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 28500

Article: 28500
Subject: Re: Altera Jam player on SHARC
From: shiva@well.com (Kenneth Porter)
Date: Mon, 15 Jan 2001 20:11:53 -0000
Links: << >>  << T >>  << A >>
Yuri G Tregubov <ytregubov@my-deja.com> wrote in 
<93uk2f$3ou$1@nnrp1.deja.com>:

>Has anybody ported Altera Jam player to the
>SHARC platform ? What are the main pitfalls?

I haven't ported Jam (I do use it on a PC) but I have ported 80186 code. I 
can tell you that the main issue is the 8/32 bit difference in chars and 
code that expects that chars are one byte wide. In particular, look out for 
code that expects that casting to (char) will reduce a 32 bit value to 8 
bits. For this I've defined a macro MASK_TO_8_BITS(v) which, for byte-like 
processors, just does a cast to char and, for the SHARC (with 32-bit 
chars), performs a mask with 0xFF.

Article: 28501
Subject: Re: CRC - from long division to XOR, how?
From: "Jamie Sanderson" <jamie@nortelnetworks.com>
Date: Mon, 15 Jan 2001 15:31:29 -0500
Links: << >>  << T >>  << A >>
<eml@riverside-machines.com.NOSPAM> wrote in message
news:3a5f7a06.46595328@news.dial.pipex.com...
> This is exactly what a parallel factorisation is. Assume you've got 8
> serial data bits going into a 4-bit LFSR. When you put in the 1st bit,
> you get 4 equations for the next state of the LFSR. Iterate 7 more
> times, for the remaining data bits, using the new values of the LFSR,
> and the last set of 4 equations is the result; these define the value
> of each bit of the LFSR after an entire byte is shifted in. You'll
> find that there are lots of common terms, and you can factorise them
> out. Eventually, you'll end up with lots of XORs, to give you a
> parallel implementation of the CRC. The TAXI app note covers it all if
> you can find it.
>
> Evan

I will continue searching for the elusive app note... Sounds like just the
thing I need.

Regards,
Jamie



Article: 28502
Subject: Re: revision control tools ??
From: Petter Gustad <spam@gustad.com>
Date: 15 Jan 2001 21:36:56 +0100
Links: << >>  << T >>  << A >>
strshn99@my-deja.com writes:

> I have a newbie question to ask you.  What kind of Revision control
> tools do you use when you are working on a design?  Any recommendation
> as to which is better?

I've used SCCS, RCS and CVS. Earlier (when disk-space was expensive) I
had only source files under revision control. Now I check in all files
which can't be easily generated in reasonable time. I tag RTL files,
its equivalent netlist, as well as SVF and MCS files for production to
know what was generated from a set of source files. 

CVS is free (so is RCS) and can operate in server mode. I've used this
(together with ssh) to work on design located 10,000 miles away.

It's also smart to include the revision numbers for the various RTL
files in the simulation dump files. Most of my source files have
something like this:

// synopsys translate_off
reg [639:0] RCS_ID;
initial 
  begin RCS_ID = "$Id: icrc.v,v 2.7 2000/07/06 09:12:42 pegu Exp $"; end
// synopsys translate_on

Then you can observe the source revision numbers in signalscan or
whatever your favorite waveform tool is.

Petter
-- 
________________________________________________________________________
Petter Gustad            8'h2B | ~8'h2B            http://www.gustad.com
#include <stdio.h>/* compile/run this program to get my email address */
int main(void) {printf ("petter\100gustad\056com\nmy opinions only\n");}

Article: 28503
Subject: Re: CRC - from long division to XOR, how?
From: Phil Short <pjs@nospam.switchingpost.com>
Date: Mon, 15 Jan 2001 21:58:48 GMT
Links: << >>  << T >>  << A >>
Jamie Sanderson wrote:
> 
> <eml@riverside-machines.com.NOSPAM> wrote in message
> news:3a5f7a06.46595328@news.dial.pipex.com...
> > This is exactly what a parallel factorisation is. Assume you've got 8
> > serial data bits going into a 4-bit LFSR. When you put in the 1st bit,
> > you get 4 equations for the next state of the LFSR. Iterate 7 more
> > times, for the remaining data bits, using the new values of the LFSR,
> > and the last set of 4 equations is the result; these define the value
> > of each bit of the LFSR after an entire byte is shifted in. You'll
> > find that there are lots of common terms, and you can factorise them
> > out. Eventually, you'll end up with lots of XORs, to give you a
> > parallel implementation of the CRC. The TAXI app note covers it all if
> > you can find it.
> >
> > Evan
> 
> I will continue searching for the elusive app note... Sounds like just the
> thing I need.
> 
> Regards,
> Jamie

