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
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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 40725

Article: 40725
Subject: the server to access to this newgroup
From: e-engineer@eastday.com (Sniper Daryl)
Date: 13 Mar 2002 22:10:34 -0800
Links: << >>  << T >>  << A >>
if anyone would tell me how to set in Outlook Express

Article: 40726
Subject: Difference between Virtex-II(E) und Virtex-E
From: Martin Sauer <msauer@gmx.net>
Date: Thu, 14 Mar 2002 08:00:14 +0100
Links: << >>  << T >>  << A >>
Hi,

can you tell me the difference between the Xilinix Virtex-II and the 
Virtex-E Series?
Thanks for your answer.

bye

martin


Article: 40727
Subject: Re: How to Align 7x DDR Data Input to a XC2V6000-5?
From: Mark <mark@pac.net>
Date: Thu, 14 Mar 2002 07:35:11 GMT
Links: << >>  << T >>  << A >>
I see from your description, that you use phase shifts to sample the differences between the
different toggling signals.  I hadn't considered the "combination" of these elements.  Thank
you very much for sharing your ideas!


John_H wrote:

>                  ______        ______        ______
> clk10x         _/      \______/      \______/
>                __ _____________ _____________ _____
> toggle         __X_____________X_____________X_____
>                 _   _   _   _   _   _   _   _   _
> clk35x_shift  _/ \_/ \_/ \_/ \_/ \_/ \_/ \_/ \_/ \_
>               ______ _______________ ___________ __
> togp          ______X_______________X___________X__
>               ________ ___________ _______________
> togn          ________X___________X_______________X
>                      _             _             _
> tog_diff      ______/ \___________/ \___________/ \
>                                      ___
> samplep       ______________________/   \__________
>                        ___
> samplen       ________/   \_______________________/
>                _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> ddr_data      X6X0X1X2X3X4X5X6X0X1X2X3X4X5X6X0X1X2X
>
> You're right about the toggle period.
> Though it's tough to show prop delays (the combinatorial tog_diff doesn't show a delay in
> the diagram) it seems that the pulse that covers timeslot 3 and 4 in this phase shift
> version of clk35x is independent of which clk35x transition you use to sample the
> difference.  You can apply the same timing diagram to the clk10x90deg signal with an
> unshifted clk35x and get similar results but the timing analysis is up to you.  Fun stuff!
>
> Mark wrote:
>
> > I think that with the toggling signals, they have a period of twice the clock period.
> > It seems when using clk35x_p and clk35x_n, toggle35x_p and toggle35x_n can only resolve
> > down to one clk35x resolution.  The 90 degree phase shift sounds like one way to find a
> > difference.  I believe that I once saw an appnote that described using 90 deg phases to
> > lock(?) onto an input signal.
> >
> > Now that you mention it, I do remember reading messages about clock skew and jitter in
> > FPGAs.  Also, several weeks ago, I looked into the Virtex-II skew/jitter when
> > multiplying up the clock by 7x to the rate of 420 MHz.


Article: 40728
Subject: Virtex BUFGDLL
From: "H.L" <alphaboran@yahoo.com>
Date: Thu, 14 Mar 2002 09:44:25 +0200
Links: << >>  << T >>  << A >>
Hello all,
I use a BUFGDLL in a Virtex-E FPGA to succeed a proper distribution of a
clock (155MHz). I instantiate BUFGDLL in my code and I did the port map. In
this way I have not access to the RST pin of the DLL (I want to set it '0'
as xilinx suggests). I checked that DLL's CLKIN,CLKFB and CLK0 pins are OK
but I cant check if the RST pin is '0' by default. How can I check if the
RST pin is grounded or not?

