you're doing. I run linux on my 486 PCs since one year and I like it; but I feel it's time to upgrade to a new CPU, given also NT is coming. I am an hardware and firmware developer with previous experience in designing, building and programming: - TI34010 GSP - DSP56000..56002 - TI320C30 - all flavors of Transputers (they called it a RISC back in 1987.. :-) - AD1848 SoundPort I have a full set of EDA tools available, both on PC and on workstations. My current job involves PCB design and DSP programming. Is this useful to you? Greetings Alberto Vignani vignani.crf@csp.it ---------------------------------------------------------------------- "This phrase begun to translate itself into English, then suddenly ha cambiato idea" From broadley@neurocog.lrdc.pitt.edu Tue Jan 18 14:25:03 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA27815; Tue, 18 Jan 1994 19:25:10 -0500 Message-Id: <199401190025.AA27815@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA06880; Tue, 18 Jan 94 19:25:03 -0500 From: Bill Broadley Subject: PING To: riscy@sunsite.unc.edu Date: Tue, 18 Jan 1994 19:25:03 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 311 Hello just seeing if the list is up. Any new information? Is the list so quite because we have ended up with a design people are unhappy with? -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From andy@waldorf-gmbh.de Wed Jan 19 09:54:54 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA09349; Wed, 19 Jan 1994 02:53:51 -0500 Received: by mail.Germany.EU.net(EUnetD-2.3.1.c) via EUnet id fA20591; Wed, 19 Jan 1994 08:54:50 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 19 Jan 94 08:54:54 +0100 Message-Id: <9401190754.AA09532@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: PING > Hello just seeing if the list is up. Yes, I guess so. > Any new information? No. I must apologize, but NEC Germany did not receive the promised design kit yet and has no information when it will be available. I hate that shit and I'm *very* near to give up. > Is the list so quite because we have ended up with a design people are > unhappy with? Don't know. If you like, get another design. Sorry, I can imagine that you won't be happy with this statement, but it's the truth. I gave my best, but this don't seem to be sufficient. Andy From tthorn%hof@hof.daimi.aau.dk Wed Jan 19 04:57:30 1994 Received: from hof (hof.daimi.aau.dk) by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA16344; Wed, 19 Jan 1994 04:57:30 -0500 Received: by hof (Linux Smail3.1.28.1 #14) id m0pMZei-0000ZfC; Wed, 19 Jan 94 10:56 MET Message-Id: Date: Wed, 19 Jan 94 10:56 MET From: tthorn%hof@hof.daimi.aau.dk (Tommy Thorn) To: Multiple recipients of list Subject: Re: PING In-Reply-To: <9401190754.AA09532@resi.waldorf-gmbh.de> References: <9401190754.AA09532@resi.waldorf-gmbh.de> Reply-To: Tommy.Thorn@daimi.aau.dk Andreas Busse writes: > No. I must apologize, but NEC Germany did not receive > the promised design kit yet and has no information when > it will be available. > I hate that shit and I'm *very* near to give up. > > > Is the list so quite because we have ended up with a design people are > > unhappy with? > > Don't know. If you like, get another design. > Sorry, I can imagine that you won't be happy with this > statement, but it's the truth. I gave my best, but this > don't seem to be sufficient. Well, I fully understand. It's a sad situation indeed. I still like the design very much, though I would have liked an external cache, but I like lower cost more. We are all aware that you been doing an immense amount of work to make this a success, and I appreciate this very much. I just sorry I'm not able to contribute to the HW part. I the riscy project dies, I'll be sad, but I'll still be looking for a replacement motherboard with a resonable price and docs. I do alot of compiler work, and optimizing for the intel just make me feel bad. The MIPS (esp. 4200) is still my favorite, due to the superior price/performance relation. If we are to belive some of the market analysis and SGI's own commercial, mips motherboards will become generally available anyway, so a Linux (and VSTa ;^) port might still happen. I think we should give it more time still. Either we finally do get something from NEC (or SGI), or alternatives might crop up we could consider. If anyone have any better ideas, I'm eager to hear them. -- Tommy.Thorn@daimi.aau.dk Staff-programmer Aarhus University, Ny Munkegade 116 Phone: +45 89423223 DK-8000 Aarhus C, Denmark. Fax: +45 86135725 PGP Public Key fingerprint: E7B1175FC30D9E96B67AF61D89A70A1F From tor@tss.no Wed Jan 19 12:22:14 1994 Received: from benoni.Uit.No by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA17730; Wed, 19 Jan 1994 05:32:05 -0500 Received: from benoni by ppenoni.uit.no with SMTP (PP) id <20839-0@ppenoni.uit.no>; Wed, 19 Jan 1994 11:31:59 +0000 Received: from unas.tss.no by benoni.uit.no (5.65+IDA/Babel-1.15/ABaa-1.2/Ultrix) id AAbenoni20835; Wed, 19 Jan 1994 11:31:54 +0100 Received: by unas.tss.no (4.0/ABaa-1.3mini) id AA02837; Wed, 19 Jan 94 11:22:16 +0100 Message-Id: <9401191022.AA02837@unas.tss.no> From: tor@tss.no (Tor Arntsen) Date: Wed, 19 Jan 1994 11:22:14 +0100 X-Mailer: Mail User's Shell (7.2.5 10/14/92) To: riscy@sunsite.unc.edu Subject: Re: PING I'm still here, and I'm still interested. And patient. Tor From berge@sierra.nl Wed Jan 19 12:46:33 1994 Received: from gatekeeper.sierra.nl by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA18418; Wed, 19 Jan 1994 05:48:07 -0500 Received: from quint.sierra.nl.sierra.nl by gatekeeper.sierra.nl (4.1/2.00) id AA11381; Wed, 19 Jan 94 11:46:33 +0100 Date: Wed, 19 Jan 94 11:46:33 +0100 From: berge@sierra.nl (Wim van den Berge) Message-Id: <9401191046.AA11381@gatekeeper.sierra.nl> To: riscy@sunsite.unc.edu Subject: Re: PING Andreas Busse writes: > No. I must apologize, but NEC Germany did not receive > the promised design kit yet and has no information when > it will be available. > I hate that shit and I'm *very* near to give up. > > > Is the list so quite because we have ended up with a design people are > > unhappy with? > > Don't know. If you like, get another design. > Sorry, I can imagine that you won't be happy with this > statement, but it's the truth. I gave my best, but this > don't seem to be sufficient. I, and I assume everybody else on this list, do appriciate the time and effort you put into this project. As a software kind of guy I am unable to help very much with the hardware of the riscy project, but I do like the design we ended up with. Tommy Thorn writes: > I the riscy project dies, I'll be sad, but I'll still be looking for a > replacement motherboard with a resonable price and docs. I do alot of > compiler work, and optimizing for the intel just make me feel bad. The > MIPS (esp. 4200) is still my favorite, due to the superior > price/performance relation. I could not agree more, if riscy fails I will be looking for a MIPS 4200 based replacement too. Come on people, I think this project has too much going for it to quit now, I have been on this list for 5 months now and I believe there is more than enough talent and know-how gathered here to make this project a success. Lets hang in there and try to convince the big boys (NEC etc) to take us serious. Any suggestions ??? Wim van den Berge Sierra Semiconductor B.V. Tel : +31 73408888 (:> \/\/i/\/\ Fax : +31 73423155 (_) E-mail : berge@sierra.nl =======""====== From andy@waldorf-gmbh.de Wed Jan 19 13:16:37 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA19445; Wed, 19 Jan 1994 06:15:36 -0500 Received: by mail.Germany.EU.net(EUnetD-2.3.1.c) via EUnet id fD02364; Wed, 19 Jan 1994 12:16:33 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 19 Jan 94 12:16:37 +0100 Message-Id: <9401191116.AA00389@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: PING Hi everybody ! Thanks for the motivation ! To keep things running while we are waiting for the design, I'll try to make the promised cost estimation. It's not that easy since NEC/Mips use some parts I never heard about... Perhaps somebody knows what a "XC1736APD8C" is ??? There's no explanation in the part list I got from Mips (and there's obviosly nobody I can ask). As soon I got prices for the most important parts, I'll post a summary. I cannot really promise this, but you can expect prices next week, perhaps on Monday. If someone needs a technical description, let me know. Bye, Andy From hodgen@mailhost.uni-koblenz.de Wed Jan 19 06:24:18 1994 Received: from mailhost.uni-koblenz.de by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA19732; Wed, 19 Jan 1994 06:24:18 -0500 Received: from ozzy.uni-koblenz.de by mailhost.uni-koblenz.de with smtp (Smail3.1.28.1) id m0pMb10-0001bKC; Wed, 19 Jan 94 12:24 MET Message-Id: Received: by ozzy.uni-koblenz.de (4.1/KO-2.0) id AA10915; Wed, 19 Jan 94 12:25:33 +0100 From: Subject: Re: PING To: riscy@sunsite.unc.edu Date: Wed, 19 Jan 94 12:25:31 MET In-Reply-To: <9401190754.AA09532@resi.waldorf-gmbh.de>; from "Andreas Busse" at Jan 19, 94 3:00 am X-Mailer: ELM [version 2.3 PL11] > No. I must apologize, but NEC Germany did not receive > the promised design kit yet and has no information when > it will be available. > I hate that shit and I'm *very* near to give up. I hate it too, I can't wait to get a MIPS. > > Is the list so quite because we have ended up with a design people are > > unhappy with? Who's unhappy? You? Speak up those who are unhappy. We don't even know what the design is *because* NEC hasn't sent anything. How can you be unhappy with something you haven't seen? > Don't know. If you like, get another design. > Sorry, I can imagine that you won't be happy with this > statement, but it's the truth. I gave my best, but this > don't seem to be sufficient. It's more than enough Andy. Come on people, show a little support. -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | Have you heard about the Uni Koblenz, | or Fight-o-net 2:245/5600.11 | riscy project? A GPL based Rheinau 1, | Voice: +49 261 9119-645 | design for an R4000 PC MB. 56075 Koblenz. | Fax: +49 261 9119-499 | MIPS, Power to the people! From Jarle.F.Greipsland@idt.unit.no Wed Jan 19 13:32:51 1994 Received: from pat.idt.unit.no by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA20106; Wed, 19 Jan 1994 06:33:02 -0500 Received: from drue.idt.unit.no by pat.idt.unit.no with SMTP id AA05727 (5.65c8/IDA-1.4.4 for ); Wed, 19 Jan 1994 12:32:51 +0100 Received: by drue.idt.unit.no id AA13265 (5.65c8/idt-C-2.0); Wed, 19 Jan 1994 12:32:51 +0100 Date: Wed, 19 Jan 1994 12:32:51 +0100 From: Jarle.F.Greipsland@idt.unit.no Message-Id: <199401191132.AA13265@drue.idt.unit.no> To: riscy@sunsite.unc.edu Cc: andy@waldorf-gmbh.de In-Reply-To: Andreas Busse's message of Wed, 19 Jan 1994 06:16:27 -0500 <9401191116.AA00389@resi.waldorf-gmbh.de> Subject: Re: PING Andreas Busse writes: ab> Perhaps