There is no reason that I can think of that the TAXI chip would have
anything to say about this topic, although there might be something
about generating and checking CRCs in either the SCC manual or the
Ethernet chip manuals (it definitely appears in chapter 11 of the
Motorola McMOS Handbook, published in 1974, as well as possibly the
original Ethernet (DIX) specifications).

An alta-vista on '+CRC +polynomial +remainder' turns up a large
number of references on the basic CRC techniques (needed as a
precursor to emls' description), some of them being:

http://www.cl.cam.ac.uk/Research/SRG/bluebook/21/crc/crc.html

http://users.originet.com.br/martinnorris/crcjava.html

http://ftp.rad.com/networks/1994/err_con/crc_how.htm

There are probably a lot more down this same vein.  Once these
techniques
are understood, the algorithm that Gray uses is constructed by
performing
the one-bit serial technique four times in sequence, as eml is
describing
above.

--

Phil

Article: 28504
Subject: Re: grey code counters
From: Theron Hicks <hicksthe@egr.msu.edu>
Date: Mon, 15 Jan 2001 17:06:27 -0500
Links: << >>  << T >>  << A >>
Also, the switching noise may be substantially less as only one bit is changing
at a time.

Peter Alfke wrote:

> Good Question:
> In a Grey counter, only one bit changes on any count. So when you run two
> Grey counters asynchronously ( like the address counters in an asynchronous
> FIFO) you can compare the two counters reliably, even during the
> transition, because there will never be a "far-out" code.
> You cannot do that with a binary counter, e.g. when it changes from 7 to 8.
> During that transition, you might accidentally read any possible 4-bit
> code.
>
> Peter Alfke
> =====================
> erika_uk@my-deja.com wrote:
>
> > me i'll ask, why grey counter and not a simple binary ?
> >
> > In article <3A60E39B.B6CDBAF@earthlink.net>,
> >   lenihan3weNOSPAM@earthlink.net wrote:
> > > This sounds convenient, but the problem remains .... if a normal
> > binary up
> > > counter looks like this:
> > >
> > > always @(posedge clk) count <= count +1;
> > >
> > > and we want the synthesis tool to turn this into a grey code counter
> > for us:
> > >
> > > always @(posedge clk) count <= count +1 /* synthesis counter = grey
> > */ ;
> > >
> > > .... then there's going to be a big discrepency between RTL & gate-
> > level
> > > simulation results, if our simulator checks the counter!
> > >
> > > >
> > > > What I'd like to see (and I think I suggested this to the Synplicity
> > > > guys) is an attribute or something in the synthesis tool that lets
> > you
> > > > select gray-code or binary-code counters.
> > > >
> > > > That would be useful.
> > > >
> > > > -- a
> > > > ----------------------------
> > > > Andy Peters
> > > > Sr. Electrical Engineer
> > > > National Optical Astronomy Observatory
> > > > 950 N Cherry Ave
> > > > Tucson, AZ 85719
> > > > apeters (at) n o a o [dot] e d u
> > >
> > > --
> > > ==============================
> > > William Lenihan
> > > lenihan3weNOSPAM@earthlink.net
> > > .... remove "NOSPAM" when replying
> > > ==============================
> > >
> > >
> >
> > Sent via Deja.com
> > http://www.deja.com/


Article: 28505
Subject: Re: Alliance for Linux - not a technical issue
From: shiva@well.com (Kenneth Porter)
Date: Mon, 15 Jan 2001 22:27:10 -0000
Links: << >>  << T >>  << A >>
Simon Gornall <simon@unique-id.com> wrote in
<3A6026B5.D66C4E92@unique-id.com>: 

>But we are talking about the Alliance tools. Should Xilinx ever
>distribute (hint! hint!) Linux tools, I'd expect them to be in .rpm
>format. I don't mean 'expect' as in "they'd better be", I mean it in the
>sense "most distributors do"

RPM's are definitely desirable if you limit yourself to Red Hat and 
derivative distributions. Developers outside the RH world tend to prefer 
tarballs as they're more portable. Note that Linux binaries can be used 
with FreeBSD and Solaris with suitable compatibility libraries installed.
 
