Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarApr2017

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search

Messages from 9300

Article: 9300
Subject: Re: The case for free operating systems and EDA
From: "rk" <stellare@erols.com.NOSPAM>
Date: 6 Mar 1998 00:33:44 GMT
Links: << >>  << T >>  << A >>
peter:
: :There are lots of stupid things in NT but none of them make it in any
: ;way unsuitable for EDA.

wen:
: Oh yes.  For one, it prevents me from using the tools if I am not sitting
: in front of of the machine.

rk:
solution: sit in front of the machine.

i think nt (and probably linux) is winner here over unix since with nt, you
can simply fold up the machine, put it in your backpack, hop on the plane,
wait for the announcement that you can pull out your 'puter, and get back
to work.  i travel a modest amount on business, 6-8 times per year, and i
do some personal travelling.  i see a lot of pc's with windows and mac's
being used everywhere.  never saw anybody doing unix on the plane.  and i
frequently pack up a pc, stick it in a box, get to where i'm going, take it
out, turn it on, don't miss a beat.  NOT being tied to the network makes it
the portable computer and i don't lose any of the environment i'm used to. 
can you do this with unix box?  can't with the way ours are setup but not a
unix guru, just a key tapper.

also, what about the remote tools for the pc?  they're not very expensive
although i don't have any first-hand experience with them.  perhaps someone
else can comment on their performance.

then again, tools for pc are cheap.  got a set for home pc.  don't gotta go
to work.  and can upload/download files no prob over the internet.  hooking
up drag and drop from home pc to work 
pc now, it'll be even easier to communicate.
 
--------------------------------------------------------------
rk

"there's nothing like real data to screw up a great theory" 
- me (modified from original, slightly more colorful version)
--------------------------------------------------------------
 
Article: 9301
Subject: Re: Using Java for PLI?
From: Jian_Zhang <jzhang@optilink.dsccc.com>
Date: Thu, 05 Mar 1998 17:36:25 -0800
Links: << >>  << T >>  << A >>
Thanks for the reply, Bob.  

What I'm looking for is to use object oriented portion of C++ (classes,
inheritance, etc.) to build my models using PLI for verilog or FLI for
VHDL.  The existing standard PLI only supports procedure oriented C
programming and C++ classes are not allowed.  

Furthermore, I am also hopping to be able to present simulation output
using some kind of GUI.  I know there are tools like i-Logix' HDL
express, etc doing things like that.  But I need to use standard
simulators such as Modelsim, VSS, etc as my primary simulation
environment and I need to have more control to the code and using
standard interfaces like PLI/FLI.

Although I am hardware designer, I started using Java as my GP
programming language early last year and have been hooked since.

Jian Zhang
Petaluma, CA

P.S. I am not doing hardware/software co-simulations.  I am an ASIC RTL
designer and need to verify only the RTL designs I did.



> 
> I don't know of anyone currently using Java (sort of a back burner
> project
> of mine :), but people DO use C++ for PLI/FLI code (also for system test
> and verification).
> 
> There will be presenting a paper one method of "hooking up" tests
> written in C (or C++) to Verilog at the upcoming IVC conference in
> Santa Clara.
> 
> There's also a commercial product available that pretty much does the
> same thing (VCPU from SimTech).
> 
> Is this what you're talking about?
> 
>   --Bob
> 
> --
> Bob Beckwith
> To reply, remove NOSPAM. from the email address above.
Article: 9302
Subject: XC4000EX input hysteresis
From: "Allan Herriman" <allan.herriman@fujitsu.com.au.NOSPAM>
Date: 6 Mar 1998 01:44:01 GMT
Links: << >>  << T >>  << A >>
The xilinx 4000EX data sheet says that there is an input hysteresis of
"about 300mV", however there is no guaranteed figure in the DC
characteristics section.

Does anyone know what is guaranteed, and if it applies to all input pins in
CMOS or TTL modes?

I have a slow slewing signal (~5V/30ns) that I wish to use as a clock.  I
don't wish to incur the cost of an external buffer.

TIA,
Allan.
Article: 9303
Subject: Pricing info for OTP FPGA
From: "C. F. Fung" <ecffung@ntu.edu.sg>
Date: 6 Mar 1998 01:47:03 GMT
Links: << >>  << T >>  << A >>
I am looking for an OTP FPGA at the size of about 30,000 to 50,000 gates.
Anybody knows which manufacturer provide the lowest cost parts for this
sort of FPGA? I've got a rough quote for Actel's A32300DX series. It costs
about $500. Considering it is an OTP part, it doesn't seem very attractive
to
me in terms of pricing.

C. Fung
Article: 9304
Subject: Re: Pricing info for OTP FPGA
From: "rk" <rich.katz@gsfc.nasa.gov.NOSPAM>
Date: 6 Mar 1998 02:38:11 GMT
Links: << >>  << T >>  << A >>
cf:
: I am looking for an OTP FPGA at the size of about 30,000 to 50,000 gates.
: Anybody knows which manufacturer provide the lowest cost parts for this
: sort of FPGA? I've got a rough quote for Actel's A32300DX series. It
costs
: about $500. Considering it is an OTP part, it doesn't seem very
attractive
: to
: me in terms of pricing.

rk:

hmmm ... haven't bought any a32300dx's yet.  but here's a comment or two,
perhaps it could help.

one, you have to define 'gates.'  the a32300dx family's gate count may be
higher than some other manufacturers (without trying to start a gate count
war).  of course, if you use their new 42mx family, you can double your
gate count for free!  (to be fair, they quote it two ways, as pld gates and
gate array gates.  but it's a good trick to double my design efficiency in
gates per day).  don't know the pricing for the 42mx stuff, but it is a
smaller die, and presumably cheaper than the dx.  42mx = 0.45 um.  dx = 0.6
um.

might want to try some different packaging options.  sometimes that helps.

an option for some of the actel parts is the -F version.  you can guess
what the 'f' stands for.  but, if you are not pushing the technology for
speed, and these parts are say 30-40% slower than standard speed, you may
be able to get a bargain.

also, you can look at the pasic3 series from quicklogic.  these parts are
0.35 um, 4 layer metal.  should be cheap.  and, relative to another
thread/brewing flame war, the die is small.  i opened up some ql3025's
(quoted at 25kgates, iirc) and the die is tiny, no doubt about that.

if you'll be purchasing in volume, you may want to prototype with otp fpga.
 and then plunk down a quick turn asic such as the cx2041 from chipx.  i
know that you can get the pinouts to match up as well as some of the
footprints.

of course, gate count is in the idea of the beholder.

hope this helps,

good luck!
 