somebody knows what a "XC1736APD8C" is ??? There's ab> no explanation in the part list I got from Mips (and there's ab> obviosly nobody I can ask). The XC1736APD8C is a 36Kbit serial EPROM from Xilinx Inc. XC - Xilinx Corp. 1736 - 17 == serial EPROM family, 36==#Kbits A - no programmable reset polarity PD - plastic dual-inline package 8 - 8 pins C - commercial variant This is probably for the boot code loading which I think take place on a few dedicated pins (ModeClock and ModeIn?). -jarle ---- "People who deal with bits should expect to get bitten." -- Jon Bentley From andy@waldorf-gmbh.de Wed Jan 19 13:39:15 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA20316; Wed, 19 Jan 1994 06:38:11 -0500 Received: by mail.Germany.EU.net(EUnetD-2.3.1.c) via EUnet id fE03433; Wed, 19 Jan 1994 12:39:09 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 19 Jan 94 12:39:15 +0100 Message-Id: <9401191139.AA00441@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: PING > Who's unhappy? You? Speak up those who are unhappy. We don't even know > what the design is *because* NEC hasn't sent anything. How can you be > unhappy with something you haven't seen? That's actually not true. Everyone who has ever seen a Mips Magnum 4000PC/50 or even worked on that box knows what we (hopefully) will get. I like it :-) > It's more than enough Andy. Come on people, show a little support. Thanks ! Andy From andy@waldorf-gmbh.de Mon Jan 24 15:03:25 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA25362; Mon, 24 Jan 1994 08:02:15 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:6/EUnetD-2.4.0.m) via EUnet id OAA07105; Mon, 24 Jan 1994 14:03:14 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Mon, 24 Jan 94 14:03:25 +0100 Message-Id: <9401241303.AA09043@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: EMULEX Hi there, besides the fact that I got an order number for the ARC 100/150 kits (which at least means that I can order it now!) I'm looking for a german distributor of EMULEX. If someone knows, please mail me. Thanks ! Andy From andy@waldorf-gmbh.de Mon Jan 24 16:10:15 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA29571; Mon, 24 Jan 1994 09:09:13 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:6/EUnetD-2.4.0.n) via EUnet id PAA11200; Mon, 24 Jan 1994 15:10:05 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Mon, 24 Jan 94 15:10:15 +0100 Message-Id: <9401241410.AA09245@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: andy@waldorf-gmbh.de, ronald@osf.uci.kun.nl Subject: Re: EMULEX Cc: riscy@sunsite.unc.edu > This is GREAT! at last some movement. Sorry I cannot give the answer on > EMULEX. Hope to run Linux on R4k2 soon. I cannot promise a R4K2 solution since the R4K2 needs 3.3V supply. At least some sort of adaptor with voltage converters must be made to allow this. Perhaps someone else can do that ? I'd rather recommend the use of the IDT Orion (R4600) which runs with 5V and is faster than the R4200. Cheers, Andy From andy@waldorf-gmbh.de Mon Jan 24 16:44:22 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA02358; Mon, 24 Jan 1994 09:43:14 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:7/EUnetD-2.4.0.n) via EUnet id PAA12880; Mon, 24 Jan 1994 15:44:10 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Mon, 24 Jan 94 15:44:22 +0100 Message-Id: <9401241444.AA09315@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: Emulex Forget my question about Emulex. I found the distributor by accident :-) If I remember right someone asked for the SCSI Controller on the Mips board. It's the EMULEX FAS216. I have currently no documentation for it, so please don't ask me for details. In summary, here's again the list of the major parts: CPU: R4000PC or R4400PC or R4600PC or R4200 with voltage conversion adaptor (to be designed). SCSI: Emulex FAS216 Ethernet: National 83932BVF Serial I/O: Western Digital 16C552 Keyboard I/O: Intel 8242 Audio: Crystal CS4215 Video: Inmos G364 Realtime Clk: Dallas DS1287 Floppy: Intel 82077A EISA chipset: Intel 82352 (2x), 82357, 82358 Bus Ctrl.: NEC uPD31431 (2x), uPD31432 I'm still working on the cost estimation. As soon it's finished I'll let you know. I'll give you prices w/o CPU plus the german prices for the CPUs. I guess you probably can get them for less money in US. (R4600 is currently us$493 in qtys of 10, R4400 costs twice). Andy From atk@agua.colorado.edu Mon Jan 24 02:11:48 1994 Received: from boulder.Colorado.EDU by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11418; Mon, 24 Jan 1994 11:11:35 -0500 Received: from agua.Colorado.EDU by boulder.Colorado.EDU with SMTP id AA03438 (5.65c/IDA-1.4.4 for ); Mon, 24 Jan 1994 09:11:25 -0700 Received: by agua.colorado.edu (cu.generic.890828) Date: Mon, 24 Jan 94 09:11:48 MST From: Alan Krantz Message-Id: <9401241611.AA06156@agua.colorado.edu> To: andy@waldorf-gmbh.de, riscy@sunsite.unc.edu Subject: Re: EMULEX Yes i would think the R4600 would be a much better solution. They are very similar in cost and the R4600 is about 3 times faster (for fp). atk From amoss@cs.huji.ac.il Mon Jan 24 20:31:07 1994 Received: from cs.huji.ac.il by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA13551; Mon, 24 Jan 1994 11:31:15 -0500 Received: from picton.cs.huji.ac.il by cs.huji.ac.il with SMTP id AA22521 (5.65c/HUJI 4.153 for ); Mon, 24 Jan 1994 18:31:09 +0200 Received: from localhost by picton.cs.huji.ac.il with SMTP id AA11988 (5.65c/HUJI 4.1 for riscy@sunsite.unc.edu); Mon, 24 Jan 1994 18:31:08 +0200 Message-Id: <199401241631.AA11988@picton.cs.huji.ac.il> To: riscy@sunsite.unc.edu Subject: Re: EMULEX In-Reply-To: Your message of Mon, 24 Jan 1994 11:16:35 -0500 . <9401241611.AA06156@agua.colorado.edu> From: Amos Shapira Date: Mon, 24 Jan 1994 18:31:07 +0200 Sender: amoss@CS.HUJI.AC.IL atk@agua.colorado.edu writes: |Yes i would think the R4600 would be a much better solution. They are |very similar in cost and the R4600 is about 3 times faster (for fp). | |atk Maybe when we see what are the costs of the other parts and if someone can find specific details about the prices of CPU's in the U.S. and Europe (because I'm NOT going to buy the CPU in Israel, no thanks :) we will be able to make a wiser decision. Cheers, --Amos --Amos Shapira (Jumper Extraordinaire) | "War does not determine who is right, C.S. System Group, Hebrew University, | but who is left" Jerusalem 91904, ISRAEL | amoss@cs.huji.ac.il | -- Anonymous? From andy@waldorf-gmbh.de Mon Jan 24 19:18:53 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA19621; Mon, 24 Jan 1994 12:18:04 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:7/EUnetD-2.4.0.n) via EUnet id SAA22555; Mon, 24 Jan 1994 18:18:45 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Mon, 24 Jan 94 18:18:53 +0100 Message-Id: <9401241718.AA09498@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: amoss@cs.huji.ac.il Subject: Re: EMULEX Cc: riscy@sunsite.unc.edu > Maybe when we see what are the costs of the other parts and if someone > can find specific details about the prices of CPU's in the U.S. and > Europe (because I'm NOT going to buy the CPU in Israel, no thanks :) > we will be able to make a wiser decision. Oh, perhaps it was a bit misleading what I said... We *will* offer CPUs, but you *need not* to buy them from us. Cheers, Andy From andy@waldorf-gmbh.de Tue Jan 25 12:32:03 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA06171; Tue, 25 Jan 1994 05:30:57 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:7/EUnetD-2.4.0.n) via EUnet id LAA24677; Tue, 25 Jan 1994 11:31:55 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Tue, 25 Jan 94 11:32:03 +0100 Message-Id: <9401251032.AA10562@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: ARC 100 Prices Hi everybody, I have made a very rough cost estimation. Some of the prices are varying from day to day. Others aren't really up-to-date because I couldn't get the latest price in time. So don't wonder, the price will change but I don't know how much. However, our estimated retail prices: Motherboard w/o RAM and CPU, DM 2000.-- / us$1185.-- incl. SCSI, serial/parallel, FDC, Ethernet, EISA Ctrl. etc. Video sub board, incl. 2 Meg VRAM DM 700.-- / us$410.-- Audio sub board DM 950.-- / us$550.-- Ethernet AUI sub board DM 70.-- / us$40.-- Keyboard Ctrl. sub board DM 55.-- / us$33.-- current CPU prices in qtys of 10: (no retail price, this is the price we pay) R4400PC/50 DM 1647,-- / us$968.-- R4600-100 DM 839.-- / us$493,-- Well, this is NO ANNOUNCEMENT, rather a RFC. The estimation is based on Qtys of 100, except the CPUs. So far, so good. Now send flames :-) Cheers, Andy From df@eyrie.demon.co.uk Wed Jan 26 01:18:04 1994 Received: from eyrie.demon.co.uk by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA27072; Wed, 26 Jan 1994 01:34:54 -0500 Received: by eyrie.demon.co.uk (Smail3.1.28.1 #9) id m0pOxxN-00014PC; Wed, 26 Jan 94 00:18 GMT Message-Id: From: df@eyrie.demon.co.uk (Derek Fawcus) Subject: Re: Emulex - Alternative support chips? To: riscy@sunsite.unc.edu Date: Wed, 26 Jan 1994 00:18:04 +0100 (GMT) In-Reply-To: <9401241444.AA09315@resi.waldorf-gmbh.de> from "Andreas Busse" at Jan 24, 94 09:43:52 am X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 878 I know that the intention was to use the ARC set design, but how much scope is there for change - assuming that it'd be worth the cost? Two items that come to mind were devices mentioned earlier on the list. > Serial I/O: Western Digital 16C552 > Floppy: Intel 82077A Could these be replaced by the SMC FDC37c666 (plus gain a // port) ? > Keyboard I/O: Intel 8242 > Realtime Clk: Dallas DS1287 Replace by the VLSI VL82C114 (RTC / NVRAM / KBD CONT / PS/2 MOUSE). I assume that the Dallas device includes a built in Lithium Battery, and I know that the VLSI device will need an external Battery. But then I'd prefer the external one! Just Idea's incase they can reduce costs, as the mentioned devices should (potentially after configuration commands), be identical to the normal devices. DF -- Derek Fawcus (G7FVS) df@eyrie.demon.co.uk From andy@waldorf-gmbh.de Wed Jan 26 10:15:22 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA03185; Wed, 26 Jan 1994 03:14:34 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:13/EUnetD-2.4.1.c) via EUnet id JAA12231; Wed, 26 Jan 1994 09:15:10 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 26 Jan 94 09:15:22 +0100 Message-Id: <9401260815.AA12641@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: andy@waldorf-gmbh.de, broadley@neurocog.lrdc.pitt.edu Subject: Re: ARC 100 Prices Cc: riscy@sunsite.unc.edu > Do I understand this correctly? us$1185=410+550+40+33+motherboard$$$???? > So the motherboard alone cost $152?? (I hope so) Sorry, NO. Motherboard price *without* any of the other cards is us$1185. > > Video sub board, incl. 2 Meg VRAM DM 700.-- / us$410.