Best Regards,
Harris





Article: 40729
Subject: Spartan-XL, SpartanII and Spartan-IIE bitstream format question ...
From: meng.engineering@bluewin.ch (Markus Meng)
Date: 13 Mar 2002 23:55:21 -0800
Links: << >>  << T >>  << A >>
Hi all,

on the Xilinx documentation I do find, that the bitstream
format has a minimum of 8 '1' at the very beginning of
the serial bitstream. Is it possible to extend the minimum
8 of '1' to any higher number before the start-pattern
is being applied?

markus

Article: 40730
Subject: XILINX XC2V6000
From: haim_moshe@hotmail.com (haim moshe)
Date: 14 Mar 2002 00:03:18 -0800
Links: << >>  << T >>  << A >>
Is the difference between the buffers LVDCI_DV2_18 and LVCMOS18 only
in the internal pair 2R resistors ?

Haim

Article: 40731
Subject: use virtex2 DCM as delay line
From: nahum_barnea@yahoo.com (Nahum Barnea)
Date: 14 Mar 2002 01:23:16 -0800
Links: << >>  << T >>  << A >>
Hi.
I want to use virtex2 DCM as a delay line on a clock signal.
This means that I want to disable any loop to prevent any
added jitter to the clock. I need only specified delay
in the 50 ps steps.

How can it be done ?


ThankX ,
Nahum.

Article: 40732
Subject: Xilinix FPGA width 5V IO
From: "Martin Sauer" <msauer@gmx.net>
Date: Thu, 14 Mar 2002 10:31:13 +0100
Links: << >>  << T >>  << A >>
Hello,

one question: Is it possible to connect a 5V IO Device direct to the Xilinx
Virtex FPGA series?

Thanks for your answer.

bye

martin sauer




Article: 40733
Subject: Where can I get the information on implementing CPU with FPGA?
From: "RSM" <ryu21c@unitel.co.kr>
Date: Thu, 14 Mar 2002 18:44:27 +0900
Links: << >>  << T >>  << A >>
Thanks for your answer in advance!






Article: 40734
Subject: Re: Where can I get the information on implementing CPU with FPGA?
From: Christian Plessl <plessl@remove.tik.ee.ethz.ch>
Date: Thu, 14 Mar 2002 11:28:50 +0100
Links: << >>  << T >>  << A >>
RSM wrote:

> Thanks for your answer in advance!

www.fpgacpu.org (excellent starting point!)
www.gaisler.com (LEON CPU)
www.opencores.org

regards,
 chris

Article: 40735
Subject: Re: Synthesis tools comparison?
From: "Arash Salarian" <arash.salarian@epfl.ch>
Date: Thu, 14 Mar 2002 11:39:58 +0100
Links: << >>  << T >>  << A >>
hmmmmmmmm, quite strange! It can explain why I could not find any useful
benchmark on the web for the recent verstions of these tools....

But the question remains, and it's that if there is no benchmark
comparision, how would one choose one tool over another one? Just looking at
the feature list and vendor claims does not look like the best way...

A year ago, I used one of my big DSP designs as benchmark for these tools
and at that time Synplify outperformed both Leonardo and FPGA compiler for a
Flex10K target. But today? Who knows how these tools compete against each
other?

Should I try to downlaod the trial version of the tools, and run all of them
on my latest designs to choose one? For each new design? :) I don't think
so...

Regards
Arash


"S. Ramirez" <sramirez@cfl.rr.com> wrote in message
news:H3Rj8.118337$Dl4.13204394@typhoon.tampabay.rr.com...
> "Arash Salarian" <arash.salarian@epfl.ch> wrote in message
> news:3c8f4cb7$1@epflnews.epfl.ch...
> > Hello,
> >
> > Do you know of any recent review/benchmark comparision of major
synthesis
> > tools for the FPGA? I'm interested in such a comparision between
Synplify,
> > LeonardoSpectrum and FPGA compiler. And is there any major perfromance
> > difference between these tools when targetting different FPGAs from
Xilinx
> > and Altera?
> >
> > Best Regards
> > Arash
>
>    You will be hard pressed to find these comparisons, since the license
> agreements of these vendors say something close to "Licensee shall not ..
> disclose the results of any benchmarking of the SOFTWARE, or use such
> results for its own competing software development activities, without the
> prior written permission of Blah Blah."
>    Don't blame me, I'm just the messenger!
> Simon Ramirez, Consultant
> Synchronous Design, Inc.
> Oviedo, FL  USA
>
>