--------------------------------------------------------------
rk

"there's nothing like real data to screw up a great theory" 
- me (modified from original, slightly more colorful version)
--------------------------------------------------------------
Article: 9305
Subject: Re: Using Java for PLI?
From: Bob Beckwith <beckwith@NOSPAM.whinny.tdh.qntm.com>
Date: Thu, 05 Mar 1998 22:26:47 -0500
Links: << >>  << T >>  << A >>
Jian_Zhang wrote:
> 
> Thanks for the reply, Bob.
> 

You're welcome!

> What I'm looking for is to use object oriented portion of C++ (classes,
> inheritance, etc.) to build my models using PLI for verilog or FLI for
> VHDL.  The existing standard PLI only supports procedure oriented C
> programming and C++ classes are not allowed.

I know of people that have done this with verilog. You may have to
resort
to some fairly sophisticated hackery (an oxymoron?) in order to make it
work, but it definitely can be done. This biggest problem is static
object
constructors (and there may be some issues if you try to mix C stdio
with
C++ streams) but you can certainly write C++ classes to encapulate the
PLI functionality in a straightforward manner. What you can NOT do with
the PLI is modify the netlist topology (you could write even more
classes
that allow you do implement "hardware" widgets, but where they connect
in with your HDL code would presumably be "fixed") 

> 
> Furthermore, I am also hopping to be able to present simulation output
> using some kind of GUI.  I know there are tools like i-Logix' HDL
> express, etc doing things like that.  But I need to use standard
> simulators such as Modelsim, VSS, etc as my primary simulation
> environment and I need to have more control to the code and using
> standard interfaces like PLI/FLI.
> 

You should look into the MVC (model/view/controller) methodology that
evolved with Smalltalk. In a nutshell, each entity (i.e. object) has
3 parts (listed above). The model - which is the actual object, a view
which is its visual representation and a controller (a way of making the
model do things and also vary model parameters). You could adopt this
and implement in parts. Once you've got all three going, you'll have
your
GUI.
    
> Although I am hardware designer, I started using Java as my GP
> programming language early last year and have been hooked since.
> 
> Jian Zhang
> Petaluma, CA
> 
> P.S. I am not doing hardware/software co-simulations.  I am an ASIC RTL
> designer and need to verify only the RTL designs I did.
> 

I am a verification engineer, and I won't speak for co-simulation at
this
point, but I'd MUCH rather write my tests in C (or C++) than in an HDL.
I
know there are those who would disagree with me (and some rather
strongly,
I'm sure), but this is my preference. They're welcome to theirs.

-- 
Bob Beckwith
To reply, remove NOSPAM. from the email address above.
Article: 9306
Subject: Leonardo/Xilinx BUFGLS question
From: janovetz@ews.uiuc.edu (Jacob W Janovetz)
Date: 6 Mar 1998 04:36:07 GMT
Links: << >>  << T >>  << A >>
Hello,

   I read in the Xilinx documentation that the 4000XL series allows
mapping an internal signal to an available BUFGLS.
(page 4-43 9/96 databook).  I have not been able to do this, however.
In Leonardo, I have tried assigning a BUFGLS to the signal and 
even tried mapping a component in.  However, when it comes time for
the Xilinx tools to place and route, I get the following error:

--------------------------------------------------
Processing global clock buffers...
ERROR:baste:263 - The LOC constraint "P175" (a IOB location) is not valid for
   IPAD symbol `ix2282' (pad signal=dsptx_wr), which is being mapped to the
   following site types:
        CLKIOB
--------------------------------------------------


   Can anyone lend some advice?  Thanks!


    Cheers,
    Jake
--
   janovetz@uiuc.edu    | Once you have flown, you will walk the earth with
 University of Illinois | your eyes turned skyward, for there you have been,
                        | there you long to return.     -- da Vinci
        PP-ASEL         | http://www.ews.uiuc.edu/~janovetz/index.html
Article: 9307
Subject: Re: Leonardo/Xilinx BUFGLS question
From: janovetz@ews.uiuc.edu (Jacob W Janovetz)
Date: 6 Mar 1998 05:04:02 GMT
Links: << >>  << T >>  << A >>
janovetz@ews.uiuc.edu (Jacob W Janovetz) writes:

>   I read in the Xilinx documentation that the 4000XL series allows
>mapping an internal signal to an available BUFGLS.
>(page 4-43 9/96 databook).  I have not been able to do this, however.
>In Leonardo, I have tried assigning a BUFGLS to the signal and 
>even tried mapping a component in.  However, when it comes time for
>the Xilinx tools to place and route, I get the following error:

>--------------------------------------------------
>Processing global clock buffers...
>ERROR:baste:263 - The LOC constraint "P175" (a IOB location) is not valid for
>   IPAD symbol `ix2282' (pad signal=dsptx_wr), which is being mapped to the
>   following site types:
>        CLKIOB
>--------------------------------------------------


>   Can anyone lend some advice?  Thanks!




Well, I think I just answered my own question.  I had to map the
pin to an IBUF first, then to a BUFGLS.  This helped....

    Cheers,
    Jake

--
   janovetz@uiuc.edu    | Once you have flown, you will walk the earth with
 University of Illinois | your eyes turned skyward, for there you have been,
                        | there you long to return.     -- da Vinci
        PP-ASEL         | http://www.ews.uiuc.edu/~janovetz/index.html
Article: 9308
Subject: Re: The case for free operating systems and EDA
From: Mark Willey <willey@etla.ml.org>
Date: 6 Mar 1998 06:00:05 GMT
Links: << >>  << T >>  << A >>
In comp.arch.fpga Thomas Pornin <pornin@news.ens.fr> wrote:
: In article <6dn451$b0c$1@nntp.Stanford.EDU>,
: Mark Willey  <willey@etla.ml.org> wrote:
:>The tools I refer to are:
: [...several useful tools...]
:>One of the most important tools I use is X windows.  It does me absolutely
:>no good to have an NT box or server across campus.  However, if I have a
:>UNIX box, I just log in remotely and run my CAD tool while displaying the
:>application GUI on my local X server.

: Actually, all (or almost all) the tools you describe are available for NT
: for free; the "export DISPLAY" one exists also and is called NTrig (or
: something like that)(and it costs much !)(there is also something free I
: saw once, but it was not very efficient in terms of network consumption).

The tools that are available for NT form that list (and *no*, not all of
them are) are often quite buggy.  Perl under NT, for example has a lot of
problems.  I used many of these tools under NT last summer and was
thoroughly disgusted.

: So you can get many posix tools and make your NT look like a Unix. The

You can get them, but they are often very buggy - and don't fit together as
seamlessly as they do under UNIX.  POSIX under NT always has been a sick
joke.

: point is: if you want something like Unix, why do you use NT ? Get Linux,
: it is faster and cheaper, and at least as reliable as NT.

Agreed.