-- > > Does this make sense compared to the nice S3-928 video boards > (probably even eisa) for $300 or so? (the s3 greatly reduces the > load on the bus) No, it doesn't make sense, you're right. But you don't need to buy the ARC video board. You can use *any* video card you like, even a $15 no-name 256k VGA card. > > Audio sub board DM 950.-- / us$550.-- > > OUCH sounds way to much. 44.1 KHz 16 bit stereo play/record > cost less then $200 for ISA maybe even eisa. Sound boards don't > push the bus very hard. Right again. The audio sub board is so expensive because of 4 (four!) large Altera FPGAs for DMA handling. Get something else if you like. I don't recommend to use the ARC board (although it sounds good, I have it in my Mips box). > > Keyboard Ctrl. sub board DM 55.-- / us$33.-- > > Seems a prime candidate for putting on the motherboard, but it's cheap > either way. You're right again, but our goal is *not* to redesign the board. That's why the whole thing is splitted into several boards. > current CPU prices in qtys of 10: > (no retail price, this is the price we pay) > > R4400PC/50 DM 1647,-- / us$968.-- > R4600-100 DM 839.-- / us$493,-- > [comparison deleted] > I believe Drew offered to handle the linux scsi driver for the NCR53cf94 > (very nice chip from what I hear) chip if we decided to use it. I'm not > sure this offer still stands. Someone wrote that the Emulex FAS216 is compatible with the NCR 53cf94. > A: Change the motherboard to use NCR chip and take advantage of Drew's > help? > B: Or leave the motherboard design as is an write our own scsi driver for > this unknown (to me) emulex chip? I worry about the emulex because > I haven't heard of it and linux doesn't support it yet. See above. > I personally could probably afford something more along the lines of a > R4200 ($150) + motherboard ($150) + cheap eisa video ($150) + keyboard $33 Forget that. I can't offer a board for $150 if I pay $300 for the bus controller chips alone. To give you an idea how the boards are calculated, I give you the formula: Part costs + Board costs + Production & Testing ---------------------- our costs our costs * 1.38 = retail price. Perhaps you can imagine that we *need* some so-called profit to handle support, repair service and so on. As you can see, our initial costs of around us$10,000 for buying the design, setting up production etc. aren't included in the calculation. What do you think how many of these boards we could sell to the riscy group ? 10 or 20 ? So, please, don't talk about profit. There is none. Cheers, Andy From andy@waldorf-gmbh.de Wed Jan 26 10:18:46 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA03343; Wed, 26 Jan 1994 03:17:56 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:13/EUnetD-2.4.1.c) via EUnet id JAA12384; Wed, 26 Jan 1994 09:18:38 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 26 Jan 94 09:18:46 +0100 Message-Id: <9401260818.AA12660@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: andy@waldorf-gmbh.de, broadley@neurocog.lrdc.pitt.edu Subject: Re: ARC 100 Prices Cc: riscy@sunsite.unc.edu Bill Broadley wrote: > Please send the previous message back to me, or to the list (it was meant > to go to the list, as well as you). Here it comes. Cheers, Andy ------------------------------snip------------------------------ > I have made a very rough cost estimation. > Some of the prices are varying from day to day. > Others aren't really up-to-date because I couldn't > get the latest price in time. > So don't wonder, the price will change but I don't > know how much. Rough estimation is appreciated. > Motherboard w/o RAM and CPU, DM 2000.-- / us$1185.-- > incl. SCSI, serial/parallel, > FDC, Ethernet, EISA Ctrl. etc. Do I understand this correctly? us$1185=410+550+40+33+motherboard$$$???? So the motherboard alone cost $152?? (I hope so) > Video sub board, incl. 2 Meg VRAM DM 700.-- / us$410.-- Does this make sense compared to the nice S3-928 video boards (probably even eisa) for $300 or so? (the s3 greatly reduces the load on the bus) > Audio sub board DM 950.-- / us$550.-- OUCH sounds way to much. 44.1 KHz 16 bit stereo play/record cost less then $200 for ISA maybe even eisa. Sound boards don't push the bus very hard. > Ethernet AUI sub board DM 70.-- / us$40.-- Reasonable. > Keyboard Ctrl. sub board DM 55.-- / us$33.-- Seems a prime candidate for putting on the motherboard, but it's cheap either way. > current CPU prices in qtys of 10: > (no retail price, this is the price we pay) > > R4400PC/50 DM 1647,-- / us$968.-- > R4600-100 DM 839.-- / us$493,-- I have a postscript doc comparing the 4600 to the 4400, and is intended to guide system designers who wish to do a single system design which can be used with either the R4200 or the R4600 processor. (sounds like it's written for us) sgigate.sgi.com:/pub/doc/r4x00/R4x00_common_design.ps Heres a quick comparison: R4600 R4400PC Cache 16/16 2 way associative 16/16 direct mapped Cache miss 3 cycles 5-8 cycles. load latency 2 cycles 3 cycles Branch latency 2 cycles 4 cycles Uncached store 4 doublewords (4 addresses) 1 doubleword doubles as write buffer Int mult 4 cycles to issue 1 cycle to issue Int divide 4 cycles to issue 69 cycles to issue I don't think theres any reason to consider the R4400PC. I think the R4200 will be VERY important to the affordability of the project. So the question should be R4200 or R4600? I believe Drew offered to handle the linux scsi driver for the NCR53cf94 (very nice chip from what I hear) chip if we decided to use it. I'm not sure this offer still stands. I ask anyone out there which would be the bigger win? A: Change the motherboard to use NCR chip and take advantage of Drew's help? B: Or leave the motherboard design as is an write our own scsi driver for this unknown (to me) emulex chip? I worry about the emulex because I haven't heard of it and linux doesn't support it yet. I personally could probably afford something more along the lines of a R4200 ($150) + motherboard ($150) + cheap eisa video ($150) + keyboard $33 From broadley@neurocog.lrdc.pitt.edu Wed Jan 26 06:21:09 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11119; Wed, 26 Jan 1994 11:21:10 -0500 Message-Id: <199401261621.AA11119@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA12025; Wed, 26 Jan 94 11:21:09 -0500 From: Bill Broadley Subject: Arc 100 prices.. To: riscy@sunsite.unc.edu Date: Wed, 26 Jan 1994 11:21:09 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 2992 > I have made a very rough cost estimation. > Some of the prices are varying from day to day. > Others aren't really up-to-date because I couldn't > get the latest price in time. > So don't wonder, the price will change but I don't > know how much. A rough estimation is appreciated. > Motherboard w/o RAM and CPU, DM 2000.-- / us$1185.-- > incl. SCSI, serial/parallel, > FDC, Ethernet, EISA Ctrl. etc. Do I understand this correctly? us$1185=410+550+40+33+motherboard$$$???? So the motherboard alone cost $152?? (I hope so) > Video sub board, incl. 2 Meg VRAM DM 700.-- / us$410.-- Does this make sense compared to the nice S3-928 video boards (probably even eisa) for $300 or so? (the s3 greatly reduces the load on the bus) > Audio sub board DM 950.-- / us$550.-- OUCH sounds way to much. 44.1 KHz 16 bit stereo play/record cost less then $200 for ISA maybe even eisa. Sound boards don't push the bus very hard. > Ethernet AUI sub board DM 70.-- / us$40.-- Reasonable. > Keyboard Ctrl. sub board DM 55.-- / us$33.-- Seems a prime candidate for putting on the motherboard, but it's cheap either way. > current CPU prices in qtys of 10: > (no retail price, this is the price we pay) > > R4400PC/50 DM 1647,-- / us$968.-- > R4600-100 DM 839.-- / us$493,-- I have a postscript doc comparing the 4600 to the 4400, and is intended to guide system designers who wish to do a single system design which can be used with either the R4200 or the R4600 processor. (sounds like it's written for us) sgigate.sgi.com:/pub/doc/r4x00/R4x00_common_design.ps Heres a quick comparison: R4600 R4400PC Cache 16/16 2 way associative 16/16 direct mapped Cache miss 3 cycles 5-8 cycles. load latency 2 cycles 3 cycles Branch latency 2 cycles 4 cycles Uncached store 4 doublewords (4 addresses) 1 doubleword doubles as write buffer Int mult 4 cycles to issue 1 cycle to issue Int divide 4 cycles to issue 69 cycles to issue I don't think theres any reason to consider the R4400PC. I think the R4200 will be VERY important to the affordability of the project. So the question should be R4200 or R4600? I believe Drew offered to handle the linux scsi driver for the NCR53cf94 (very nice chip from what I hear) chip if we decided to use it. I'm not sure this offer still stands. I ask anyone out there which would be the bigger win? A: Change the motherboard to use NCR chip and take advantage of Drew's help? B: Or leave the motherboard design as is an write our own scsi driver for this unknown (to me) emulex chip? I worry about the emulex because I haven't heard of it and linux doesn't support it yet. I personally could probably afford something more along the lines of a R4200 ($150) + motherboard ($150) + cheap eisa video ($150) + keyboard $33 -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From andy@waldorf-gmbh.de Wed Jan 26 18:23:04 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11218; Wed, 26 Jan 1994 11:21:56 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:13/EUnetD-2.4.1.e) via EUnet id RAA09987; Wed, 26 Jan 1994 17:22:52 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Wed, 26 Jan 94 17:23:04 +0100 Message-Id: <9401261623.AA13789@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Who's there ??? Hi everybody ! Now that I posted a rough price estimation I'd like to hear what you all think about, more exactly: I need votes. If those two or three guys who sent comments are the only ones on the list who are still interested, I'd say: forget about riscy. Cheers, Andy From tim@ubitrex.mb.ca Wed Jan 26 08:22:18 1994 Received: from ubitrex.mb.ca (ubiserver.ubitrex.mb.ca) by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA20614; Wed, 26 Jan 1994 15:22:36 -0500 Received: from ska.ubitrex.mb.ca ([192.75.16.23]) by ubitrex.mb.ca (4.1/SMI-4.1) id AA24567; Wed, 26 Jan 94 14:22:18 CST Date: Wed, 26 Jan 94 14:22:18 CST From: tim@ubitrex.mb.ca (Tim Braun) Message-Id: <9401262022.AA24567@ubitrex.mb.ca> To: riscy@sunsite.unc.edu Subject: Re: ARC 100 Prices Let me be the first to say, Wow. That's higher than we had hoped. Tim Braun | Ubitrex Corporation | Voice: 204-942-2992 ext 228 1900-155 Carlton St | FAX: 204-942-3001 Winnipeg, Manitoba, Canada R3C 3H8 | Email: tim@ubitrex.mb.ca From bwojcik@cs.tamu.edu Thu Jan 27 14:45:52 1994 Received: from cs.tamu.edu (clavin.cs.tamu.edu) by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA16805; Thu, 27 Jan 1994 21:47:17 -0500 Received: from sparc68.cs.tamu.edu (813@sparc68.cs.tamu.edu [128.194.132.168]) by cs.tamu.edu (8.6.4/8.6.4) with ESMTP id UAA22179 for ; Thu, 27 Jan 1994 20:46:25 -0600 From: Brian J Wojcik Received: from localhost (bwojcik@localhost) by sparc68.cs.tamu.edu (8.6.4/8.6.4) id UAA10847 for riscy@sunsite.unc.edu; Thu, 27 Jan 1994 20:45:54 -0600 Message-Id: <199401280245.UAA10847@sparc68.cs.tamu.edu> Subject: Re: Who's there ??? To: riscy@sunsite.unc.edu Date: Thu, 27 Jan 1994 20:45:52 -0600 (CST) In-Reply-To: <9401261623.AA13789@resi.waldorf-gmbh.de> from "Andreas Busse" at Jan 27, 94 09:23:18 pm X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 822 > > > Hi everybody ! > > Now that I posted a rough price estimation I'd like > to hear what you all think about, more exactly: > > I need votes. > > If those two or three guys who sent comments are > the only ones on the list who are still interested, > I'd say: forget about riscy. > > Cheers, > Andy > I've been watching this discussion for a while with moderate interest. However, I would be hard pressed to come up with $800-$1000 anything over a thousand dollars would be definitely out of the question. As I am just a student though, I'm not sure I could come up with even $400 dollars for a motherboard ;). I wonder if this group's lost its focus on producing a cheap mips board or if (more likely) cis dealing with definitions of cheap other than mine. -Regards, -Brian Wojcik -bwojcik@cs.tamu.edu From johnsonm Thu Jan 27 16:50:26 1994 Received: from localhost.