>Under Linux it's the opposite by convention - you install into
>/opt/<package>/{lib,bin,man,etc...} or /usr/local/<package>/... I think
>the /usr/local... is no longer part of the LFS spec though, I think
>/opt/<package> is the recommended layout now. 

/opt is a relatively new thing inherited from Sun. /usr/local is now 
reserved strictly for administrator-installed packages, and packages should 
not default to using /usr/local. (For example, I use /usr/local/sbin for my 
personally-written admin scripts.)

>Two "absolutely safe" methods come to mind for supporting linux:
> o Xilinx could install their own versions of the shared libs 
>   they use, and use LD_LIBRARY_PATH as above, or
> o They could link the programs statically.

The latter is preferable. If a specific library is needed, static binding 
will trim off the unneeded parts. Further, if an app spawns a system 
process, you don't want a custom LD_LIBRARY_PATH passed through to the 
child process, causing it to use the wrong library.

Article: 28506
Subject: Re: Virtex-II officially launched
From: Jim Granville <jim.granville@designtools.co.nz>
Date: Tue, 16 Jan 2001 11:56:09 +1300
Links: << >>  << T >>  << A >>
Peter Alfke wrote:
> 
<snip>
> Highest on my list of exciting features is the Digitally Controlled
> Impedance, which effectively puts the series termination resistor
> right into the output driver, or the parallel termination right into
> the input buffer ( all optionally of course). It will be a gods-end
> for people putting >500-pin packages on a pc-board, not having to
> bother with resistor packs...

 Just getting my head around this feature.

 The notes show a block of 8, sharing a pair of RefUP and RefDown 
bias resistors.
 So, what is the Static ( DC ) power component when 
using this feature ?

 - jg

Article: 28507
Subject: http://www.datasheetlocator.com/nl
From: nokiogl@yahoo.com (NokioGL)
Date: Mon, 15 Jan 2001 23:05:45 GMT
Links: << >>  << T >>  << A >>

Hello people,

just came by to inform you about this site. It's a very good and
advanced site related to the seminconductor industry, maybe it might
be interesting for you to check it out.. :)

http://www.datasheetlocator.com/nl

Greetings,
Charlie


Article: 28508
Subject: Re: grey code counters -- BUG FIX (OOPS!)
From: eteam <eteam@aracnet.com>
Date: Mon, 15 Jan 2001 16:42:18 -0800
Links: << >>  << T >>  << A >>
Apologies to all, there was a bug in the code previously posted.
There goes my reputation, shot to pieces!

Sadly, netscape translates TAB characters to spaces, and messes up the columnation...

- fixed a bug in the most significant bit-pair
- parameterised the counter width
- mapped the reset/init function to look up table logic, rather than async clear/preset

If anyone wants the MAX+2 source code file and/or simulation file, drop me an email and
I'll email it out to you.  Posting binaries on this newsgroup is a no-no.

-- Bob Elkind, eteam@aracnet.com

========= START OF CODE ===========================

constant	last = 4;	-- specify the counter width in bit-pairs (i.e. if last == 4, then counter is 8 bits wide)
						-- this parameter can just as easily be passed from the subroutine "call"

SUBDESIGN grey8
%
REVISED -- MSBs aware, so that most-significant bit-pair
counts on every "carry in".

8-bit grey code counter, generalised to be extensible.

structure is a unidirectional ripple carry, similar
to conventional binary counter.

counter is up *only*, not bidirectional.

Bob Elkind, eteam@aracnet.com
%


(
	clock				: INPUT;		-- main clock
	enable				: input=vcc;	-- clock enable (or count enable) (or could use the cry[0] term)
	reset				: input=gnd;	-- reset the puppy, default is off
	out[(last*2-1)..0]	: output;		-- outputs
)

variable
	up[last..1]			: srffe;
	temp[last..1][1..0]	: node;
	cry[last..0]		: node;
	cnt[last..1][1..0]	: dffe;

begin
	up[].clk		= clock;
	up[].ena		= enable;

	cnt[][].clk		= clock;
	cnt[][].ena		= enable;

	cry[0]			= vcc;	-- carry in to LSB is always asserted

%
The grey counter is segmented into bit pairs.
For each bit-pair, there are 4 count values, and 2 direction-of-count
values, hence 8 states.