: Anyway, I once met a guy who was doing some development on FPGAs using
: only NT -- and he was quite happy and efficient, more than he could be
: with a Unix shell. I think I represent the opposite: I am painfully slow

Um...  yes, more than with a UNIX shell...  but who just uses a UNIX shell?
People use the GUIs that run on UNIX.

: To sum up: some people prefer NT, other prefer Unix. So it is a pity that
: tools are not as available under Unix as they are under NT.

Well, the tools are coming to NT.  They really aren't there as of yet.  The
shame is that they are missing the easier target of Linux.

Mark

Article: 9309
Subject: Re: The case for free operating systems and EDA
From: Mark Willey <willey@etla.ml.org>
Date: 6 Mar 1998 06:18:31 GMT
Links: << >>  << T >>  << A >>
In comp.arch.fpga rk <stellare@erols.com.NOSPAM> wrote:
: peter:
: : :There are lots of stupid things in NT but none of them make it in any
: : ;way unsuitable for EDA.

: wen:
: : Oh yes.  For one, it prevents me from using the tools if I am not sitting
: : in front of of the machine.

: rk:
: solution: sit in front of the machine.

Uh... who wants to do that?  I'd rather do my work from wherever I want to
be, such as on the train, at home, or out in the garden.  Not in the
office breathing recycled air and sitting in uncomfortable ergo chairs.

: i think nt (and probably linux) is winner here over unix since with nt, you
: can simply fold up the machine, put it in your backpack, hop on the plane,
: wait for the announcement that you can pull out your 'puter, and get back
: to work.  i travel a modest amount on business, 6-8 times per year, and i
: do some personal travelling.  i see a lot of pc's with windows and mac's
: being used everywhere.  never saw anybody doing unix on the plane.  and i
: frequently pack up a pc, stick it in a box, get to where i'm going, take it
: out, turn it on, don't miss a beat.  NOT being tied to the network makes it
: the portable computer and i don't lose any of the environment i'm used to. 
: can you do this with unix box?  can't with the way ours are setup but not a
: unix guru, just a key tapper.

I think the thing that you're missing here is that Linux *is* a UNIX.
NetBSD *is* a UNIX.  And it runs on more hardware than MS Windows and Mac
combined, including PCs.  So any claims about the portability of the PC is
even more true for the UNIX world.  Case in point, you can disconnect from
the network on a Mac or UNIX, move to a new location, and reconnect to the
network w/o rebooting.  Last time I checked, Windows could not do that.

: also, what about the remote tools for the pc?  they're not very expensive
: although i don't have any first-hand experience with them.  perhaps someone
: else can comment on their performance.

They...  suck.  ;-)

: then again, tools for pc are cheap.  got a set for home pc.  don't gotta go
: to work.  and can upload/download files no prob over the internet.  hooking
: up drag and drop from home pc to work 
: pc now, it'll be even easier to communicate.

Who wants to upload/download files?  Ugh.  Welcome to the 80's, man...  ;-)
Just NFS mount the work drive or telnet in and run the job on the remote
machine.

The key point of the message:  PC != Windows.  Linux/*BSD = UNIX.  UNIX !=
workstation.  UNIX runs on a PC, too.  Only faster/better.  :)

Mark "Not a fanatic, really! ;-)" Willey

Article: 9310
Subject: CFP -- Save the date: On-Line Calendar of Science & Technology Events.
From: "N. Gat" <ng@oksi.com>
Date: Thu, 05 Mar 1998 23:06:08 -0800
Links: << >>  << T >>  << A >>
Hello,

You may want to post your announcement at the TechExpo Calendar of
Science & Technology events at http://www.techexpo.com (select
TechCALENDAR).   Over 1,000 conferences are currently posted, updated
daily, the site is   visited daily by 1,000 to 3,000 scientists,
engineers, and technical   managers.  Posting is free to qualified
organizations.

Regards,

N. Gat, Ph.D.

(I apologise if this message reached you twice, or if I miised your
prior posting at TechExpo)



Article: 9311
Subject: Re: Version Control for schematics?
From: mushh@jps.net (David Decker)
Date: Fri, 06 Mar 1998 08:30:05 GMT
Links: << >>  << T >>  << A >>
I have had a couple of email responses from people who are
interested in schematic version control. One is working on his own
custom solution as a background task. I thank those that responded.
I encourage others to post follow-up messages to this news group,
however, so the venders can see there is a demand.

I also had a phone call from the PVCS people. They claim that PVCS can
automatically check in and out multiple dirs, and that PVCS will
automatically recognize extra files being checked in, and will
properly include them in the version. So one main function I was
looking for is already there. I won't need .zip or TAR.

PVCS also claims they can provide the rest of my whish list, but at
a cost of me paying for a PVCS expert to come to my company for two
days to write the necessary scripts. I'm looking into it.

That's it for this update.
Dave Decker
Diablo Research Co. LLC

Please use only one 'h' in mush. I'm trying to reduce the spam.



"Animals .  .  . are not brethren they are not 
underlings;  they are other nations, 
caught with ourselves in the net of life and time, 
fellow prisoners of the splendor and travail of 
the earth."
Henry Beston -  The Outermost House
Article: 9312
Subject: Re: The case for free operating systems and EDA
From: z80@ds2.com (Peter)
Date: Fri, 06 Mar 1998 10:44:00 GMT
Links: << >>  << T >>  << A >>

I should have phrased it differently. NT is easy to crash if there are
hardware problems, and of course hardware problems can be caused by
software.

Let's say you have a normal user program. 

It is very hard for that program to crash NT in the traditional way:
by crapping into memory which doesn't belong to it - although
reportedly this is just about possible, as someone showed by writing a
DOS app which writes 0x00 into every single address in its 1MB address
space.

But let's say that program sends an invalid command to the SCSI
controller, and messes it up. Then NT dies with it, too.

And of course a DOS or win16 app can just do

 cli
loop:
 jmp loop

and you have a dead machine. A win32 app cannot do this however.


Peter.

Return address is invalid to help stop junk mail.
E-mail replies to zX80@digiYserve.com but
remove the X and the Y.
Article: 9313
Subject: Re: The case for free operating systems and EDA
From: z80@ds2.com (Peter)
Date: Fri, 06 Mar 1998 10:44:01 GMT
Links: << >>  << T >>  << A >>

>CVS
>diff
>gcc
>emacs/vi
>make
>perl
>shell scripts
>screen
>grep

It's funny - I use none of the above, except make. VI is for
masochists :) And I have designed some 300 different board-level
products, a dozen FPGAs, a few ASICs.

It just shows how far apart the "workstation" and "PC" worlds can be,
even today.


Peter.