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA17113; Thu, 27 Jan 1994 21:50:27 -0500 Message-Id: <199401280250.AA17113@SunSITE.Unc.EDU> To: Multiple recipients of list Reply-To: johnsonm Subject: Re: Who's there ??? In-Reply-To: Your message of "Thu, 27 Jan 94 21:22:20 EST." <9401261623.AA13789@resi.waldorf-gmbh.de> Date: Thu, 27 Jan 94 21:50:26 -0500 From: ""Michael K. Johnson"" Andreas Busse writes: >Now that I posted a rough price estimation I'd like >to hear what you all think about, more exactly: > > I need votes. > >If those two or three guys who sent comments are >the only ones on the list who are still interested, >I'd say: forget about riscy. Well, we were talking about prices around (below) $1000 including CPU before. The prices you have quoted are well above my budget, and I think the original point of the riscy group isn't well served by those prices. Whatever happened to CPU's for $85? Before Waldorf came into the picture, powerPC was rejected for CPU prices lower than the ones that you posted. DEC has an alpha chip which is almost the entire motherboard on one chip, with significantly higher performance, from the figures I've seen, for prices similar to the CPU alone as you posted. I mentioned $1000 for the whole motherboard, not including RAM, but including *everything* else (including CPU), as the highest I would possibly consider, although I don't remember if that was in private email to you (Andy) or publically on the group. I did see several similar posts, and it seemed some time ago that that was the consensus. At those prices, yeah, I'd have to say, forget about riscy, as far as I'm concerned. Sorry, michaelkjohnson From ksh@charybdis.prl.ufl.edu Thu Jan 27 17:02:17 1994 Received: from charybdis.prl.ufl.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA17983; Thu, 27 Jan 1994 22:00:06 -0500 Received: by charybdis.prl.ufl.edu (4.1/1.34) id AA22112; Thu, 27 Jan 94 22:02:17 EST Date: Thu, 27 Jan 94 22:02:17 EST From: ksh@charybdis.prl.ufl.edu (Kevin S Ho) Message-Id: <9401280302.AA22112@charybdis.prl.ufl.edu> To: broadley@neurocog.lrdc.pitt.edu Cc: riscy@sunsite.unc.edu In-Reply-To: <199401261621.AA11119@SunSITE.Unc.EDU> (message from Bill Broadley on Thu, 27 Jan 1994 21:17:05 -0500) Subject: Re: Arc 100 prices.. I say...looking at the prices: The motherboard looks great. let's go witht he 4600. However.....peripherals listed are WAY overpriced. I say that for those let's roll our own. Now for another question. SOFTWARE. I assume that someone out there has a prototype board. I hope so, because I don't know how to write OSes KsH From broadley@neurocog.lrdc.pitt.edu Thu Jan 27 17:49:07 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA23101; Thu, 27 Jan 1994 22:49:08 -0500 Message-Id: <199401280349.AA23101@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA08477; Thu, 27 Jan 94 22:49:07 -0500 From: Bill Broadley Subject: Re: ARC 100 Prices To: riscy@sunsite.unc.edu Date: Thu, 27 Jan 1994 22:49:07 -0500 (EST) In-Reply-To: <9401260818.AA12660@resi.waldorf-gmbh.de> from "Andreas Busse" at Jan 26, 94 09:18:46 am X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 804 > > Motherboard w/o RAM and CPU, DM 2000.-- / us$1185.-- > > incl. SCSI, serial/parallel, > > FDC, Ethernet, EISA Ctrl. etc. This is to much for me. Where does the cost come from? SCSI seems cheap, serial seems cheap. A rough break down would help play the price/performance gain. If enough people support it I wish you the best of luck, for me I'll either have to not go riscy or go for less ambitious motherboard. Would dropping the EISA/bus completely make a big difference? Switching to ISA? Are any other chipsets/cheaper in contention? The sgi reference designs? Changing the memory interface from 128 to 64? to 32? -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From drew@romeo.cs.Colorado.EDU Thu Jan 27 14:13:21 1994 Received: from romeo.cs.colorado.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA28842; Thu, 27 Jan 1994 23:14:57 -0500 Received: from localhost (localhost [127.0.0.1]) by romeo.cs.Colorado.EDU (8.6.4/8.6.4) with SMTP id VAA01486; Thu, 27 Jan 1994 21:13:23 -0700 Message-Id: <199401280413.VAA01486@romeo.cs.Colorado.EDU> X-Authentication-Warning: romeo.cs.Colorado.EDU: Host localhost didn't use HELO protocol To: andy@waldorf-gmbh.de Cc: riscy@sunsite.unc.edu Subject: Re: ARC 100 Prices In-Reply-To: Your message of "Thu, 27 Jan 1994 20:59:11 EST." <9401260815.AA12641@resi.waldorf-gmbh.de> Date: Thu, 27 Jan 1994 21:13:21 -0700 From: Drew Eckhardt -------- > I believe Drew offered to handle the linux scsi driver for the NCR53cf94 > (very nice chip from what I hear) chip if we decided to use it. I'm not > sure this offer still stands. Someone wrote that the Emulex FAS216 is compatible with the NCR 53cf94. Could be. Emulex makes clones of a few of the low-midrange NCR chips, as well as their own high-end SCSI chips. Unfortunately, when I asked for people to commit firmly to using one of the chips in the NCR53c90 family, I saw only one response and nothing out of Waldorf. With no consensus, I wasn't about to waste my time on a driver that wouldn't be used. I made the offer when I still had plenty of free time - enough time to wire wrap a prototype board, write the driver, and debug it, and wasn't working on any other interesting projects. Now, I'm working full time as a programmer, if I wasn't I'd still have enough consulting to keep me busy, I'm in the middle of snowboarding season (12" of fresh powder on Loveland Pass!), and am working on the NCR53c800/700 series driver under Linux. The marketoids bill the NCR53c810 as having an onboard "2 MIPS processor", it's busmastering, runs code out of main memory, takes about 600ns to start a SCSI command, has no limit on the number of discontiguous scatter/gather buffers, and can do just about everything without host processor intervention. And writing a SCSI driver that runs mostly on the SCSI chip, with a scheduler to handle multiple threads of execution (one per command), is much more fun than writing a SCSI driver for a simple chip that has nothing more than a simple onboard sequencer :-) Not to mention the fact that hardware was provided for the NCR53c810 driver, meaning a hot 486 PCI board (Real nice. 0ws cache, interleaved main memory so writes alternate between 0 and 1 wait states, allegedly transfers from main memory to PCI video boards at > 30M/sec) :-) From t146678@cc.tut.fi Fri Jan 28 10:04:25 1994 Received: from cc.tut.fi by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11689; Fri, 28 Jan 1994 01:06:11 -0500 Received: by cc.tut.fi id AA21020 (5.65b/IDA-1.4.3 for riscy@sunsite.unc.edu); Fri, 28 Jan 94 08:04:25 +0200 From: Tuominen Juha Message-Id: <9401280604.AA21020@cc.tut.fi> Subject: General info wanted To: riscy@sunsite.unc.edu Date: Fri, 28 Jan 1994 08:04:25 +0200 (GMT) X-Mailer: ELM [version 2.4 PL21] Content-Type: text Content-Length: 366 I've been very busy due to my work and hadn't had time to read the message flow on the riscy mailing list, so I'm just asking could someone send me (or maybe to this mailing list also) a brief description what's been done so far and what's going to happen in the near future regarind this project. -- Doctor, doctor! I suck. You've gotta help me. From andy@waldorf-gmbh.de Fri Jan 28 10:40:41 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA20468; Fri, 28 Jan 1994 03:25:53 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:15/EUnetD-2.4.1.e) via EUnet id JAA07922; Fri, 28 Jan 1994 09:26:49 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Fri, 28 Jan 94 09:40:41 +0100 Message-Id: <9401280840.AA16536@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: ARC 100 Prices Hi everybody, thanks for the votes. Although there weren't much of them it's enough for me to understand that the ARC design isn't of much interest. I DO understand that most or nearly all members of the list cannot afford $2000 or the like. Ok, that's your problem :-) My problem is that I can't offer the board for less money. Someone asked if skipping EISA would save money - yes it will, but it won't reduce the price by 50%. Another problem is that Waldorf has no time to do two re-designs, one for our own purposes, one for riscy. Even if we had, we would come into the same loop - you'll ask me why a dump ISA board costs us$xxx. You can't buy a board for $500 if the CPU plus bus controller chipset is $800. Prices will change, but this will take a while. Some time ago the group decided to wait for R4000 instead of using old-fashioned R30xx derivates. I did not mind. Now we have a R4000 design, surely not the best (fastest) one available, and perhaps not the cheapest one, but we can have it. Any alternative in sight ? NO. Correct me if I'm wrong, please. What do we do now ? Tell me. I can only say: Waldorf will buy this design, no matter how riscy decides to go on (or stop). Don't get me wrong: I would LIKE to offer a complete board incl. CPU/SCSI/VIDEO/SERIAL and so on for $1000. But I can't. However, whatever the group decides to do now, I'll support it if I can. Cheers, Andy From hodgen@mailhost.uni-koblenz.de Fri Jan 28 04:36:46 1994 Received: from mailhost.uni-koblenz.de by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA24057; Fri, 28 Jan 1994 04:36:46 -0500 Received: from ozzy.uni-koblenz.de by mailhost.uni-koblenz.de with smtp (Smail3.1.28.1) id m0pPp3a-0001jXC; Fri, 28 Jan 94 10:00 MET Message-Id: Received: by ozzy.uni-koblenz.de (4.1/KO-2.0) id AA22747; Fri, 28 Jan 94 10:02:02 +0100 From: Subject: Re: Who's there ??? To: riscy@sunsite.unc.edu Date: Fri, 28 Jan 94 10:02:01 MET In-Reply-To: <9401261623.AA13789@resi.waldorf-gmbh.de>; from "Andreas Busse" at Jan 27, 94 9:22 pm X-Mailer: ELM [version 2.3 PL11] > Now that I posted a rough price estimation I'd like > to hear what you all think about, more exactly: First, I think the prices, particularly for peripherals are too high. The board is expensive enough, then Video, CPU, memory etc. BTW, no-one has even said if normal PC SIMM will fit on the board. Does it need its own SIMMs? Not using PC memory will almost double the cost. > I need votes. > > If those two or three guys who sent comments are > the only ones on the list who are still interested, > I'd say: forget about riscy. I'm still commited to getting away from Intel computers but unless the price can be reduced I wont be able to aford it. Look, motherboard + CPU + Video + Ethernet + keyboard = DM 2825. 