Article: 40736
Subject: Re: Synthesis tools comparison?
From: "Tim" <tim@rockylogic.com.nooospam.com>
Date: Thu, 14 Mar 2002 11:08:15 -0000
Links: << >>  << T >>  << A >>
S. Ramirez wrote

>    You will be hard pressed to find these comparisons, since the license
> agreements of these vendors say something close to "Licensee shall not ..
> disclose the results of any benchmarking of the SOFTWARE, or use such
> results for its own competing software development activities, without the
> prior written permission of Blah Blah."

You could fall back on the legal principle that "common abuse is not
actionable".  From what we hear about some of the tools, common abuse
would be a fair summary :-)





Article: 40737
Subject: where to start with constraining..
From: "Josh Pfrimmer" <yeah_spam_me@thisaddress.com>
Date: Thu, 14 Mar 2002 03:14:04 -0800
Links: << >>  << T >>  << A >>
Hi all,
    For a class (calmer down.. I'm not looking for answers to my homework)
I've designed an 8-bit CPU.  Fifteen instructions, pre-fetching, branch
predicting, load-store, etc. etc.  So far, it cruises along at ~50MHz on an
XC4010XL-3.  This is better than what's necessary for the class, but I'd
like to push it further.

    Now that I have a ballpark speed, what's the best way to get started
specifying constraints in a .ucf?  How do I get the (old, old, old)
Foundation 3.1 software in the labs to tell me what the worst paths are, so
that I can start picking out false paths, etc.

    Incidentally, the post-layout timing report gives a figure of ~20MHz fo
the maximum clock speed, but on the board, I'm well over 45.. is this
typical?  How can I get more accurate numbers out of the reports?

Thanks for your attention
JoshP



Article: 40738
Subject: Re: How would I know somebody has copied my files in Unix?
From: "e.i.chester" <e.i.chester@ncl.ac.uk>
Date: Thu, 14 Mar 2002 11:47:56 +0000
Links: << >>  << T >>  << A >>
Timmestein wrote:
[snip]

> Kelvin Hsu wrote:
> [snip]

or maybe you didn't bother setting permissions properly on your files and a
simple cp/scp worked just fine. get a book on *nix operating systems, and
then one about posting to relevant groups in usenet.


edic


Article: 40739
Subject: Re: Synthesis tools comparison?
From: Mike Treseler <tres@tc.fluke.com>
Date: Thu, 14 Mar 2002 06:52:31 -0800
Links: << >>  << T >>  << A >>
Arash Salarian wrote:

> Should I try to downlaod the trial version of the tools, and run all of them
> on my latest designs to choose one? 

That would answer the question for you.

I'm not sure the difference between the 
two products is worth spending much time over.
Consider other factors like existing licenses, 
local experts etc.

          -- Mike Treseler

Article: 40740
Subject: Proto boards for labs
From: Nitin Chandrachoodan <nitin@eng.umd.edu>
Date: Thu, 14 Mar 2002 15:45:44 GMT
Links: << >>  << T >>  << A >>

Hello,

We are trying to decide on boards for use in a course in the 
Electrical Engg. department at our university. The 
requirements we have come up with so far are as follows:

1. Easy to get started and working with it.
2. Relatively inexpensive
3. Can be used to implement reasonably large designs, 
especially DSP filters, FFTs etc. Here  we would like to 
implement the designs without having to go to too much low-
level optimization (bit-serial implementations etc.) unless 
absolutely necessary, as this would change the focus from 
learning about FPGA implementation to low-level design.
4. Can be used with existing tools (we have Synopsys DC and 
FPGA compiler II, so we would prefer not to have to purchase 
Foundation etc. unless necessary)

It seems that the Xilinx Spartan chips can be used with the 
Webpack which is free, and they also have very large 
capacities (100k-200k for XC2S200). Two boards I have found 
so far with these are

1. XESS XSA-100 (100k gate, $279) from xess.com
2. Digilab XLA 2 (XC2S200 Spartan II - 200k gates, $99) from 
digilentinc.com

The digilent board seems to have better features at a lower 
price, but XESS appears more popular overall. Are there any 
other compelling reasons for using one or the other?