Return address is invalid to help stop junk mail.
E-mail replies to zX80@digiYserve.com but
remove the X and the Y.
Article: 9314
Subject: Re: The case for free operating systems and EDA
From: "rk" <rich.katz@gsfc.nasa.gov.NOSPAM>
Date: 6 Mar 1998 11:26:00 GMT
Links: << >>  << T >>  << A >>
thomas:
: : Anyway, I once met a guy who was doing some development on FPGAs using
: : only NT -- and he was quite happy and efficient, more than he could be
: : with a Unix shell. I think I represent the opposite: I am painfully
slow

mark: 
: Um...  yes, more than with a UNIX shell...  but who just uses a UNIX
shell?
: People use the GUIs that run on UNIX.

rk:
a lot of people don't use the gui and are command line oriented.  a big
plus when you want to operate, say, via telnet, and don't have access to a
high speed link to support a gui system.  and, for hdl-based eda, you want
all of the cpu cycles doing real work, not drawing pictures.

Article: 9315
Subject: Re: The case for free operating systems and EDA
From: svoiski@mcr.spb.ru.nospam
Date: Fri, 06 Mar 98 14:33:19 +0300
Links: << >>  << T >>  << A >>
> rk:
> solution: sit in front of the machine.
>
> i think nt (and probably linux) is winner here over unix since with nt, you
> can simply fold up the machine, put it in your backpack, hop on the plane,
> wait for the announcement that you can pull out your 'puter, and get back
> to work.  i travel a modest amount on business, 6-8 times per year, and i
> do some personal travelling.  i see a lot of pc's with windows and mac's
> being used everywhere.  never saw anybody doing unix on the plane.  and i
> frequently pack up a pc, stick it in a box, get to where i'm going, take it
> out, turn it on, don't miss a beat.  NOT being tied to the network makes it
> the portable computer and i don't lose any of the environment i'm used to.
> can you do this with unix box?  can't with the way ours are setup but not a
> unix guru, just a key tapper.
>

There is a significant advantage of Linux over Windows concerning portables.
I was very surprised when I touched the processor inside my desktop when
it was running Linux. It was COLD. So was the power regulator. Under
Windows95 the processor is always hot, even when OSR2 System Monitor shows 1%
CPU utilisation for an hour. Same about NT.

Of course, when CPU-intensive application runs under Linux, the processor
warms up. But for typing text on the portable Linux should give much longer
battery life.

Misha.


Article: 9316
Subject: Re: The case for free operating systems and EDA
From: "rk" <rich.katz@gsfc.nasa.gov.NOSPAM>
Date: 6 Mar 1998 11:51:14 GMT
Links: << >>  << T >>  << A >>
peter:
: : : :There are lots of stupid things in NT but none of them make it in
any
: : : ;way unsuitable for EDA.
 
wen:
: : : Oh yes.  For one, it prevents me from using the tools if I am not
sitting
: : : in front of of the machine.
 
rk:
: : solution: sit in front of the machine.
 
mark:
: Uh... who wants to do that?  I'd rather do my work from wherever I want
to
: be, such as on the train, at home, or out in the garden.  Not in the
: office breathing recycled air and sitting in uncomfortable ergo chairs.

rk:
perhaps you should have read the next line of my post.  can you do your eda
on a train w/ unix?  sorry, haven't seen that, my experience (somewhat
limited) usually had unix on fairly big machines.  and i haven't seen the
standard eda applications running on mac or pc hardware running unix os's. 
is that available?

rk eloquently said: 
: : i think nt (and probably linux) is winner here over unix since with nt,
you
: : can simply fold up the machine, put it in your backpack, hop on the
plane,
: : wait for the announcement that you can pull out your 'puter, and get
back
: : to work.  i travel a modest amount on business, 6-8 times per year, and
i
: : do some personal travelling.  i see a lot of pc's with windows and
mac's
: : being used everywhere.  never saw anybody doing unix on the plane.  and
i
: : frequently pack up a pc, stick it in a box, get to where i'm going,
take it
: : out, turn it on, don't miss a beat.  NOT being tied to the network
makes it
: : the portable computer and i don't lose any of the environment i'm used
to. 
: : can you do this with unix box?  can't with the way ours are setup but
not a
: : unix guru, just a key tapper.


mark: 
: I think the thing that you're missing here is that Linux *is* a UNIX.

rk:
if Linux *is* a UNIX, can i just go down to the local synopsys store, order
me up a copy of UNIX synopsys, drop in the cd onto my linux/x86 portable
machine, and start running?

mark:
: NetBSD *is* a UNIX.  And it runs on more hardware than MS Windows and Mac
: combined, including PCs.  So any claims about the portability of the PC
is
: even more true for the UNIX world.

rk:
hmmm ... just a little old engineer here, but i seem to remember, let me
know if it's not true, that s/w needs to be written or tweaked for
different versions of unix.  can i take code from an hp box and run it on a
sun box, or a dec box, or a ibm box, or an cray box, or a silicon graphics
box (haven't followed these two too close since merger) or a xenix box? 
anyways, that's s/w portability.

mark:
:                                      Case in point, you can disconnect
from
: the network on a Mac or UNIX, move to a new location, and reconnect to
the
: network w/o rebooting.  Last time I checked, Windows could not do that.

rk:
while i like all three environments, unix, mac, and win, for particular
uses, being able to change physical location, disconnect and reconnect to a
network without rebooting, doesn't seem to be a very important criteria for
determining whether an operating system wins the $ if you have to choose
the three.  in fact, for a machine that is being moved or is portable, i
would want to turn it off.  in fact, i would insist on it.  i think, rather
than the cost of a reboot, which just takes a minute or so, i would worry
more about those high-speed rotating pie tins covered with rust that we
store our information on.  reloading that is a pain.  and losing info is
worse.

rk: 
: : also, what about the remote tools for the pc?  they're not very
expensive
: : although i don't have any first-hand experience with them.  perhaps
someone
: : else can comment on their performance.

mark: 
: They...  suck.  ;-)

rk:
haven't tried them myself, so i'll take your word for it.  but, i do have
experience with the laptops and their fine, easy to carry around.  and with
the move to win '95/nt, we've been writing remote capability into our
software, can operate it from anywhere there's an internet connection. 
it's GREAT! ;)  <you're supposed to chime in, 'less filling.'>

rk: 
: : then again, tools for pc are cheap.  got a set for home pc.  don't
gotta go
: : to work.  and can upload/download files no prob over the internet. 
hooking
: : up drag and drop from home pc to work 
: : pc now, it'll be even easier to communicate.

mark: 
: Who wants to upload/download files?  Ugh.  Welcome to the 80's, man... 
;-)
: Just NFS mount the work drive or telnet in and run the job on the remote
: machine.

rk:
me.  don't want to tie up the phone line running programs.  just let the
machine whirrr away and be happy.  and for a lot of graphical-based work,
doing that over a 28.8 kbps modem is sooooooooooooooooooooooooooo
sloooooooooooooooooooooooooooooooow.  yuch.  welcome to the '90s, you can
take your computing cycles with you.  anywhere.  

i'm thinking the model of the processor in a room where remote users use a
terminal to get access.  what is that, 1962? my memory fades ;)

telnet is nicer for some applications, haven't seen it for windows.

mark: 
: The key point of the message:  PC != Windows.  Linux/*BSD = UNIX.  UNIX
!=
: workstation.  UNIX runs on a PC, too.  Only faster/better.  :)


rk:
if this logic is true, then i can go down to the synopsys store, on the way
back stop off at the linux store, and then load linux and synopsys on the
pc hardware and run, no muss, no fuss.  correct?  and all of the other unix
programs, right?

mark: 
: Mark "Not a fanatic, really! ;-)" Willey

rk:
really?