16 MB memory ~ DM 1000. So I'm looking at close to DM 4000 for a decent configuration. That's a *big* investment, especially if I can't use my PC memory (which I already have plenty of). OK, it's still less than a Pentium PCI board but by the end of the year (it'll take that long to port Linux won't it?) they'll be priced like 486s now. I don't want a Goddamn Intel again though!!! I really want this project to work. I know Andy has put in a lot of time and effort into this _but_ if we had a design (as origionally desired) with everything on the motherboard, the costs go down immensely. As Andy said just the EISA chipset is $300. The problem is, no manufacturer will produce a design for such a board. Either you hardware people out there manage to design one (I have no hardware education/experience) or we all just shut up and go buy Intel boards (which should please Bill Broadly, they're very cheap Bill). -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | Have you heard about the Uni Koblenz, | or Fight-o-net 2:245/5600.11 | riscy project? A GPL based Rheinau 1, | Voice: +49 261 9119-645 | design for an R4000 PC MB. 56075 Koblenz. | Fax: +49 261 9119-499 | MIPS, Power to the people! From andy@waldorf-gmbh.de Fri Jan 28 12:37:29 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AB26108; Fri, 28 Jan 1994 05:25:42 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:15/EUnetD-2.4.1.e) via EUnet id LAA13609; Fri, 28 Jan 1994 11:26:16 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Fri, 28 Jan 94 11:37:29 +0100 Message-Id: <9401281037.AA17011@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: Re: Who's there ??? > First, I think the prices, particularly for peripherals are too high. Agreed, but can't change it. > The board is expensive enough, then Video, CPU, memory etc. BTW, no-one > has even said if normal PC SIMM will fit on the board. Does it need > its own SIMMs? Not using PC memory will almost double the cost. It wants 36bit 60ns SIMMs, 70ns perhaps will do too. > OK, it's still less than a Pentium PCI board but by the end of the year > (it'll take that long to port Linux won't it?) they'll be priced like 486s > now. Again, I'm sure the prices for the R4000 stuff will fall too. > I don't want a Goddamn Intel again though!!! That's why we are talking about Mips, isn't it ? > I really want this project to work. I know Andy has put in a lot of time and > effort into this _but_ if we had a design (as origionally desired) with > everything on the motherboard, the costs go down immensely. Agreed, but the design we can get isn't that way. > As Andy said just the EISA chipset is $300. Wrong, I meant the NEC chip set which can't be skipped in this design. The EISA chipset can be skipped, but it would make a redesign necessary. > The problem is, no manufacturer will produce > a design for such a board. Either you hardware people out there manage to > design one (I have no hardware education/experience) or we all just shut up > and go buy Intel boards (which should please Bill Broadly, they're very cheap > Bill). Agreed. Well, if the ARC board is too expensive, why don't you get the documentation for the NEC chipset and do your own design ? It shouldn't be *that* complicated since the NEC chips do most, if not all of the really complicated things. If it works, Waldorf will perhaps produce it. Why not... But please don't expect a lot of support from me. I can't, for example, distribute circuit diagrams since Waldorf will *license* them from NEC. They aren't GPL'ed, you know ? I give you an idea what you *could* do: - Get the R4x00 spec sheets, from NEC, IDT, Toshiba or the like. - Get the NEC ARC chipset specs. You'll see, they do nearly all essential things. No need to think about complicated DRAM control with interleaving, burst mode and so on. Everything built-in. Should even be possible to use your old 9bit SIMMs (but you'll need lots of them). - Add an ISA chipset. So far I know, this is supported by the NEC chips too. - Leave everything else *away*, so that everyone can use his old stuff. - Watch out for someone who could do CAD stuff. - Ask me again if Waldorf can produce that board. We might be interested... Costs ? $500 for the CPU (will fall) $300 for the NEC chips (will fall) $100 for all the other glue (won't fall) $50 for a multilayer printed circuit board ---- $950 for a board comparable with all these ISA Intel Boards. Works, but doesn't do anything but consuming power :-) BUT WAY FASTER ! Then plug in your old video, scsi, serial, IDE and so on cards. Ready. Cheers, Andy From bergstro@src.honeywell.com Fri Jan 28 03:17:24 1994 Received: from src.honeywell.com (moon.src.honeywell.com) by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA19054; Fri, 28 Jan 1994 10:17:29 -0500 Return-Path: Received: from potatoe.src.honeywell.com by src.honeywell.com (4.1/smail2.6.3/SRCv0.25); Fri, 28 Jan 94 09:17:25 CST id AA09074 for riscy@sunsite.unc.edu at sunsite.unc.edu Posted-Date: Fri, 28 Jan 94 09:17:24 CST Received: by potatoe.src.honeywell.com (4.1/SMI-3.2) id AA28362; Fri, 28 Jan 94 09:17:24 CST Date: Fri, 28 Jan 94 09:17:24 CST From: bergstro@src.honeywell.com (Pete Bergstrom) Message-Id: <9401281517.AA28362@potatoe.src.honeywell.com> To: riscy@sunsite.unc.edu In-Reply-To: <199401280250.AA17113@SunSITE.Unc.EDU> (johnsonm@SunSITE.Unc.EDU) Subject: Reactions to price (was Re: Who's there ???) I'm afraid that the price is out of my range for a motherboard & cpu. I was expecting a total price of maybe $800 - at least I thought that was the general area when I first started listening to this list. Pete From amoss@cs.huji.ac.il Fri Jan 28 20:11:57 1994 Received: from cs.huji.ac.il by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA27154; Fri, 28 Jan 1994 11:12:26 -0500 Received: from picton.cs.huji.ac.il by cs.huji.ac.il with SMTP id AA00949 (5.65c/HUJI 4.153 for ); Fri, 28 Jan 1994 18:12:00 +0200 Received: from localhost by picton.cs.huji.ac.il with SMTP id AA01850 (5.65c/HUJI 4.1 for riscy@sunsite.unc.edu); Fri, 28 Jan 1994 18:11:59 +0200 Message-Id: <199401281611.AA01850@picton.cs.huji.ac.il> To: riscy@sunsite.unc.edu Subject: Re: ARC 100 Prices In-Reply-To: Your message of Fri, 28 Jan 1994 03:39:29 -0500 . <9401280840.AA16536@resi.waldorf-gmbh.de> From: Amos Shapira Date: Fri, 28 Jan 1994 18:11:57 +0200 Sender: amoss@CS.HUJI.AC.IL andy@waldorf-gmbh.de writes: |and perhaps not the cheapest one, but we can have it. |Any alternative in sight ? NO. Correct me if I'm wrong, please. | |What do we do now ? Tell me. I can only say: Waldorf will |buy this design, no matter how riscy decides to go on (or stop). | |Don't get me wrong: I would LIKE to offer a complete board |incl. CPU/SCSI/VIDEO/SERIAL and so on for $1000. But I can't. | |However, whatever the group decides to do now, I'll support |it if I can. | |Cheers, |Andy Thanks very much for the offer, and I suspect I won't be the only one to appreciate your efforts for having a cheap riscy board, but as the others said this is a bit more than we expected. Unless a cheaper offer comes up I won't be able (and won't have a justification) to buy that board. Good luck, --Amos --Amos Shapira (Jumper Extraordinaire) | "War does not determine who is right, C.S. System Group, Hebrew University, | but who is left" Jerusalem 91904, ISRAEL | amoss@cs.huji.ac.il | -- Anonymous? From broadley@neurocog.lrdc.pitt.edu Fri Jan 28 17:17:44 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA15633; Fri, 28 Jan 1994 22:17:57 -0500 Message-Id: <199401290317.AA15633@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA15976; Fri, 28 Jan 94 22:17:44 -0500 From: Bill Broadley Subject: Wild idea. To: riscy@sunsite.unc.edu Date: Fri, 28 Jan 1994 22:17:44 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1601 Waldorf has it's own agenda which seems to include MIPS. Mips seems that the design complexity if forcing the use of some fairly complex/expensive chips. ($2711 for 0 ram, no L2 cache, 4600 with video, audio, and ethernet) Are any of the hardware designers out there interested in doing our own design with an eye towards price performance? Please email me and/or the list if you feel you could help design a motherboard. I've talked to a few dec engineers and they seem to think that a very cheap motherboard using the ALPHA 21066 should be possible. They sent me I believe almost everything I need to make a motherboard. (Not that I have any skill at such things.) Big features of the 21066: Embedded graphics controller (on chip) L2 cache support (directly connects to sram.) PCI I/O controller on chip 8k/8k L1 cache 3.3 V chip (cooler) 70 Specint 105 Specfp (better then any 100 Mhz mips) This has several big advantages of the mips design: Cheaper, faster, PCI. Of course the disadvantages: No cheap production at Waldorf Software will be pretty tricky getting started/booted. Seems like you basically hook up the PCI bus, vram, sram, and dram to the pins, and you have a motherboard. So any hardware designers interested in donating time? I have parts list, parts descriptions, schematics, and a product brief on the ALPHA evaluation board. It's a wild idea just thought I'd thow it out to the list. -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From antsu@sandra.pp.fi Sat Jan 29 09:04:04 1994 Received: from eunet.fi by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11225; Sat, 29 Jan 1994 09:04:04 -0500 Received: from sandra.pp.fi by eunet.fi with UUCP id AA18558 (5.65c+l/IDA-1.4.4 for sunsite.unc.edu!riscy); Sat, 29 Jan 1994 16:03:08 +0200 Received: by sandra.pp.fi (/\==/\ Smail3.1.24.1 #24.3) id ; Sat, 29 Jan 94 15:59 EET Message-Id: Date: Sat, 29 Jan 94 15:59 EET From: antsu@sandra.pp.fi (Antti-Pekka Virtanen) To: broadley@neurocog.lrdc.pitt.edu, riscy@sunsite.unc.edu Subject: Re: Wild idea (really ?) Hi! Regarding the DEC Alpha... > Big