Sequence is:

  value   direction  carry out
    0       up          0
    1       up          0
    3       up          0
    2       up          1
    2      down         0
    3      down         0
    1      down         0
    0      down         1

This is sequence is implemented as a case statement or state machine,
for each of the bit-pair segments.

Also note that a "carry out" from the bit-pair to the next more significant
bit-pair occurs on the same "states" which produce count direction changes.
%

FOR i IN 1 to last GENERATE

case cnt[i][] is
	when 0 =>
		if up[i] or (i == last) then
--		if up[i] then
			temp[i][]	= 1;
		else
			temp[i][]	= 0;
		end if;
	when 1 =>
		if up[i] then
			temp[i][]	= 3;
		else
			temp[i][]	= 0;
		end if;
	when 2 =>
		if up[i] and (i != last) then
--		if up[i] then
			temp[i][]	= 2;
		else
			temp[i][]	= 3;
		end if;
	when 3 =>
		if up[i] then
			temp[i][]	= 2;
		else
			temp[i][]	= 1;
		end if;
end case;

if cry[i-1] then							-- if carry in, then update count and maybe assert carry out
	cnt[i][].d	= temp[i][] and !reset;
	up[i].s		= (cnt[i][]==0) or reset;		-- change direction to up
	up[i].r		= (cnt[i][]==2) and !reset; 	-- change direction to down
	cry[i]		= !cnt[i][0] and (cnt[i][1]==up[i]) and !reset;		-- carry out from this bit-pair
else
	cnt[i][].d	= cnt[i][] and !reset;		-- no carry in, so no count
	up[i].s		= reset;
	up[i].r		= gnd;
	cry[i]		= gnd;
end if;

out[(i*2-1)..(i*2-2)]	= cnt[i][];			-- map bit-pairs of counter to function outputs

END GENERATE;

end;

Article: 28509
Subject: Re: Virtex-II officially launched
From: Allan Herriman <allan_herriman.hates.spam@agilent.com>
Date: Tue, 16 Jan 2001 11:45:13 +1100
Links: << >>  << T >>  << A >>
Hi Peter,

Can you please let me know if there's an option to allow the multipliers
to be clocked?  The picture in the datasheet makes it look like the
multiplier is a purely combinatorial function of the inputs.  The timing
information supports this.

I am concerned that the relatively slow clock to output delay of the
block rams coupled with the multiplier delay will limit the clock speed
to about 100MHz in filter applications.

Thanks,
Allan.

Peter Alfke wrote:
> 
> Ray, this may be the rare case where your long ( and distinguished !
> ) experience misleads you.
> In the 13 years I have been at Xilinx, I have never seen us as well
> prepared for the introduction of a new family, as we are today with
> Virtex-II.
> We have had software since last October, silicon since late
> November, and have been testing it furiously ever since. We finished
> writing a 536-page Virtex-II Handbook in December, and shipped many
> thousands of printed copies in early January. We have had extensive
> training for our FAEs last fall, and we are putting together the
> slides for a public seminar right now.
> Highest on my list of exciting features is the Digitally Controlled
> Impedance, which effectively puts the series termination resistor
> right into the output driver, or the parallel termination right into
> the input buffer ( all optionally of course). It will be a gods-end
> for people putting >500-pin packages on a pc-board, not having to
> bother with resistor packs...
> Hundreds of 18 x 18 multipliers (<4ns) are nice, as are 16 global
> clocks in all devices, each with an input mux that can glitch-free
> select between two sources. Ripple carry delay is 45 ps per bit, a
> 24-bit synchronous counter runs at 300 MHz, etc.
> Exciting stuff. As I mentioned,earlier, I am working on a 1 GHz
> frequency counter, on a 200 MHz asynchronous FIFO, and on
> metastability testing.
> 
> Sorry for the blatant propaganda. Got carried away by my enthusiasm.
> 
> I had been impatiently waiting for the Ides of January, for a long
> time !
> 
> Peter Alfke, Xilinx Applications
> =====================================
> Ray Andraka wrote:
> 
> > erika_uk@my-deja.com wrote:
> > >
> > > Hi,
> > >
> > > Does the tool handle it well ?
> >
> > Like any other new architecture, I suspect it will take a while
> > for the tools to catch up to the new silicon.  In the past,
> > it has taken about a year after the introduction of the
> > silicon before the tools were close to ready for prime time.
> > Macro libraries lag even further behind.  In the case of Virtex,
> > I didn't start recommending them for customers until a little
> > over a year ago...because of tools issues.
> >
> > >
> > > --Erika
> > > In article <979567585.218715@news2.cybercity.dk>,
> > >   "Rune Baeverrud" <fpga@no.spam.iname.com> wrote:
> > > > Hello,
> > > >
> > > > Xilinx Virtex-II has now been officially announced.
> > > >
> > > > Check out the press release:
> > > > http://www.xilinx.com/prs_rls/vtx2ship.htm
> > > >
> > > > and the Virtex-II Handbook:
> > > > http://www.xilinx.com/products/virtex/handbook/index.htm
> > > >
> > > > Some highlights are:
> > > > - digitally controlled impedances for input and output pins
> > > > - new resources for clock management and clock synthesis
> > > > - digital spread spectrum clocking
> > > > - encrypted bitstreams
> > > > - dedicated multipliers
> > > >
> > > > Go and see for yourself!
> > > >
> > > > Regards,
> > > > Rune Baeverrud
> > > >
> > > >
> > >
> > > Sent via Deja.com
> > > http://www.deja.com/
> >
> > --
> > -Ray Andraka, P.E.
> > President, the Andraka Consulting Group, Inc.
> > 401/884-7930     Fax 401/884-7950
> > email ray@andraka.com
> > http://www.andraka.com  or http://www.fpga-guru.com