--------------------------------------------------------------
rk

"there's nothing like real data to screw up a great theory" 
- me (modified from original, slightly more colorful version)
--------------------------------------------------------------
 
Article: 9317
Subject: Re: The case for free operating systems and EDA
From: Stephan Hegel <ea273@fen.baynet.de>
Date: Fri, 06 Mar 1998 15:44:52 +0100
Links: << >>  << T >>  << A >>
Peter wrote:
> 
> >CVS
> >diff
> >gcc
> >emacs/vi
> >make
> >perl
> >shell scripts
> >screen
> >grep
> 
> It's funny - I use none of the above, except make. 
No revision control ? No batch mode ?
Most of the cpu and time consuming jobs here are done by batch
(shell scripts, etc) overnight, e.g. Synthesis, ATPG, Simulations, 
Analysis, etc.
Don't tell me that you turn your PC off overnight (perhaps to get
your NT free off memory leaks ;)) and thus waste a lot of time ...

> VI is for masochists :) 
Or for those who need a small and fast editor and don't want to 
lose time by moving their hands to the mouse and back during 
touch typing, e.g. to copy just one line :).
 
> It just shows how far apart the "workstation" and "PC" worlds can be,
> even today.
Yes, I agree here :)).

Stephan.
Article: 9318
Subject: ISPD 98 - Hotel/Registration deadlines loom near!!
From: ispd98@ee.iastate.edu (Symposium 98 Acct)
Date: 6 Mar 1998 14:56:07 GMT
Links: << >>  << T >>  << A >>
	INTERNATIONAL SYMPOSIUM ON PHYSICAL DESIGN 1998
		   Embassy Suites, Monterey, CA
			  April 6-8, 1998

		 http://www.ee.iastate.edu/~ispd98

	***************************************************
	        ** HOTEL RESERVATION DEADLINE: Mar 9 **
	       ** Early registration deadline: Mar 10 **
	***************************************************

			  ADVANCE PROGRAM

The International Symposium on Physical Design provides a high-quality forum
for the exchange of ideas and results in critical areas related to the physical
design of VLSI systems.  This meeting evolved from the ACM/SIGDA Physical
Design Workshops held during the years 1987-1996. The first Symposium in 1997
was highly successful and drew such a large number of attendees that
registration had to be closed a month early. The scope of this symposium
includes all aspects of physical design, from interactions with behavior- and
logic-level synthesis, to back-end performance analysis and verification. 



				MONDAY, April 6


0915-0930       Welcome
			M. Sarrafzadeh, General Chair (Northwestern)
			D. F. Wong, Program Chair (UT-Austin)

0930-1030       Keynote Address 
		"Perspectives on Systems at 1 GHz and beyond"
		Dave LaPotin (IBM Austin Research Laboratory)

1030-1100       BREAK

1100-1230       Session 1:   Floorplanning and Placement
		Chairs: C.K. Cheng (UCSD) and Jochen Jess (Eindhoven)

		"On Wirelength Estimations for Row-Based Placement"
			A.B. Kahng, S. Mantik, I.L. Markov, A. Zelikovsky (UCLA)
		"Performance-Driven Soft-Macro Clustering and Placement by 
		 Preserving HDL Design Hierarchy"
			H.-P. Su, A.C.-.H. Wu, Y.-L. Lin (Tsing Hua)
		"Nostradamus: A Floorplanner of Uncertain Design"
			K. Bazargan, S. Kim, M. Sarrafzadeh (Northwestern)

 
1230-1430       Lunch (Pinot Noir Room)
                Special Address: "Impact of Web Technologies on EDA System
		Architectures"
			A. R. Newton (UCB)

1430-1600       Tutorial: "Timing Metrics for Physical Design of Deep
		Submicron Technologies " 
		Presenter:  L. Pileggi (CMU)
		Panelists:  J. Cong (UCLA)
			    S. Otto (Intel)
			    A. Yang (Washington)

1600-1630       BREAK

1630-1730	Special Address: "Moore's Law and Physical Design of ICs"
		  W. Maly (CMU) 

1730-1830       Session 2:   Interconnect Optimization
		Chairs: M. Alexander (Washington State) and
						Y.-L. Lin (Tsing Hua)

  		"Greedy Wire-Sizing is Linear Time"
			C.C.N. Chu, D.F. Wong (UT-Austin)
    		"An Efficient Technique for Device and Interconnect 
		 Optimization in Deep Submicron Designs"
			J. Cong, L. He (UCLA)

1900-2100       Dinner (Pinot Noir Room)
                Special Address:
		"Consorting with the Consortia: Cooperative Research
					 For Fun and Profit"
  		W. H. Joyner (SRC)
                   


				TUESDAY, April 7


0830-0930       Session 3:   Layout Methodologies for RF Circuits 
		Chairs:  M. Pedram (USC) and W. Dai (UCSC)

		"Device-Level Early Floorplanning Algorithms for RF Circuits"
			M. Aktuna, R.A. Rutenbar, L. R. Carley	(CMU)
		"A Layout Approach to Monolithic Microwave IC"
			A. Nagao, T. Kambe (SHARP); I. Shirakawa (Osaka)


0930-1030	Session 4: Framework and Benchmarks
		Chairs: D. Hill (Synopsys) and L. Jones (Motorola)

		"CHDStd--Application Support for Reusable Hierarchical
		 Interconnect Timing Views"
			S. Grout, G. Ledenbach, R. G. Bushroe, P. Fisher, 
                        A. Chokhavtia (Sematech); D. Cottrell, D. Mallis
			(Silicon Integration Initiative); S. DasGupta,
			J. Morrell (IBM)
		"The ISPD Circuit Benchmark Suite"
			C.J. Alpert (IBM)
		

1030-1100       BREAK
1100-1230       Panel: "Given that SEMATECH is levelling the semiconductor
	                technology playing field, will corporate CAD (in
	                particular, PD) tools be an enabler/
	                differentiator of technology in the future?"
		Organizer: S. DasGupta (IBM)
		Panelists:
		B. Beers, IBM
		M. Khaira, Intel
		R. Abrishami, Fujitsu Microelectronics
		J. Hutt, Synopsys
		L. Sheffer, Cadence
		D. Guiou, Mentor Graphics
		V. Kulkarni, Avant!