features of the 21066: > Embedded graphics controller (on chip) > L2 cache support (directly connects to sram.) > PCI I/O controller on chip > 8k/8k L1 cache > 3.3 V chip (cooler) > 70 Specint 105 Specfp (better then any 100 Mhz mips) Yes, this sounds a lot more interesting than the MIPS families. We have been thinking about using the DEC evaluation board design. (Which contains 32MB (?) memory, NCR SCSI, PCI and ISA, correct me if I'm wrong. This is taken from a short product brief. Don't know any details.) The main difference here is that we have thought to plug in the 21068 which is about 1/2 price 1/2 performance [1] , but you could allways plug in the 21066 or any newer one with same pinout... (Initially we have meant this to be an embedded controller for our internal product purposes... that explains the 21068 :) [1] 21068 at 66MHz 30 SPECint92; graphics accelerator delivers 30M Winmarks with VRAM. > Seems like you basically hook up the PCI bus, vram, sram, and dram to the > pins, and you have a motherboard. More or less... > So any hardware designers interested in donating time? I have parts list, > parts descriptions, schematics, and a product brief on the ALPHA > evaluation board. If you can, it would be most wellcome if you can send me copies [2]. I have been trying to get same things from DEC, but the local DEC seems a bit confused regarding chip level designs. They would sell me a complete workstation with a pleasure though... [2] The address can be found below. The OS thing is a bit more complicated. I think the easiest way to do OS dev. is to get Windows NT or OSF/1 and then use the DEC provided SW tools (21A03-11 21066/68 Design kit, about US * $50 * or so...) maybe, just maybe. Any OS gurus listening? Our initial idea was to run this just without any OS as a dedicated microcontroller board , just a hell lot faster than some good old HC11/16/XX :) There seems to be (or anytime soon) GCC and tools for Alpha... The 21068 samples in qty 1 cost about US $450 but I suppose the price to go down very rapidly. This doesn't mean that we will have a design ready any time soon, but if there is some interest it might be of help. Also, I don't want to start any flame war between different architectures; mips will be of interest to Andy, and alpha to me... In any case, just some thoughts... Regards, Antti-Pekka-- Antti-Pekka Virtanen E-mail: antsu@sandra.pp.fi Arholankuja 3 a 21 Packet: OH1YF@OH1RBU.TKU.FIN.EU FIN-21100 Naantali Finland Tel. : Intl.+358 21 751888 At the University of Turku : antsu@utu.fi BSD/386 Software Archives : antsu@nic.funet.fi Aumec Systems Oy "The company" Fiskarsinkatu 7D FIN-20750 Turku Finland "If the human brain were so simple that we could understand it, we would be so simple we couldn't." From johnsonm Sat Jan 29 06:17:34 1994 Received: from localhost.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA16592; Sat, 29 Jan 1994 11:17:35 -0500 Message-Id: <199401291617.AA16592@SunSITE.Unc.EDU> To: Multiple recipients of list Reply-To: johnsonm Subject: Re: Wild idea. In-Reply-To: Your message of "Fri, 28 Jan 94 22:20:39 EST." <199401290317.AA15633@SunSITE.Unc.EDU> Date: Sat, 29 Jan 94 11:17:34 -0500 From: ""Michael K. Johnson"" Bill Broadley writes: >I've talked to a few dec engineers and they seem to think that a very >cheap motherboard using the ALPHA 21066 should be possible. They sent >me I believe almost everything I need to make a motherboard. (Not >that I have any skill at such things.) > >Big features of the 21066: > Embedded graphics controller (on chip) > L2 cache support (directly connects to sram.) > PCI I/O controller on chip > 8k/8k L1 cache > 3.3 V chip (cooler) > 70 Specint 105 Specfp (better then any 100 Mhz mips) > >This has several big advantages of the mips design: > Cheaper, faster, PCI. > >Of course the disadvantages: > No cheap production at Waldorf > Software will be pretty tricky getting started/booted. I think this is worth looking into further. I think we should check the prices of available motherboards, though, while we consider one of our own, as it may be just as cheap overall to get an existing board. DEC intends for this chip to be in strong competition with the pentium, so prices for the whole motherboard could be pretty low, even for a pre-assembled one. I would hope that we could get 21066 motherboards for < $1000 without ram, but with CPU, at some point. If not, DEC is going to have trouble competing with intel, unfortunately. Then we just have to port Linux to it... I wonder if DEC would be interested in helping us with that? There was someone from DEC who posted recently saying that if there was enough demand, he'd talk to his management about a Linux port. Maybe if there isn't enough demand for that, but some demand, they'd help others do a port. By writing PALcode that approximates the MM of the 386, a porting effort could be made a lot easier. We would also be able to use Drew's PCI scsi driver, with few or no changes, which would be a big win. A RISC chip with 386-like MM and all the major devices already supported in Linux/86 would make for relatively easy porting, other than places where going from 32 to 64 bits causes problems, which shouldn't be too many places. Hopefully, GCC will have improved code generation for the ALPHA by that time... I vote that we look into this, at least. michaelkjohnson From tthorn%hof@hof.daimi.aau.dk Sat Jan 29 14:48:49 1994 Received: from hof (hof.daimi.aau.dk) by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA04836; Sat, 29 Jan 1994 14:48:49 -0500 Received: by hof (Linux Smail3.1.28.1 #14) id m0pQLiQ-000IBVC; Sat, 29 Jan 94 20:52 MET Message-Id: Date: Sat, 29 Jan 94 20:52 MET From: tthorn%hof@hof.daimi.aau.dk (Tommy Thorn) To: Multiple recipients of list Subject: Re: Who's there ??? In-Reply-To: <9401261623.AA13789@resi.waldorf-gmbh.de> References: <9401261623.AA13789@resi.waldorf-gmbh.de> Reply-To: Tommy.Thorn@daimi.aau.dk Andreas Busse writes: > Hi everybody ! > > Now that I posted a rough price estimation I'd like > to hear what you all think about, more exactly: > > I need votes. > > If those two or three guys who sent comments are > the only ones on the list who are still interested, > I'd say: forget about riscy. I simply cannot raise more than $1,000, so for now I'll forget about riscy. I'm still interested in anything non-intel RISC based EISA/ISA/PCI motherboard. If anyone knows alternatives, I'm sure this is still an excelent place to discuss them. Sorry, /Tommy From broadley@neurocog.lrdc.pitt.edu Sat Jan 29 17:47:09 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA21548; Sat, 29 Jan 1994 22:47:10 -0500 Message-Id: <199401300347.AA21548@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA21557; Sat, 29 Jan 94 22:47:09 -0500 From: Bill Broadley Subject: 21066 board (was Wild idea) To: riscy@sunsite.unc.edu Date: Sat, 29 Jan 1994 22:47:09 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 1430 I received a fair amount of support email about about the 2106[68] board. Much more supportive then the last time I mailed about it. As a rough guess I'd say that an under $1k board should be possible. Most importantly there support from people who can help with the hardware end. The following 2 people asked I mail out a copy of our evaluation board Antti-Pekka Virtanen Juha Tuominen Arholankuja 3 a 21 Pikkupiiankatu 3 A 3 FIN-21100 Naantali Finland SF-33580 TAMPERE FINLAND I'm not familiar with mailing overseas so the labels will look exactly like the above. Please email me if you think they might be wrong. Anybody else? It's an inch or 2 thick, I'll visit the copy center monday and the post office and mail them off. They will only be useful to the hardware hackers (i.e. computer printed parts lists, cryptic (to me) descriptions, and a schematic thats many pages long.) I'll try to post a parts list on monday so we can get a better idea about price. The current evalution board uses 32 bit simms, L2 cache, video, audio, pci, isa, keyboard, serial, parallel, mouse. Basically everything we could want. I'm not sure we can afford all the above features but we can find out how much they will cost. -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From broadley@neurocog.lrdc.pitt.edu Sat Jan 29 17:53:00 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA22020; Sat, 29 Jan 1994 22:53:02 -0500 Message-Id: <199401300353.AA22020@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA21591; Sat, 29 Jan 94 22:53:00 -0500 From: Bill Broadley Subject: 21066 prices To: riscy@sunsite.unc.edu Date: Sat, 29 Jan 1994 22:53:00 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 771 The only solid info I have is: DECchip 21066-- $385 per unit in quantities of 5,000 DECchip 21068-- $221 per unit in quantities of 5,000. But it's from September. I've heard they are cheaper, and shipping in volume. They have already released a 200 Mhz version of the chip, and I believe the price will keep dropping as others in the market do (i.e. pentium). The Chip supports 10 loads on a 33 Mhz pci bus, some video acceleration/vram support, and L2 support. The bus speed is 33 Mhz which should mean that the vabrication of the motherboard based on the schematic I have should not be to tricky. -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From amoss@cs.huji.ac.il Sun Jan 30 15:22:00 1994 Received: from cs.huji.ac.il by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA19524; Sun, 30 Jan 1994 06:22:13 -0500 Received: from picton.cs.huji.ac.il by cs.huji.ac.il with SMTP id AA12126 (5.65c/HUJI 4.153 for ); Sun, 30 Jan 1994 13:22:03 +0200 Received: from localhost by picton.cs.huji.ac.il with SMTP id AA04341 (5.65c/HUJI 4.1 for riscy@sunsite.unc.edu); Sun, 30 Jan 1994 From antsu@sandra.pp.fi Sun Jan 30 09:50:22 1994 Received: from eunet.fi by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA26424; Sun, 30 Jan 1994 09:50:22 -0500 Received: from sandra.pp.fi by eunet.fi with UUCP id AA25398 (5.65c+l/IDA-1.4.4 for sunsite.unc.edu!riscy); Sun, 30 Jan 1994 16:50:06 +0200 Received: by sandra.pp.fi (/\==/\ Smail3.1.24.1 #24.3) id ; Sun, 30 Jan 94 16:40 EET Message-Id: Date: Sun, 30 Jan 94 16:40 EET From: antsu@sandra.pp.fi (Antti-Pekka Virtanen) To: riscy@sunsite.unc.edu Subject: Re: 21066 price In-Reply-To: Mail from 'Bill Broadley ' dated: Sat, 29 Jan 1994 22:54:05 -0500 > The only solid info I have is: > DECchip 21066-- $385 per unit in quantities of 5,000 > DECchip 21068-- $221 per unit in quantities of 5,000. Great! So they have come down from what I had from June... Greetings, Antti-Pekka-- Antti-Pekka Virtanen E-mail: antsu@sandra.pp.fi Arholankuja 3 a 21 Packet: OH1YF@OH1RBU.TKU.FIN.EU FIN-21100 Naantali Finland Tel. : Intl.