Article: 28510
Subject: Re: Virtex-II officially launched
From: Peter Alfke <peter.alfke@xilinx.com>
Date: Mon, 15 Jan 2001 18:02:28 -0800
Links: << >>  << T >>  << A >>
One pair of reference resistors serves a whole bank, i.e. half a device
edge, i.e. between 20 and 150 pins, depending on the package.
The analog control circuit inside the chip pulls the resistors to a
common point.
Which means, these two resistors in series effectively span Vcc to
ground.
So 50 + 50 Ohm = 100 Ohm, equals  25 mA @2.5 V =  63 mW per bank or
half-edge.

Peter Alfke

Jim Granville wrote:

>
>  Just getting my head around this feature.
>
>  The notes show a block of 8, sharing a pair of RefUP and RefDown
> bias resistors.
>  So, what is the Static ( DC ) power component when
> using this feature ?
>
>  - jg


Article: 28511
Subject: Re: Virtex-II officially launched
From: Austin Lesea <austin.lesea@xilinx.com>
Date: Mon, 15 Jan 2001 18:06:42 -0800
Links: << >>  << T >>  << A >>
Jim,

The power dissipated in the reference resitors is less than Vcco/2
squared, divided by the resistance value, times 2.  It is sampled in time,
and is not ON continuously.  This less than 100 mW peak for each bank.  I
will make the actual meaasurements soon, but I think the time ratio is
~1/4, so all told it is closer to 25 mW per bank.

Then power is only dissipated in each IOB based on what the equivalent
resistors would have done anyway outside the package.

It is real power, so it is taken into account in the SSO table (DCI
numbers are smaller than non-DCI not due to SSO, but due to added power
and current in the supply wires), and also accounted for in the power
calculator spreadsheet.

Some IO standards are real power hogs (for example SSTL3_II. both
transmitting and receiving) so they number of these may be limited due to
the power dissipation.  Yet, there would be 5 of these SSTL3_II per power
ground pin pair in a BF957 package, with 10 power ground pin pairs in a
bank, so that would allow for 50*8, or 400 of these power hungry IO's
using DCI.  Now, you still have to consider the dissipation of all that
heat, and the power estimator will hand that to you.

All told, the box still gets just as hot as before (plus ~8*25 mW for the
reference resistors), but the resistors are gone.