  
1230-1330       Lunch (Atrium Court)

1330-1430       Session 5:  "PD for Manufacturability" 
		Chairs: M. Weisel (Intel), R. Rutenbar (CMU)
		"Critical Area Computation--A New Approach"
			E. Papadopoulou (IBM), D.T. Lee (Northwestern)
		"Filling and Slotting: Analysis and Algorithms"
			G. Robins, A. Singh (Virginia); H. Wang (UCLA);
			A. Zelikovsky (Virginia)


1430-1530	Special Address: "Global Wires: Harmful?"
		R. Otten (Delft)

1530-1600	BREAK

1600-1645       Session 6: Poster Presentations 
		Chairs: S. Sapatnekar (Minnesota) and G. Robins (Virginia)
	
		"Partioning Using Second-Order Information and Stochastic-Gain
		 Functions"
			S. Dutt (UIC), H. Theny (Intel)
		"A Parallel Algorithm for Zero Skew Clock Tree Routing"
			Z. Xing, P. Banerjee (Northwestern)
		"On Convex Formulation of the Floorplan Area Minimization
			Problem"
			T. Chen, M. Fan (Georgia Tech)
		"A Pattern Matching Algorithm for Verification and Analysis 
		 of Very Large IC Layouts"
			M. Niewczas, W. Maly, A. Strojwas (CMU)
		"LIBRA--A Library-Independent Framework for Post-Layout
		 Performance Optimization"
			R. Huang (UCSB), Y. Wang (Avant!), K.-T. Cheng (UCSB)
		"Estimation of Maximum Current Envelope for Power Bus 
		 Analysis and Design"
			S. Bobba, I.N. Hajj (UIUC)
       		"New Efficient Algorithms for Computing	Effective Capacitance"
			S. Muddu (SGI)
		"Calculation of Ramp Response of Lossy 	Transmission Lines Using
		 Two-Port Network Functions"
			P. Heydari, M. Pedram (USC)
		"Switch-Matrix Architecture and Routing for FPDs"
			G.-M. Wu, Y.-W. Chang (Chiao-Tung)

1645-1745       Poster Session


1900-2200       Banquet (Rancho Canada Golf Club)




				WEDNESDAY, April 8


0830-1000       Session 7: Efficient Representation in Placement 
		Chairs: R. Otten (Delft) and C. Sechen (Washington)

		"Sequence-Pair Based Placement Method for Hard/Soft/Pre-placed
		  Modules"
			H. Murata, E.S. Kuh (UCB)
		"Rectilinear Block Placement using Permutation-Pair"
			J. Xu, C.K. Cheng (UCSD)
		"Topology Constrained Rectilinear Block Packing for Layout
		Reuse"
			M. Kang, W. Dai (UCSC)


1000-1030       BREAK

1030-1200       Panel:  "Process development and its impact on Physical Design"
	   	Moderator: N. Sherwani (Intel)
	   	Panel members: J. Cong (UCLA)
			       D. Lapotin (IBM)
			       J. Rey (Cadence)

1230-1400       Lunch (Pinot Noir Room)

1400-1530      	Tutorial: "Why Clustering is the Key to Partitioning"
		Presenter: A. Kahng (UCLA)
		Panelists: C. Alpert (IBM)
			   G. Janac (Cadence)
			   J. Lillis (UIC)

1530-1700	Session 8: Routing Algorithms
		Chairs: J. Cong (UCLA) and J. Fishburn (Lucent)

		"Chip-Level Area Routing"
			L.-C. Liu, H.-P. Tseng, C. Sechen (Washington)
		"Routing Tree Topology Construction to Meet Interconnect
		 Timing Constraints"
			H. Hou (Iowa State), S. Sapatnekar (Minnesota)
		"Analysis, Reduction and Avoidance of Crosstalk on VLSI Chips"
    			T. Stoehr, M. Alt (IBM); A. Hetzel (Bonn), J. Koehl (IBM)

1700            Symposium Closes

%%==========================================================================%%
%%                          Symposium Organization                          %%
%%==========================================================================%%

General Chair
       M. Sarrafzadeh (Northwestern) 
Past Chair
       A. B. Kahng (UCLA) 
Steering Committee
       J. P. Cohoon (Virginia) 
       S. DasGupta (IBM) 
       M. Marek-Sadowska (UC Santa Barbara) 
       B. Preas (Xerox PARC) 
       E. Yoffa (IBM) 
Technical Program Chair
       D. F. Wong (UT-Austin) 
Technical Program Committee
       M. J. Alexander (Washington State) 	C. K. Cheng (UC San Diego) 
       J. Cong (UCLA) 				W. W.-M Dai (UC Santa Cruz) 
       J. Fishburn (Lucent) 			D. Hill (Synopsys) 
       J. A. G. Jess (Eindhoven) 		L. Jones (Motorola) 
       S. M. Kang (UIUC) 			Y.-L. Lin (Tsing Hua) 
       M. Pedram (USC) 				R. Rutenbar (CMU) 
       C. Sechen (Washington) 			M. Wiesel (Intel) 
       T. Yoshimura (NEC) 
Publication Chair
       D. Hill (Synopsys) 
Panel Chair
       N. Sherwani (Intel) 
Local Arrangements Chair
       R.-S. Tsay (Axis Systems) 
Publicity Chair
       S. Sapatnekar (Minnesota) 
Treasurer
       S. Souvannavong 

Sponsors
       ACM Special Interest Group on Design Automation
       in cooperation with 
       IEEE Circuits and Systems Society and IEEE Computer Society 

Additional Support From:
   Avant! Corporation 
   Ambit Design Systems

%%==========================================================================%%
%%                   Hotel Accommodations and Travel                        %%
%%==========================================================================%%

ISPD-98 is being held at the Embassy Suites Monterey Bay in Monterey,
California, located on the beautiful Monterey Peninsula, two blocks from
the beach, at the intersection of Canyon Del Rey and Del Monte Boulevard.
The address is
	Embassy Suites Monterey Bay Hotel & Conference Center
	1441 Canyon Del Rey
	Seaside, California 93955
	Tel: (408) 393 1115
	Fax: (408) 393 1113
For hotel reservations, call (408) 393 1115 or (800)362 2779.

A block of rooms is being held for the nights of Sunday through Wednesday 
(April 5 through April 8).  Room rates are $125 per night for a single room
and $145 per night for a double room.  Any individual cancellations within 72
hours from the date of arrival will be billed for (1) night's stay, plus tax.