+358 21 751888 At the University of Turku : antsu@utu.fi BSD/386 Software Archives : antsu@nic.funet.fi "If the human brain were so simple that we could understand it, we would be so simple we couldn't." From antsu@sandra.pp.fi Sun Jan 30 09:51:09 1994 Received: from eunet.fi by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA26469; Sun, 30 Jan 1994 09:51:09 -0500 Received: from sandra.pp.fi by eunet.fi with UUCP id AA25402 (5.65c+l/IDA-1.4.4 for sunsite.unc.edu!riscy); Sun, 30 Jan 1994 16:50:11 +0200 Received: by sandra.pp.fi (/\==/\ Smail3.1.24.1 #24.3) id ; Sun, 30 Jan 94 16:46 EET Message-Id: Date: Sun, 30 Jan 94 16:46 EET From: antsu@sandra.pp.fi (Antti-Pekka Virtanen) To: broadley@neurocog.lrdc.pitt.edu, riscy@sunsite.unc.edu Subject: Re: 21066 board (was Wild idea) In-Reply-To: Mail from 'Bill Broadley ' dated: Sat, 29 Jan 1994 22:48:37 -0500 > I received a fair amount of support email about about the 2106[68] board. > Much more supportive then the last time I mailed about it. As a rough > guess I'd say that an under $1k board should be possible. Most importantly > there support from people who can help with the hardware end. And, it would be of much help if there is a pcb design in electronic form that we could use directly... (I think the DEC one is available in the software kit that costs the $50 or so, at least this is what I was told by the local DEC) > The current evalution board uses 32 bit simms, L2 cache, video, audio, pci, > isa, keyboard, serial, parallel, mouse. Basically everything we could > want. I'm not sure we can afford all the above features but we can > find out how much they will cost. Sounds like a good design to me :) Greetings, Antti-Pekka -- Antti-Pekka Virtanen E-mail: antsu@sandra.pp.fi Arholankuja 3 a 21 Packet: OH1YF@OH1RBU.TKU.FIN.EU FIN-21100 Naantali Finland Tel. : Intl.+358 21 751888 At the University of Turku : antsu@utu.fi BSD/386 Software Archives : antsu@nic.funet.fi "If the human brain were so simple that we could understand it, we would be so simple we couldn't." From hodgen@mailhost.uni-koblenz.de Sun Jan 30 16:07:59 1994 Received: from mailhost.uni-koblenz.de by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA26742; Sun, 30 Jan 1994 16:07:59 -0500 Received: from ozzy.uni-koblenz.de by mailhost.uni-koblenz.de with smtp (Smail3.1.28.1) id m0pQjN2-0001lUC; Sun, 30 Jan 94 22:07 MET Message-Id: Received: by ozzy.uni-koblenz.de (4.1/KO-2.0) id AA01944; Sun, 30 Jan 94 22:09:50 +0100 From: Subject: Re: 21066 board (was Wild idea) To: riscy@sunsite.unc.edu Date: Sun, 30 Jan 94 22:09:49 MET In-Reply-To: <199401300347.AA21548@SunSITE.Unc.EDU>; from "Bill Broadley" at Jan 29, 94 10:48 pm X-Mailer: ELM [version 2.3 PL11] > The current evalution board uses 32 bit simms, L2 cache, video, audio, pci, > isa, keyboard, serial, parallel, mouse. Basically everything we could > want. I'm not sure we can afford all the above features but we can > find out how much they will cost. If you make L2 cache an option and do away with PCI (If everything is on the motherboard, who needs it?) you'd have a very good starting point. Good to see that there is life after MIPS. If this gets going I'll have to change my signature :) Count me in. -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | Have you heard about the Uni Koblenz, | or Fight-o-net 2:245/5618.50 | riscy project? A GPL based Rheinau 1, | Voice: +49 261 9119-645 | design for an R4000 PC MB. 56075 Koblenz. | Fax: +49 261 9119-499 | MIPS, Power to the people! From johnsonm Sun Jan 30 13:38:37 1994 Received: from localhost.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA09786; Sun, 30 Jan 1994 18:38:40 -0500 Message-Id: <199401302338.AA09786@SunSITE.Unc.EDU> To: Multiple recipients of list Reply-To: johnsonm Subject: Re: 21066 board (was Wild idea) In-Reply-To: Your message of "Sun, 30 Jan 94 16:09:51 EST." Date: Sun, 30 Jan 94 18:38:37 -0500 From: ""Michael K. Johnson"" hodgen@mailhost.uni-koblenz.de writes: >If you make L2 cache an option and do away with PCI (If everything is on the >motherboard, who needs it?) you'd have a very good starting point. Good to >see that there is life after MIPS. If this gets going I'll have to change >my signature :) What do you mean by "everything is on the motherboard"? Whose definition of everything? What if person X wants to add a device? I'd be willing to drop ISA support, myself, but there's no point in dropping PCI support, since it is supported by the CPU. Also, with the ALPHA chip, I think that not having an L2 cache would be a real waste of processor. I don't know of any reason that it couldn't be made an option, I'm just doubting your sanity for not wanting it ;-) Remember, this CPU (at least one of the two we are considering) is ~twice as fast as the MIPS, and can use a lot of fast access to memory... michaelkjohnson From pham@pX4.stfx.ca Sun Jan 30 15:54:58 1994 Received: from pX4.StFX.CA by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA11175; Sun, 30 Jan 1994 18:55:06 -0500 Received: by pX4.stfx.ca (5.65/DEC-Ultrix/4.3) id AA06507; Sun, 30 Jan 1994 19:54:59 -0400 Message-Id: <9401302354.AA06507@pX4.stfx.ca> From: Hai Pham Date: Sun, 30 Jan 1994 19:54:58 -0400 To: johnsonm@SunSITE.Unc.EDU, Multiple recipients of list Subject: Re: 21066 board (was Wild idea) On Jan 30, 18:39, ""Michael K. Johnson"" wrote: Subject: Re: 21066 board (was Wild idea) > >Also, with the ALPHA chip, I think that not having an L2 cache would >be a real waste of processor. I don't know of any reason that it >couldn't be made an option, I'm just doubting your sanity for not >wanting it ;-) Remember, this CPU (at least one of the two we are >considering) is ~twice as fast as the MIPS, and can use a lot of fast >access to memory... Besides, doesn't it have the cache controller & some sort of video accelerator built in or something? Cache chips aren't very expensive once the cache controller has been taken care off. hai -- +--------------------------------+----------------------------------------+ | Hai Pham | email: pham@px1.stfx.ca (preferred)| | Department of Physics | pham@helios.stfx.ca (home) | | Saint Francis Xavier University| x92haa@essex.stfx.ca (evil VMS) | | Nova Scotia, Canada. | LINUX + 386sx = Useful boat anchor. | +--------------------------------+----------------------------------------+ From johnsonm Sun Jan 30 14:06:30 1994 Received: from localhost.oit.unc.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA12582; Sun, 30 Jan 1994 19:06:31 -0500 Message-Id: <199401310006.AA12582@SunSITE.Unc.EDU> To: Multiple recipients of list Reply-To: johnsonm Subject: Re: 21066 board (was Wild idea) In-Reply-To: Your message of "Sun, 30 Jan 94 19:54:58 -0400." <9401302354.AA06507@pX4.stfx.ca> Date: Sun, 30 Jan 94 19:06:30 -0500 From: ""Michael K. Johnson"" Hai Pham writes: >Besides, doesn't it have the cache controller & some sort of video >accelerator built in or something? Cache chips aren't very expensive once >the cache controller has been taken care off. Right. That was the point that I was trying to make, and failed utterly to mention in the process. Thanks. Yes, how much can 256 or 512 K of SRAM cost? And yes, video is built in. I'd guess that there's not a lot of acceleration, except maybe things like bitblt, knowing DEC's stance on acceleration. Essentially, the cpu is a better accelerator for most things than an accelerated card, given sufficient CPU/video memory bandwidth. However, I haven't seen any specs. I *would* expect to be able to do decent opaque moves with this, though. I'd expect to pay $50-100 max for cache ram and $100-200 max for VRAM. Don't know if the design requires VRAM or DRAM, but VRAM gives much better refresh rates and greater bandwidth, and is to be highly prefered. Video, PCI, and L2 cache controllers are all built in. Sounds good to me... michaelkjohnson From broadley@neurocog.lrdc.pitt.edu Sun Jan 30 15:02:36 1994 Received: from neurocog.lrdc.pitt.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA16652; Sun, 30 Jan 1994 20:02:38 -0500 Message-Id: <199401310102.AA16652@SunSITE.Unc.EDU> Received: by neurocog.lrdc.pitt.edu (1.37.109.4/16.2) id AA02065; Sun, 30 Jan 94 20:02:36 -0500 From: Bill Broadley Subject: PCI, L2, and ISA To: riscy@sunsite.unc.edu Date: Sun, 30 Jan 1994 20:02:36 -0500 (EST) X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 2285 I responded via private email, I'll repost it. My justification for an L2 cache: No L2 cache bothered me about the mips design. It reduced the 60 specint92 60 specfp92 to about 35/35. About equivalent to reducing a 586 class cpu to 486 class (int wise). The mips 4000 runs at 100 Mhz internally and 50 Mhz externally, and gets almost twice the speed from the L2 cache. The alpha 21066 runs at 166 Mhz internal and 33 Mhz external I'd expect it to be even more dependent on L2 cache. The smallest L2 cache dec ships is the 256k (even on the 100 Mhz alpha). Given the Alpha chip supports a L2, and cache ram chips are reasonably cheap I'd think it would be a big loss to not have an L2 cache. Typically when buying a pc mothebroard 64k to 256k cache upgrades are about $30 or so (right?). I'd guess we could get 512k L1 for well under $100 since 33 Mhz doesn't need a particularly fast L2 cache. If it doubles the speed of a $300-$500 cpu it's well worth it. My justification for PCI: As for pci it would be a shame to throw away a state of the art bus just to save the price of the connectors/board space since it's supported by the alpha chip. PCI allows 10 "loads" on the bus. Each connector, and each chip on the bus is a load. This translates to 3 or 4 PCI slots maximum. So for the price of the connectors, and the increased functionality of a state of the art bus makes it a good investment, and a shame to leave out. My justifcation to NOT have ISA: ISA will cost the price of the connectors, board space, an ISA<->PCI bridge chip, and a load on the PCI bus. The benefits are adding any of the millions of the ISA cards, but this isn't as good as it sounds. ISA is as ugly bus you have to worry about 16 MB addressing limit, irq and I/O conflicts. We can't use any on board BIOS, we wont have drivers for any ISA cards, and we have limited software development resources to write them. If the design is to expensive I'd vote for the ISA to be the first to go. Real workstation do fine without it, and it's the last remnant of the UGLY intel architecture to haunt the riscy design. -- Bill Broadley@{neurocog,schneider3,lrdc5}.lrdc.pitt.edu (in order of preference) Linux is great. Bike to live, live to bike. PGP-ok From hodgen@mailhost.uni-koblenz.de Mon Jan 31 05:00:19 1994 Received: from mailhost.uni-koblenz.de by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA24623; Mon, 31 Jan 1994 05:00:19 -0500 Received: from ozzy.uni-koblenz.de by mailhost.uni-koblenz.de with smtp (Smail3.1.28.1) id m0pQvQT-0001kQC; Mon, 31 Jan 94 11:00 MET Message-Id: Received: by ozzy.uni-koblenz.de (4.1/KO-2.0) id AA04004; Mon, 31 Jan 94 11:02:10 +0100 From: Subject: Re: 21066 board (was Wild idea) To: riscy@sunsite.unc.edu Date: Mon, 31 Jan 