Enjoy,

Austin

Jim Granville wrote:

> Peter Alfke wrote:
> >
> <snip>
> > Highest on my list of exciting features is the Digitally Controlled
> > Impedance, which effectively puts the series termination resistor
> > right into the output driver, or the parallel termination right into
> > the input buffer ( all optionally of course). It will be a gods-end
> > for people putting >500-pin packages on a pc-board, not having to
> > bother with resistor packs...
>
>  Just getting my head around this feature.
>
>  The notes show a block of 8, sharing a pair of RefUP and RefDown
> bias resistors.
>  So, what is the Static ( DC ) power component when
> using this feature ?
>
>  - jg


Article: 28512
Subject: Re: Virtex-II officially launched
From: Bob Perlman <bob@cambriandesign.com>
Date: Tue, 16 Jan 2001 05:58:15 GMT
Links: << >>  << T >>  << A >>
On Mon, 15 Jan 2001 10:14:39 -0800, Peter Alfke
<peter.alfke@xilinx.com> wrote:


>Highest on my list of exciting features is the Digitally Controlled
>Impedance, which effectively puts the series termination resistor
>right into the output driver, or the parallel termination right into
>the input buffer ( all optionally of course). It will be a gods-end
>for people putting >500-pin packages on a pc-board, not having to
>bother with resistor packs...

This will be extremely useful.

One question, Peter: some ASIC vendors have been offering this feature
for a while, but some implementations are problematic.  Is the Virtex
II impedance update mechanism asynchronous to the user's clock?  If
so, is there a guarantee that the impedance can't glitch (say, the way
a D/A might) at a critical moment?

Thanks,
Bob Perlman


Article: 28513
Subject: Re: grey code counters
From: Peter Alfke <palfke@earthlink.net>
Date: Tue, 16 Jan 2001 06:40:42 GMT
Links: << >>  << T >>  << A >>
Theoretically correct,
but in reality it depends on the complexity of the Grey-control logic.
In the design I am touting, there is actually a binary counter enclosed.
Bye-bye power saving.

Peter Alfke

Theron Hicks wrote:

> Also, the switching noise may be substantially less as only one bit is changing
> at a time.
>


Article: 28514
Subject: answer
From: chsw <chen.songwei@mail.zte.com.cn>
Date: Tue, 16 Jan 2001 01:57:10 -0800
Links: << >>  << T >>  << A >>
Hello:
 is the cpu(x86 or PowerPc or 8051) ASIC?,how is it be manufactured? Does it use the (vhdl/verilog) ?
  thanks

Article: 28515
Subject: Re: revision control tools ??
From: hamish@cloud.net.au (Hamish Moffatt VK3SB)
Date: Tue, 16 Jan 2001 11:08:15 GMT
Links: << >>  << T >>  << A >>
[Re: ClearCase]

In comp.lang.vhdl Jan Vermaete <maetej@nospam.rsd.bel.alcatel.be> wrote:
> That's possible but CVS is for free (http://www.cvshome.org)

Sure, but ClearCase is much better. For one thing, because it
has a custom virtual file system on the client (MVFS -- multi version
file system) it can do a lot of things which CVS can't do.
It's integrated quite well into NT, as well as Unix.

I guess it's expensive. We started using it for FPGA designs
within the last year. We already had it because our software
engineers have been using it for several years. I doubt we
would have bought it especially for FPGA work.


Hamish
-- 
Hamish Moffatt VK3SB <hamish@debian.org> <hamish@cloud.net.au>

Article: 28516
Subject: Re: Please explain these terms
From: "Extern" <Johan.Van.Dyck@philips.com>
Date: Tue, 16 Jan 2001 12:33:22 +0100
Links: << >>  << T >>  << A >>
Hi

I agree with Eric, you should first read some background stuff before
writing the article!

On the other hand, here are some clues for you questions:

* logical design/verification only takes the logical equations, symbols into
account like "and" or "or" but also "adders" or "muxes"  . It doesn't
include timing (and related problems), process spread, process parameters.
It is (nearly) process independent. RTL-design is a commom example.

* Physical design/verification tries to simulate/build an as accurate model
as possible of the design. It is much slower and process dependant. It is
mostly generated out of an logical design. Gate-level design is a common
example.

 * Probably you'll find much better and compleet describtions in
VHDL-study-books. Did you look on the internet too unther "vhdl-course"?