        +---------------------------------------------------------+
        |  Please make room reservations directly with the hotel  |
        |  at either 1-408-393-1115 or 1-800-362-2779, mentioning |
        |  ``ISPD'' to get the special rate.                      |
        +---------------------------------------------------------+

The number of rooms available at this rate is limited, and are only being 
held through March 9. Early room reservation is highly recommended.  

%%==========================================================================%%
%%                   ISPD-98 Advance Registration Form                      %%
%%==========================================================================%%

Name: _______________________________________________________

Company/University: _________________________________________

Responsibility/Title: _______________________________________

Address: ____________________________________________________

City: _______________________________ State: ________________

Country: ______________________ Postal Code: ________________

Phone: ________________________ Fax: ________________________

Email: ______________________________________________________

Food Choices:
      [  ] Vegetarian meals
      [  ] Non-vegetarian meals
      [  ] Either one is fine

                        Advance               Late 
                   (Through March 10)    (After March 10) 
ACM/IEEE Members       [  ]   $350           [  ] $425 
Non-Members            [  ]   $425           [  ] $500 
Full-Time Students     [  ]   $175           [  ] $225

Student ID is required if registering as a student.

ACM or IEEE Member No. _____________________________

Registration fee includes meals and Banquet.  The Banquet will be held at
the Rancho Can~ada Golf Club and ISPD will provide a bus service to the site.

Payment may be submitted via personal or company check in US funds only and
drawn on a US bank, made payable to ``ACM/International Symposium on 
Physical Design''.  Payment may also be made with credit card (circle): 

         Mastercard             Visa             American Express 

Credit Card # _______________________________________________

Expiration Date: ______________ Total Payment: ______________

Name as it appears on credit card: __________________________

Signature: ___________________________ Date: ________________

Please mail or FAX (credit card only) your completed registration form to:

   ISPD-98 Symposium Registration 
   Sally Souvannavong, Treasurer 
   P.O. Box 395 
   Pullman, WA 99163-0395 
   
   FAX: 1-509-332-6118 

Email registration will not be accepted.  Cancellations must be in writing 
and must be received by March 24, 1998.  Questions concerning symposium 
registration should be directed to Sally Souvannavong at 1-509-334-3162, 
Email: ispd98@eecs.wsu.edu. 

Article: 9319
Subject: Re: Using Java for PLI?
From: Petter Gustad <pegu@computer.orgNOSPAM>
Date: 06 Mar 1998 06:58:36 -0800
Links: << >>  << T >>  << A >>
Jian_Zhang <jzhang@optilink.dsccc.com> writes:

> Does anyone know if there exist tools or even it's possible to use Java
> as verilog PLI?  How about using C++ classes in  the code for PLI.  I
> need them to model some complex telecomm devices and it's very AWKWARD
> to use C and downright impossible to use vhdl/verilog.

Basically you can use any language as long as it support C calling
convention. E.g. in C++ you have to declare the PLI interface routines
as extern "C" (I had to make veriuser.h C++ aware), but you can still
use classes etc. in the bulk of your code. This is quite a few years
ago (I know use the method described below) and I think I had to play
some tricks to make sure the constructors was called since the main
program is not written in C++.

What we did where I work was to write a small PLI routine with a
socket interface. We have the different pattern generators and
verifiers running as separate UNIX (or they could potentially run on a
PC using INET style sockets) processes and communicating with the
Verilog model using sockets. Another advantage of this is that you
don't have to link your Verilog simulator each time you change your
test, you only compile and link the smaller C++ pattern generators.
The disadvantage is that the socket communication could become an
bottleneck, but we have limited data sent over the socket and I don't
think this is a problem. If your company has an excessive number of
simulator licenses and computers you could use this method to do
parallel simulation as well.

Petter
-- 
________________________________________________________________________
Petter Gustad     8'h2B | (~8'h2B) - Hamlet     http://home.sol.no/~pegu
Remove the obvious from my reply address to send mail -- sed s/nospam//g
Article: 9320
Subject: Re: The case for free operating systems and EDA
From: wen-king@myri.com (Wen-King Su)
Date: 6 Mar 1998 07:44:24 -0800
Links: << >>  << T >>  << A >>
In a previous article "rk" <stellare@erols.com.NOSPAM> writes:
:
;peter:
:: :There are lots of stupid things in NT but none of them make it in any
;: ;way unsuitable for EDA.
:
;wen:
:: Oh yes.  For one, it prevents me from using the tools if I am not sitting
;: in front of of the machine.
:
;rk:
:solution: sit in front of the machine.

Sorry, I can't do that.  The machine I sometimes need to use would have
so many processors, disks, and memory and powersupply that it would take
a forklift to carry.  I also routinely use manage jobs that runs on a
farm of workstations and I need to do it remotely and service them at
odd hours of the day.

:also, what about the remote tools for the pc?  they're not very expensive
;although i don't have any first-hand experience with them.  perhaps someone
:else can comment on their performance.

They don't work very well, and when I am running jobs on a farm of 20
work stations, I don't want to have 20 virtual consoles on my screen
where I have to locate and click menu buttons individually.

The problem is not with NT, the problem is with how Microsoft is having
an effect on how EDA programs were written.  There is absolutely no reason
why EDA vendors couldn't compile their program on NT but link them with
X-window libraries so that the NT machine would be useful as a remote
compute engine within a network of unix workstations.  NT still has a
long way to go to match Unix for networking and user interface.  By making
it so that engineers either has to choose between inexpensive EDA software
and versatile and mature networked user environment, Microsoft is causing
us a lot of pain.  I am happy to take NT if Microsoft would release
X-window library to the EDA vendors and encourage them them to build
programs the way I described.  But Microsoft won't do that because they
don't care about you.  They care about market share. They are a monopoly.
Article: 9321
Subject: Re: The case for Linux and EDA
From: Rick Kwan <rkwan@ix.netcom.com>
Date: Fri, 06 Mar 1998 07:46:59 -0800
Links: << >>  << T >>  << A >>
R. Mark Gogolewski wrote:
> 
> In article <34FEE93E.9EC4786F@ix.netcom.com>, Rick Kwan
>   <rkwan@ix.netcom.com> wrote:
> 
> >Pricing.  This is an interesting artificial, obsolete separator
> >between UNIX and PC platforms.  I am arguing here for equivalent
> >application pricing for UNIX and NT systems.
> 
>   <snip>
> 
> >From the software vendor's point of view, pricing should reflect
> >development and support costs.  If it is the same on both NT
> >and UNIX, then it should be priced the same.  If it is not,
> >you are getting ripped off on one platform or the other.
> 
> That is definitely the crux.  As to "rk"'s points, there is certainly
> a market out there to lower performance/cheaper software.  That
> makes the development/support costs much less.  You should get what
> you pay for.

