From: "Scott Thibault" Newsgroups: comp.arch.embedded,comp.arch.fpga Subject: Java or bytecode processors?? Date: Wed, 30 Jan 2002 08:54:55 -0500 Organization: Green Mountain Computing Systems, Inc. Message-ID: X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 X-Complaints-To: newsabuse@supernews.com Lines: 14 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.mailgate.org!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!sn-xit-01!sn-xit-02!sn-post-02!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch comp.arch.fpga:13750 What are you're thoughts about Java and other bytecode processors? Is anyone out there using them? We are developing a bytecode processor for a high-level language (i.e. NOT Java), and I'm curious as to what people think of these things. --Scott Thibault Green Mountain Computing Systems, Inc. http://www.gmvhdl.com ###### From: hmurray-nospam@megapathdsl.net (Hal Murray) Newsgroups: comp.arch.embedded,comp.arch.fpga Subject: Re: Java or bytecode processors?? Date: Wed, 30 Jan 2002 18:03:54 -0000 Message-ID: X-Newsreader: xrn 9.02 Sender: murray@glypnod (Hal Murray) References: X-Complaints-To: newsabuse@supernews.com Lines: 21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.mailgate.org!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!sn-xit-01!sn-post-02!sn-post-01!supernews.com!corp.supernews.com!glypnod!hmurray-nospam Xref: chonsp.franklin.ch comp.arch.fpga:13755 >We are developing a bytecode processor for a high-level language (i.e. NOT >Java), and I'm curious as to what people think of these things. It's a great way to save memory but you pay for it in decoding time/resources. Memory is cheap these days. Do you have a lot of code? Compare with a RISC type system, or an even wider instruction as used in (old) microcoded machines. FPGAs can do many things in parallel. It's harder to take advantage of that if you are starting from an instruction set that thinks you have a single ALU. Another potential advantage is that you might be able to get an off-the-shelf programming environment if you pick a byte code like Java that many other people are using. -- These are my opinions, not necessarily my employer's. I hate spam. ###### Message-ID: <3C590FCC.B6B7ED5B@etechnik.uni-rostock.de> Date: Thu, 31 Jan 2002 10:35:08 +0100 From: Jens Hildebrandt X-Mailer: Mozilla 4.78 [en] (WinNT; U) X-Accept-Language: en,ru MIME-Version: 1.0 Newsgroups: comp.arch.embedded,comp.arch.fpga Subject: Re: Java or bytecode processors?? References: Content-Type: multipart/mixed; boundary="------------6912D1F0AE33766409293B9D" NNTP-Posting-Host: 139.30.201.24 X-Original-NNTP-Posting-Host: 139.30.201.24 X-Trace: news.uni-rostock.de 1012469686 139.30.201.24 (31 Jan 2002 10:34:46 +0100) Lines: 48 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.mailgate.org!skynet.be!skynet.be!fu-berlin.de!zrz.TU-Berlin.DE!news.uni-rostock.de Xref: chonsp.franklin.ch comp.arch.fpga:13783 This is a multi-part message in MIME format. --------------6912D1F0AE33766409293B9D Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Scott Thibault wrote: > > What are you're thoughts about Java and other bytecode processors? > > Is anyone out there using them? > > We are developing a bytecode processor for a high-level language (i.e. NOT > Java), and I'm curious as to what people think of these things. > > --Scott Thibault > Green Mountain Computing Systems, Inc. > http://www.gmvhdl.com Some of my collegues are/have been working in that area. Here is a paper about their project: http://www-md.e-technik.uni-rostock.de/veroeff/jsm-smartdev.pdf May be you'll find that interesting. Jens --------------6912D1F0AE33766409293B9D Content-Type: text/x-vcard; charset=us-ascii; name="jens.hildebrandt.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Jens Hildebrandt Content-Disposition: attachment; filename="jens.hildebrandt.vcf" begin:vcard n:Hildebrandt;Jens tel;fax:+49 381 4983601 tel;work:+49 381 4983537 x-mozilla-html:FALSE org:University of Rostock;Dept. of Applied Microelectronics and Computer Science adr:;;Richard-Wagner-Str. 31;Rostock-Warnemuende;;18119;Germany version:2.1 email;internet:jens.hildebrandt@etechnik.uni-rostock.de title:Dipl.