Kind regards
----------------------------------------------------------------
Johan Van Dyck
Johan.Van.Dyck@philips.com
VLSI Design Engineer
Philips Leuven (Belgium)
Johan.Van.Dyck@philips.com
----------------------------------------------------------------
<strshn99@my-deja.com> wrote in message news:93sumd$qp0$1@nnrp1.deja.com...
> Dear Gurus,
>
> I am trying to write up a little article for our college little
> newspaper but I am sort of a newbie at VHDL and FPGA. So I would like
> you gurus to explain to me:
> 1) what these terms means.
> 2) what are the tools used to do these kind of things.
>
> The terms are:
>
> a) Logical Design
> b) Physical Design
> c) Logical Verification
> d) Physical Verification
>
> thanks very much
>
>
> Sent via Deja.com
> http://www.deja.com/
>



Article: 28517
Subject: negative borrow
From: karenwlead@my-deja.com
Date: Tue, 16 Jan 2001 12:06:30 GMT
Links: << >>  << T >>  << A >>
hello,

what's negative borrow ?
(it's not for a homework, we are on holidays ;-))


Sent via Deja.com
http://www.deja.com/

Article: 28518
Subject: help
From: jack <jack@aol.com>
Date: Tue, 16 Jan 2001 04:23:43 -0800
Links: << >>  << T >>  << A >>
Hello:
  i want to know what tools(including design input/synthesis/simulation/P&R/Verifying .etc)  are used
in designing ASIC/FPGA.
  can you introuduce for me ?
  thanks

Article: 28519
Subject: Re: Looking for prototyping board
From: "Henry Styles" <hes2@doc.ic.ac.uk>
Date: Tue, 16 Jan 2001 12:31:55 -0000
Links: << >>  << T >>  << A >>
I use the Virtex-based Celoxica RC1000-PP board for applications work. It
pretty much meets your specification; PCI communication, 4xRAM modules,
simple interfacing software.

http://technology.celoxica.com/boards/boards_001.asp

Henry Styles,
Imperial College,
London



Article: 28520
Subject: Xilinx UCF/ngdbuild problem
From: Nicolas Matringe <nicolas.matringe@IPricot.com>
Date: Tue, 16 Jan 2001 13:48:37 +0100
Links: << >>  << T >>  << A >>
Hi all
I am trying to implement a design but ngdbuild finds an error in the
ucf, although it's generated with the constraints editor.

Error message:
ERROR:NgdHelpers:33 - Invalid UCF/NCF file entry value "" detected on
line 45, offset 246, while reading UCF file "bonito.ucf".

The line that causes the problem:
TIMEGRP "Protocol" = FFS( "copier_cout(41)" "copier_cout(42)"
"copier_cout(43)" "cpu_out(41)" "cpu_out(42)" "cpu_out(43)"
"dram_out(41)" "dram_out(42)" "iosys_cout(41)" "iosys_cout(42)"
"pcicache_out(41)" "pcicache_out(42)" "pcicache_out(43)" );

If I remove the "", I get another error message:
ERROR:NgdHelpers:676 - Processing TIMEGRP definition "Protocol=
   FFS(copier_cout(41)  copier_cout(42)  copier_cout(43)  cpu_out(41)
cpu_out(42)  cpu_out(43)  dram_out(41)  dram_out(42)  iosys_cout(41)
iosys_cout(42)  pcicache_out(41)  pcicache_out(42)  pcicache_out(43)) "
from UCF file: Encountered bad token "eof" while parsing group name
"FFS(copier_cout(41)".

I am using Foundation 2.1i (the synthesis was done with Leonardo
Spectrum)

If anyone has a clue...
-- 
Nicolas MATRINGE           IPricot European Headquarters
Conception electronique    10-12 Avenue de Verdun
Tel +33 1 46 52 53 11      F-92250 LA GARENNE-COLOMBES - FRANCE
Fax +33 1 46 52 53 01      http://www.IPricot.com/

Article: 28521
Subject: Re: Alliance for Linux
From: Jamie Lokier <spamfilter.jan2001@tantalophile.demon.co.uk>
Date: 16 Jan 2001 14:36:08 +0100
Links: << >>  << T >>  << A >>
Andy Peters writes:
>> Andy Peters writes:
>> > ModelSim and Synplify both use dongles (which I vastly prefer over
>> > that FlexLM bogosity) but the Mac ain't got a parallel port so I'm SOL
>> > there.
>> 
>> And even if the MAC did have a parallel port...
>> 
>> I vastly prefer FlexLM because there is some hope of it working in an
>> emulator.  In my experience dongle reading code does seriously weird
>> things to your parallel port, and doesn't work in an emulator.

> Yeah, but if FlexLM is keyed to your ethernet MAC (not Macintosh!)
> address and you want to, say, do some work at home, you're hosed.