94 11:02:09 MET In-Reply-To: <199401302338.AA09786@SunSITE.Unc.EDU>; from "Michael K. Johnson" at Jan 30, 94 6:39 pm X-Mailer: ELM [version 2.3 PL11] > >If you make L2 cache an option and do away with PCI (If everything is on the > >motherboard, who needs it?) you'd have a very good starting point. Good to > >see that there is life after MIPS. If this gets going I'll have to change > >my signature :) > > What do you mean by "everything is on the motherboard"? Whose > definition of everything? What if person X wants to add a device? > I'd be willing to drop ISA support, myself, but there's no point in > dropping PCI support, since it is supported by the CPU. So long as decent audio is on the board and ethernet PCI cards aren't incredibly expensive, go ahead, kill the ISA. If person X wants to add card Y, where is he going to put it? By everything I mean the basics that everyone needs. Serial, parallel, video, SCSI. There is going to be a lot of discussion about _what_ video and if SCSI is needed (there were the last time). There will be lots of shouts for Ethernet on the board (I need it for one) or on a cheap enough PCI card. Audio tastes vary wildly too. Some only want 8kHz bleepers, others (me included) want 44.1kHz multi channel boards. I think we need to define the minimum configuration board that *everyone* (80% or so) wants to see. And the rest will have to go on cards. Otherwise there will be continual arguments "drop , I don't want it and it makes it too expensive". > Also, with the ALPHA chip, I think that not having an L2 cache would > be a real waste of processor. I don't know of any reason that it > couldn't be made an option, I'm just doubting your sanity for not > wanting it ;-) Remember, this CPU (at least one of the two we are > considering) is ~twice as fast as the MIPS, and can use a lot of fast > access to memory... Thats one hell of an assumption pal. Look at what I wrote and read it this time. I certainly wouldn't want an uncached board with the processor continually waiting for memory. But some will balk at the cost of cache chips (I'm assuming a decent sized/fast cache. ~0.25MB to 1MB) with board. They'll want to buy the motherboard without cache and bung in the chips as they can afford to. Amazing concept huh? Thinking about what someone else may want ;-) -- Wayne Hodgen | hodgen@informatik.uni-koblenz.de | Have you heard about the Uni Koblenz, | or Fight-o-net 2:245/5618.50 | riscy project? A GPL based Rheinau 1, | Voice: +49 261 9119-645 | design for an R4000 PC MB. 56075 Koblenz. | Fax: +49 261 9119-499 | MIPS, Power to the people! From andy@waldorf-gmbh.de Mon Jan 31 12:31:02 1994 Received: from mail.Germany.EU.net by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA25746; Mon, 31 Jan 1994 05:29:47 -0500 Received: by mail.Germany.EU.net with ESMTP (8.6.5:15/EUnetD-2.4.1.e) via EUnet id LAA27571; Mon, 31 Jan 1994 11:30:46 +0100 Received: from resi.waldorf-gmbh.de From: Andreas Busse Date: Mon, 31 Jan 94 11:31:02 +0100 Message-Id: <9401311031.AA21489@resi.waldorf-gmbh.de> Received: by resi.waldorf-gmbh.de (5.61/GEN-1.0.7) To: riscy@sunsite.unc.edu Subject: new R4x000 design Hi everybody, I just want to let you know that I found another (probably too expensive too :-) ) board design. Waldorf wouldn't need to produce it since you can already buy it. Call Deskstation and ask for the "Tyne Series RISC PCs". The board works with R4200s, R4400s and R4600s up to 200 Mhz, supports up to 2 MB L2 cache controlled by a special chip set, has a PCI bus plus 2 VESA and 4 ISA slots. Main Memory is up to 256 MB using standard SIMMs. A HAL and a ARC BIOS Prom is shipped together with the board. Quoted from Deskstation's European OEM Price List: R4600-100MHz 0K L2 cache: $1760.-- R4600-133MHz 512K L2 cache: $1910.-- R4600-133MHz 2M L2 cache: $2530.-- The boards are the same, ie. you can later plug in the L2 cache and move some jumpers to activate it. Don't tell me the board is too expensive. Cheers, Andy From af4@ukc.ac.uk Mon Jan 31 10:53:56 1994 Received: from mercury.ukc.ac.uk by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA27976; Mon, 31 Jan 1994 06:28:50 -0500 Message-Id: <199401311128.AA27976@SunSITE.Unc.EDU> Received: from eagle by mercury.ukc.ac.uk with UKC POP3+ id aa13417; 31 Jan 94 10:53 GMT Date: Mon, 31 Jan 94 10:53:56 GMT From: Cover Thief To: riscy@sunsite.unc.edu Subject: Re: 21066 board (was Wild idea) hodgen@mailhost.uni-koblenz.de writes: >So long as decent audio is on the board and ethernet PCI cards aren't >incredibly expensive, go ahead, kill the ISA. If person X wants to add card >Y, where is he going to put it? By everything I mean the basics that >everyone needs. Serial, parallel, video, SCSI. There is going to be >a lot of discussion about _what_ video and if SCSI is needed (there were >the last time). There will be lots of shouts for Ethernet on the board (I Well video is handled by the CPU so the only things to argue about here are the amount of VRAM and choice of RAMDAC. As for SCSI, Drew Eckhart is in favour of a rather fast NCR PCI SCSI controller for which he has written a driver. I see no reason for not going along with this unless the NCR chip is particularly expensive. >need it for one) or on a cheap enough PCI card. Audio tastes vary wildly too. >Some only want 8kHz bleepers, others (me included) want 44.1kHz multi channel >boards. I think we need to define the minimum configuration board that >*everyone* (80% or so) wants to see. And the rest will have to go on cards. Here we go again, design by committee. Wouldn't It Be Nice IF... >Otherwise there will be continual arguments "drop , I don't want it and >it makes it too expensive". True, but it's an all or nothing situation. If you want all the features and performance of a workstation then you're very unlikely to end up with a price tag of less than $1000. Something has to go. Remember that at one stage we were talking about designing a RISC based accelerator to plug into a PC ISA slot - no features, just CPU and memory. >Thats one hell of an assumption pal. Look at what I wrote and read it this >time. I certainly wouldn't want an uncached board with the processor >continually waiting for memory. But some will balk at the cost of cache chips >(I'm assuming a decent sized/fast cache. ~0.25MB to 1MB) with board. They'll >want to buy the motherboard without cache and bung in the chips as they can >afford to. Amazing concept huh? Thinking about what someone else may want ;-) OK, let's go with your 80% straw poll. Anyone not wanting second level cache? I thought not. Cache memories are reasonably cheap so penny pinching in this area is not a good idea. Also, designing the board to allow the cache (or any other feature for that matter) to be optional will probably require a modest amount of extra hardware. Allowing such options will require that parts of the board are socketed - increasing production cost and reducing reliability. Andy. --- "I spend in her hand, and spew in her lap;" --- J. Wilmot From drew@kinglear.cs.Colorado.EDU Mon Jan 31 11:52:24 1994 Received: from kinglear.cs.colorado.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA17646; Mon, 31 Jan 1994 20:52:43 -0500 Received: from localhost (localhost [127.0.0.1]) by kinglear.cs.Colorado.EDU (8.6.4/8.6.4) with SMTP id SAA13700; Mon, 31 Jan 1994 18:52:26 -0700 Message-Id: <199402010152.SAA13700@kinglear.cs.Colorado.EDU> X-Authentication-Warning: kinglear.cs.Colorado.EDU: Host localhost didn't use HELO protocol To: af4@ukc.ac.uk, riscy@sunsite.unc.edu Subject: Re: 21066 board (was Wild idea) In-Reply-To: Your message of "Mon, 31 Jan 1994 06:55:42 EST." <199401311128.AA27976@SunSITE.Unc.EDU> Date: Mon, 31 Jan 1994 18:52:24 -0700 From: Drew Eckhardt -------- Well video is handled by the CPU so the only things to argue about here are the amount of VRAM and choice of RAMDAC. As for SCSI, Drew Eckhart is in favour of a rather fast NCR PCI SCSI controller for which he has written a driver. A more accurate assessment may be "my driver is under development" _ although the code was essentially done before I got my sample hardware, now that I have the sample hardware and before the first public release, I'm doing some radical rearrangements on the guts so that it's cleaner and avoids some nasty race conditions. The "rather fast" bit is a bit of an understatement - the time needed to start a SCSI command is measured in nanoseconds rather than milliseconds as with other SCSI bus masters, there's no limit on the number of discontiguous scatter/gather buffers, with a 40Mhz clock you can do 10M/sec sync, 6M/sec async, etc. The chip's also very robust, featuring some amount of onchip filtering and active negation so you shouldn't have the cable problems that plague the newer Adaptec boards. I see no reason for not going along with this unless the NCR chip is particularly expensive. It's a nice chip, fast, elegant, requires no glue to plug it into the PCI bus. Cost should be in the $50-$60 range from the rough pricing estimates I got some time ago, I'm not sure where the price is now. -------- From drew@kinglear.cs.Colorado.EDU Mon Jan 31 12:27:33 1994 Received: from kinglear.cs.colorado.edu by SunSITE.Unc.EDU (5.65c+IDA/FvK-1.06) with SMTP id AA21428; Mon, 31 Jan 1994 21:27:40 -0500 Received: from localhost (localhost [127.0.0.1]) by kinglear.cs.Colorado.EDU (8.6.4/8.6.4) with SMTP id TAA14970; Mon, 31 Jan 1994 19:27:38 -0700 Message-Id: <199402010227.TAA14970@kinglear.cs.Colorado.EDU> X-Authentication-Warning: kinglear.cs.Colorado.EDU: Host localhost didn't use HELO protocol To: johnsonm@sunsite.unc.edu, riscy@sunsite.unc.edu Subject: Re: Wild idea. In-Reply-To: Your message of "Sat, 29 Jan 1994 11:18:18 EST." <199401291617.AA16592@SunSITE.Unc.EDU> Date: Mon, 31 Jan 1994 19:27:33 -0700 From: Drew Eckhardt By writing PALcode that approximates the MM of the 386, a porting effort could be made a lot easier. We would also be able to use Drew's PCI scsi driver, with few or no changes, which would be a big win. A RISC chip with 386-like MM and all the major devices already supported in Linux/86 would make for relatively easy porting, other than places where going from 32 to 64 bits causes problems, which shouldn't be too many places. Hopefully, GCC will have improved code generation for the ALPHA by that time... Where as many CISC chips (ie, the i86 and M68k series) provide a user interface to the VM hardware using a page-table abstraction, most RISC chips I've seen (ie, the R4000 series) don't waste space on microcode and force the user to handle TLB misses in software. I find it unlikely that the ALPHA will be any different, in which case PAL code will be unecessary and the TLB exception handler can use any sort of page table we care to use - like the i386 page table. NCR53c810 support - if the chips are running little-endian, there shouldn't be any problems.