Alternately, are there any other boards that are especially 
suited to classroom use? Apart from basic use in a classroom 
setting, we are also interested in boards that can be used 
for rapid reconfiguration etc., for use in research into 
reconfigurable computing techniques.

Any replies much appreciated. 

Thank you,

Nitin



Article: 40741
Subject: Re: Difference between Virtex-II(E) und Virtex-E
From: Austin Lesea <austin.lesea@xilinx.com>
Date: Thu, 14 Mar 2002 07:54:01 -0800
Links: << >>  << T >>  << A >>
Martin,

Sure.  As a member of the design, and the verification team for Virtex
II (and II-Pro), I think I can answer pretty fairly.

Virtex E was a shrink to 0.18u of the classic Virtex architecture and
circuitry.  Virtex E added LVDS input buffers to the original Virtex
design, but little else was changed.

Virtex II was a complete redesign in 0.15u at 1.5V for the core, that
extended the reach of the interconnect, and buffered virtually all paths
to reduce loading effects.  The CLB had more features added (more LUT
RAM modes, SRL modes), as well as the horizontal carry (useful for
p-terms).  The block RAMS got bigger, and there are more of them (4K vs
18K), as well as having three read/write modes instead of one in
addition to a 18X18 multiplier in each BRAM block.  The global clock
skew was reduced from ~ 300 ps in Virtex E to less than 120 ps in Virtex
II (largest parts).

The DLL got two new friends, a digital phase shifter (1/256 resolution
of a period) for variable or fixed phase shifting, and a digital
frequency synthesizer for M/D frequency sythesis.  The IOBs got current
sink/source differential LVDS outputs, and they have digital controlled
impedance for precise 50 ohm (or 25 to 100 ohm) t-line matching.  The
IOBs also got a full set of hardware to support double data rate (DDR)
clock forwarded interfacing.

Oh, and I almsot forgot, there is key storage for 6 keys, and it
supports Triple DES for encrypted bitstreams to protect your IP.

All of these features support the myriad of RAM interfaces, and other
wide and fast parallel IO.

The software got twice as fast, and many more powerful features were
added to make designs easier.  The new power and speed of Virtex II
allows for many ASIC like cores and functions that just can't run on
anything less powerful, or slower.

The bottom line for Virtex II is that we see it in all cell base station
designs, in data communications systems, routers, switches, enterprise
storage systems, video processing .... and many many more.

The family is extended by the introduction last week of the Virtex II
Pro family members, which included IBM Power PC uP's (yes, that is
plural as in more than one on the larger family members), and Rocket
IO(tm) gigabit serial transceivers (up to 3.125 Gbs each).

The smallest family member is the 2V40, with 40K "fpga" gates (e.g. 512
LUT's), that is about 30X more powerful than a clock distribution chip,
and is actually less expensive, too.  From there, you have the 2v80,
2v250, 2v500, 2v1000, .... up to the 2v6000, and later this year, the
2v8000.

Check out the documentation:

 http://www.xilinx.com/products/platform/

Austin



Martin Sauer wrote:

> Hi,
>
> can you tell me the difference between the Xilinix Virtex-II and the
> Virtex-E Series?
> Thanks for your answer.
>
> bye
>
> martin


Article: 40742
Subject: Re: use virtex2 DCM as delay line
From: Austin Lesea <austin.lesea@xilinx.com>
Date: Thu, 14 Mar 2002 07:56:57 -0800
Links: << >>  << T >>  << A >>
Nahum,

Contact your FAE, and request information on how to use the "Freeze_DLL"
feature.