-Ing. x-mozilla-cpt:;5712 fn:Jens Hildebrandt end:vcard --------------6912D1F0AE33766409293B9D-- ###### Message-ID: <3C591458.204081B7@alcatel-ke.de> Date: Thu, 31 Jan 2002 10:54:32 +0100 From: Tim Sinkins Reply-To: timothy.sinkins@alcatel-ke.de X-Mailer: Mozilla 4.61 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: comp.arch.embedded,comp.arch.fpga Subject: Re: Java or bytecode processors?? References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Organization: Alcatel Kommunikations Elektronik Cache-Post-Path: bommel!unknown@pc117.kecam-han.de X-Cache: nntpcache 2.4.0b2-MSI (see http://www.nntpcache.org/) Lines: 66 NNTP-Posting-Host: 193.99.158.1 X-Trace: 1012470794 read.news.de.uu.net 184 193.99.158.1 X-Complaints-To: abuse@de.uu.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeeds.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!t-online.de!newsfeed.wirehub.nl!bnewspeer00.bru.ops.eu.uu.net!bnewsifeed02.bru.ops.eu.uu.net!bnewspost00.bru.ops.eu.uu.net!emea.uu.net!read.news.de.uu.net!not-for-mail Xref: chonsp.franklin.ch comp.arch.fpga:13661 Hal Murray wrote: > > >We are developing a bytecode processor for a high-level language (i.e. NOT > >Java), and I'm curious as to what people think of these things. Yes I'm interested in this as well. I went for a job with one of these companies before Xmas. I passed as the work wasn't for me, but I came away wondering whether, in time, I was going to find my self out of a job unless I retrained in Java (I do high level apps that just happen to run on embedded processors). > It's a great way to save memory but you pay for it in decoding > time/resources. Memory is cheap these days. I think the real motivation is the saving on development costs. Obviously they aren't going to replace the micros in washing machines but for something like a bank ATM the cost of the processor is almost an irrelevance. Question is, "is there a saving in development cost"? Or is it just that old ageism thing again, all the new kids on the block come out of school already knowing Java, us oldies don't and everybody knows that a company is much better off with a room full of 20 somethings than a room full of 40 somethings, don't they :-(. After all, the 20 somthings are going to be with the company for 40 years instead of 20, aren't they! > Do you have a lot of code? > > Compare with a RISC type system, or an even wider instruction > as used in (old) microcoded machines. FPGAs can do many things > in parallel. It's harder to take advantage of that if you > are starting from an instruction set that thinks you have a > single ALU. Are you sure about this. As Java is an OO language (I think!) it would seem to make sense for the processors to be architured to take advantange of this, do they? (Genuine question, I have no idea of the answer) > Another potential advantage is that you might be able to > get an off-the-shelf programming environment if you pick > a byte code like Java that many other people are using. Yep, as above, however I don't think it's just a case of getting an off the shelf environment, but the SAME off the shelf environment (that they already use for their Web programming for example). But this goes further. Is the motivation that you can use the SAME programmers? (as for your web development), ignoring the fact that the the two disciplines require completely different analyticaly skills, or do they? Does the processor cross this bridge for you? I've no idea, if it does I'm going to start to get worried. Tim > -- > These are my opinions, not necessarily my employer's. I hate spam. ditto (perhaps I should make myself a sig ###### From: "Tauno Voipio" Newsgroups: comp.arch.embedded,comp.arch.fpga References: <3C590FCC.B6B7ED5B@etechnik.uni-rostock.de> Subject: Re: Java or bytecode processors?? Lines: 31 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Message-ID: <4G968.149$ug2.13063@read2.inet.fi> Date: Thu, 31 Jan 2002 11:14:08 GMT NNTP-Posting-Host: 62.71.164.128 X-Trace: read2.inet.fi 1012475648 62.71.164.128 (Thu, 31 Jan 2002 13:14:08 EET) NNTP-Posting-Date: Thu, 31 Jan 2002 13:14:08 EET Organization: Sonera corp Internet services Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeeds.belnet.be!news.belnet.be!news.tele.dk!small.news.tele.dk!206.252.192.28!news.stealth.net!newsfeed.song.fi!nntp.inet.fi!central.inet.fi!inet.fi!read2.inet.fi.POSTED!not-for-mail Xref: chonsp.franklin.ch comp.arch.fpga:13682 > Scott Thibault wrote: > > > > What are you're thoughts about Java and other bytecode processors? > > > > Is anyone out there using them? > > > > We are developing a bytecode processor for a high-level language (i.e. NOT > > Java), and I'm curious as to what people think of these things. > > > > --Scott Thibault Some twenty-odd years ago Western Digital made a processor called WD90. It run the UCSD Pascal P-code as its native language. Another processor resembling current bytecode processors is the HP 3000 multi-user system of the late 1970's. It was a computer too much before its time. A bait for Ms Rather: The current bytecode processors run a code very near Forth, right? A question: What is the difference between a CISC processor with peculiar instruction set and a bytecode processor? Tauno Voipio tauno voipio @ iki fi ###### From: "Scott Thibault" Newsgroups: comp.arch.embedded,comp.arch.fpga Subject: Re: Java or bytecode processors?? Date: Thu, 31 Jan 2002 09:40:16 -0500 Organization: Green Mountain Computing Systems, Inc. Message-ID: References: <3C591458.204081B7@alcatel-ke.de> X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 X-Complaints-To: newsabuse@supernews.com Lines: 67 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.ifi.unizh.ch!86815!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!sn-xit-01!sn-post-02!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch comp.arch.fpga:13753 "Tim Sinkins" wrote in message news:3C591458.204081B7@alcatel-ke.de... > > > Hal Murray wrote: > > > > >We are developing a bytecode processor for a high-level language (i.e. NOT > > >Java), and I'm curious as to what people think of these things. > > Yes I'm interested in this as well. I went for a job with one of > these companies before Xmas. I passed as the work wasn't for me, > but I came away wondering whether, in time, I was going to find my > self out of a job unless I retrained in Java (I do high level apps > that just happen to run on embedded processors). > > > It's a great way to save memory but you pay for it in decoding > > time/resources. Memory is cheap these days. > > I think the real motivation is the saving on development costs. > Obviously they aren't going to replace the micros in washing > machines but for something like a bank ATM the cost of the > processor is almost an irrelevance. > > Question is, "is there a saving in development cost"? This is the main goal in our case, at least. We've focused on a high-level scripting language to simplify software development. Code density does become important when it comes to selecting a compilation vs. bytecode processor approach. Using compilation may lead to very bulky code. > Or is it just that old ageism thing again, all the new kids on > the block come out of school already knowing Java, us oldies > don't and everybody knows that a company is much better off with > a room full of 20 somethings than a room full of 40 somethings, > don't they :-(. > After all, the 20 somthings are going to be with the company for > 40 years instead of 20, aren't they! > > > Do you have a lot of code? > > > > Compare with a RISC type system, or an even wider instruction > > as used in (old) microcoded machines. FPGAs can do many things > > in parallel. It's harder to take advantage of that if you > > are starting from an instruction set that thinks you have a > > single ALU. > > Are you sure about this. As Java is an OO language (I think!) > it would seem to make sense for the processors to be architured > to take advantange of this, do they? (Genuine question, I have > no idea of the answer) I think that is what Hal was driving at. Take for instance the Pentium. Although, the instruction set has been modified it is still fundamentally tied to the 8086 instruction set that was not designed for superscalar execution. This makes implementing instruciton level parallelism in the Pentium much more complex than necessary. So much so, that it has driven Intel to experiment with the Itanium. --Scott Thibault Green Mountain Computing Systems, Inc.