OTOH if you run it in an emulator, you get to choose the reported MAC
addess :-)

> Here's my thought on all this: choose the software, and buy whatever OS
> and machine it runs on, rather than buying (or downloading) an OS and
> installing it on a particular machine and using hacks and such in an
> effort to get the software to work on it.

That's all very well if you're only using one package.

> Yes, I'd like ModelSim PE and the Xilinx tools and Synplify to all run
> on Linux.  But my job is to design FPGAs, and if that requires running
> NT, so be it.  Yes, I will tell the vendors that I prefer Linux.  Yes, I
> will demand that the Linux versions of the tools cost the same as the NT
> versions.  No, I will not frustrate myself trying to get software
> written for one OS to run on another.

My job is to design FPGAs too, and talk to them.
This requires that I run: Linux, NT, Solaris.

Unfortunately the Solaris box isn't as fast as the Linux box, and the NT
box is far slower.  Getting access to a 3rd computer to run NT was a
pain and took several months.  "You've already got two computers".  (The
computer cost more than the software that runs on it after all).

-- Jamie

Article: 28522
Subject: Re: revision control tools ??
From: Jamie Lokier <spamfilter.jan2001@tantalophile.demon.co.uk>
Date: 16 Jan 2001 14:39:43 +0100
Links: << >>  << T >>  << A >>
strshn99  writes:
> I have a newbie question to ask you.  What kind of Revision control
> tools do you use when you are working on a design?  Any recommendation
> as to which is better?

We use CVS over ssh for our source files (VHDL, Handel-C, Perl, C),
makefiles etc. and no version control for the generated files.

-- Jamie

Article: 28523
Subject: Re: http://www.datasheetlocator.com/nl
From: Jamie Lokier <spamfilter.jan2001@tantalophile.demon.co.uk>
Date: 16 Jan 2001 15:31:53 +0100
Links: << >>  << T >>  << A >>
NokioGL  writes:
> just came by to inform you about this site. It's a very good and
> advanced site related to the seminconductor industry, maybe it might
> be interesting for you to check it out.. :)

> http://www.datasheetlocator.com/nl

I looked up Altera and it said (to paraphrase) "click on this button to
go to Altera's datasheet page".  I clicked, nothing happened.

Shame, I was hoping for a fully indexed search engine.

-- Jamie

Article: 28524
Subject: Re: Xilinx UCF/ngdbuild problem
From: Utku Ozcan <ozcan@netas.com.tr>
Date: Tue, 16 Jan 2001 16:40:29 +0200
Links: << >>  << T >>  << A >>
Nicolas Matringe wrote:

> Hi all
> I am trying to implement a design but ngdbuild finds an error in the
> ucf, although it's generated with the constraints editor.
>
> Error message:
> ERROR:NgdHelpers:33 - Invalid UCF/NCF file entry value "" detected on
> line 45, offset 246, while reading UCF file "bonito.ucf".
>
> The line that causes the problem:
> TIMEGRP "Protocol" = FFS( "copier_cout(41)" "copier_cout(42)"
> "copier_cout(43)" "cpu_out(41)" "cpu_out(42)" "cpu_out(43)"
> "dram_out(41)" "dram_out(42)" "iosys_cout(41)" "iosys_cout(42)"
> "pcicache_out(41)" "pcicache_out(42)" "pcicache_out(43)" );
>
> If I remove the "", I get another error message:
> ERROR:NgdHelpers:676 - Processing TIMEGRP definition "Protocol=
>    FFS(copier_cout(41)  copier_cout(42)  copier_cout(43)  cpu_out(41)
> cpu_out(42)  cpu_out(43)  dram_out(41)  dram_out(42)  iosys_cout(41)
> iosys_cout(42)  pcicache_out(41)  pcicache_out(42)  pcicache_out(43)) "
> from UCF file: Encountered bad token "eof" while parsing group name
> "FFS(copier_cout(41)".
>
> I am using Foundation 2.1i (the synthesis was done with Leonardo
> Spectrum)
>
> If anyone has a clue...
> --
> Nicolas MATRINGE           IPricot European Headquarters
> Conception electronique    10-12 Avenue de Verdun
> Tel +33 1 46 52 53 11      F-92250 LA GARENNE-COLOMBES - FRANCE
> Fax +33 1 46 52 53 01      http://www.IPricot.com/

Are the lines separated by newline characters? Maybe
Ngdbuild UCF parser might not handle newline characters.
Can it be the case?

Utku



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