Agreed.  And those apps didn't justify a traditional
workstation.  What has happened in the fast few years, however,
is that the workstation (POSIX) APIs have come down to the PC
in reasonably robust implementations.  Getting all the pieces
into a common design flow becomes much easier. 
 
> >(Oh, BTW, compilers and debuggers come free on Linux and
> >FreeBSD, not to mention the OS. :-)
> 
> Yeah, and that represents < 5% of development costs.  I would
> find it difficult to argue that any specific platform significantly
> changes the development costs vis-a-vis compilers, debuggers,
> etc...

You are absolutely correct.  Having made the point about costs,
I just couldn't resist the term "free".  (Note above smiley face.)
I repent.

> 
> Mark

--Rick Kwan
Article: 9322
Subject: Re: The case for Linux and EDA
From: Michael McNamara <mac@silicon-sorcery.com>
Date: 06 Mar 1998 09:25:10 -0800
Links: << >>  << T >>  << A >>
"rk" <stellare@erols.com.NOSPAM> writes:
> rk:
> agreed.  but today, the unix s/w is generally more expensive.  and the pc
> s/w is generally cheaper.  and, the high end tools tend to run on unix. 
> and the low end tools tend to run on pc.  but the goals are, as you say,
> the key, not the OS.  but with the volume of win machines in general as
> compared to unix boxes, it's simple economics that the $5k package can make
> more $ than the $90k if they both can do the job.  and here do the job
> means process in a reasonable amount of time with a reasonable amount of
> hassle.  and, actually getting back to my last msg, the small to medium
> sized designs are not as sensitive to performance as the million gate asic.
> 

	Great theroy; and it applies quite well to word processors,
spreadsheets and the like. 

	But your theory assumes completely elastic demand. 

	Something like 10% of Unix boxes (my guess) are used to run
EDA tools. I will go out on a limb here: No matter how many quality
EDA tools are ported to the intel platform (Linux or Win/NT), and no
matter how low the price, you will see far less than 1/100 of 1% of
the total number of PC's that will ever run EDA software.

	So, vendor XYZ figures they can sell 100 copies of their
combination simulator, synthesis and place & route tool on UNIX in one
year. They can change $100k for the package; so $10M total sales! 

	Now if they port to the PC, and hence cut the price to $10k
for the package, can they sell 1000 copies? Sorry, no. They may be
able to sell 200 copies.. and hence only do $2M in sales.

	That is the problem.

-- 
    /\     Michael McNamara            <mac@surefirev.com>
   /\//    SureFire Verification Inc.
  /\///\   <http://www.surefirev.com>
 _\///\/        Formerly Silicon Sorcery
  \//\/    Get my verilog emacs mode from
    \/     <http://www.surefirev.com/verilog-mode.html>
                                     
Article: 9323
Subject: Re: The case for free operating systems and EDA
From: ptkwt@user2.teleport.com (Phil Ptkwt Kristin)
Date: 6 Mar 1998 09:36:32 -0800
Links: << >>  << T >>  << A >>
In article <01bd4897$2b62d1c0$2985accf@homepc>,
rk <stellare@erols.com.NOSPAM> wrote:
>peter:
>: :There are lots of stupid things in NT but none of them make it in any
>: ;way unsuitable for EDA.
>
>wen:
>: Oh yes.  For one, it prevents me from using the tools if I am not sitting
>: in front of of the machine.
>
>rk:
>solution: sit in front of the machine.

Not always possible.  For example if you can't make it into work because
of an ice storm, you can (if you're running Linux at home) establish a ppp
connection with a system at work and set your display to your home system
(assuming you're running some kind of Unix at work) and do a reasonable
amount of work.

>
>i think nt (and probably linux) is winner here over unix since with nt, you
>can simply fold up the machine, put it in your backpack, hop on the plane,
>wait for the announcement that you can pull out your 'puter, and get back
>to work.  i travel a modest amount on business, 6-8 times per year, and i
>do some personal travelling.  i see a lot of pc's with windows and mac's
>being used everywhere.  never saw anybody doing unix on the plane.  and i
>frequently pack up a pc, stick it in a box, get to where i'm going, take it
>out, turn it on, don't miss a beat.  NOT being tied to the network makes it
>the portable computer and i don't lose any of the environment i'm used to. 
>can you do this with unix box?  can't with the way ours are setup but not a
>unix guru, just a key tapper.

Sure, not being tied to the network might make it easier to take you
laptop on the plane and work on a design, but the need for this is only
2-3 times a year for most of us.

The disadvantages of not being tied to a network greatly outweigh the so
called advantages (especially if you have multiple people on a project).

BTW: Linux does run very nicely on most laptops, so you could do the same
if you wanted.  As far as the environment issues you mention - you would
have the same problem if your laptop running NT was connected to a
network at the office.


>
>also, what about the remote tools for the pc?  they're not very expensive
>although i don't have any first-hand experience with them.  perhaps someone
>else can comment on their performance.
>
>then again, tools for pc are cheap.  got a set for home pc.  don't gotta go
>to work.  and can upload/download files no prob over the internet.  hooking
>up drag and drop from home pc to work 
>pc now, it'll be even easier to communicate.
> 

Well, the tools for running remotely with Linux are free.  You get
XWindows which allows you to do this sort of thing.  Now there is
something called VNC (Virtual Network Computer) which allows PCs running
NT, Suns running Solaris, and Linux boxen to display a desktop remotely -
and VNC is free (GPLed).


 phil
Article: 9324
Subject: Whats wrong with this method
From: "Allan Redenbaugh" <allan@ti.com>
Date: Fri, 6 Mar 1998 12:02:28 -0600
Links: << >>  << T >>  << A >>
Given the following code for a control register where a single bit has to
have
a seperate async reset :
(target device = Xilinx 4ke)

process ( strobe, reset, clear_bit0)

begin
if reset = '0' then
   cntl_reg <= DEFAULT_REG;
elsif clear_bit0 = '1' then
   cntl_reg(0) <= DEFAULT_REG(0);
elsif strobe'event and strobe = '1' then
   cntl_reg <= data;
end if;
end process;

I assign reset to GSR so I expect clear_bit0 to drive the reset line of a
dff
on bit 0 and the reset of the register bit to only have the GSR reset.

My synthesis tool (Leonardo) says that since I have nothing defined for the
upper bits under the clear_bit0 condition it defaults to a preset which is
not what I inteded.

I have pulled out bit 0 into its own process and everythings happy, I just
don't
understand why this method doesn't work.

Any ideas?

Thanks,
Allan Redenbaugh
Raytheon Systems




Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarApr2017

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search