This is a signal that is part of the test primitive that we do not
support, aminly because there was too much to do to get all the
supported features, and some of the new ones introduced.  Like most
designs, there are a lot of hidden features that we placed in there for
test, but may be useful for some applications.

The freeze stops all tap updates, so once locked, and the delay is set,
freezing suspends all further corrections for temperature and voltage
changes in the system.

Austin

Nahum Barnea wrote:

> Hi.
> I want to use virtex2 DCM as a delay line on a clock signal.
> This means that I want to disable any loop to prevent any
> added jitter to the clock. I need only specified delay
> in the 50 ps steps.
>
> How can it be done ?
>
> ThankX ,
> Nahum.


Article: 40743
Subject: Re: XST duplicates unnecessary IOB OE FFs
From: Brian Drummond <brian@shapes.demon.co.uk>
Date: Thu, 14 Mar 2002 16:08:43 +0000
Links: << >>  << T >>  << A >>
On Wed, 13 Mar 2002 17:04:37 -0600, Kevin Brace
<ihatespam99kevinbraceusenet@ihatespam99hotmail.com> wrote:

>        I am currently experiencing a problem with XST's Pack I/O
>Registers into IOBs option.

>I don't appreciate XST overriding the design trade off I made in the
>design, and do I have a way to prevent XST from duplicating the OE FF?
>I am using ISE WebPACK 4.1WP3.0's XST (XST E.33), and Spartan-II XC2S150
>is the target device.
>I feel like this OE FF duplication thing should not happen, and hope
>that the future version of XST will give its users an option to disable
>OE FF duplication if the user doesn't want it.
>
I had the opposite problem for a long time, until I realised the OE
signal _into_ the ENBFFs has to be active low (mine was active high).
Maybe changing the polarity of your OE signal would prevent this
"optimisation"? 

Or, check if your tool chain supports "dont_touch" or "preserve_signal"
attributes on specific signals?

- Brian


Article: 40744
Subject: Re: where to start with constraining..
From: Ray Andraka <ray@andraka.com>
Date: Thu, 14 Mar 2002 16:43:20 GMT
Links: << >>  << T >>  << A >>
The old 4000 series stuff seemed to be pretty conservatively timed.  It was not
unusual to be able to clock something at 50% faster than the timing report
stated in the lab under ideal conditions, however I would not put such an
overclocked design into production unless you like spending your career fixing
production problems.  The numbers reported by the timing analysis are worst
case over voltage, temperature and process.  Chances are in the lab you will
not hit the worst case on any of the three much less all of the three.

For the timing, run the static timing analyzer and set it to report paths
failing timing.  Those are sorted so that the worst paths show up first.  From
that report, you'll be able to see what CLBs the worst paths go through, so
you'll have an idea where to look.

Josh Pfrimmer wrote:

> Hi all,
>     For a class (calmer down.. I'm not looking for answers to my homework)
> I've designed an 8-bit CPU.  Fifteen instructions, pre-fetching, branch
> predicting, load-store, etc. etc.  So far, it cruises along at ~50MHz on an
> XC4010XL-3.  This is better than what's necessary for the class, but I'd
> like to push it further.
>
>     Now that I have a ballpark speed, what's the best way to get started
> specifying constraints in a .ucf?  How do I get the (old, old, old)
> Foundation 3.1 software in the labs to tell me what the worst paths are, so
> that I can start picking out false paths, etc.
>
>     Incidentally, the post-layout timing report gives a figure of ~20MHz fo
> the maximum clock speed, but on the board, I'm well over 45.. is this
> typical?  How can I get more accurate numbers out of the reports?
>
> Thanks for your attention
> JoshP


Article: 40745
Subject: Re: Proto boards for labs
From: Christopher.Saunter@durham.ac.uk (Christopher Saunter)
Date: Thu, 14 Mar 2002 16:46:19 +0000 (UTC)
Links: << >>  << T >>  << A >>
Nitin Chandrachoodan (nitin@eng.umd.edu) wrote:

: It seems that the Xilinx Spartan chips can be used with the 
: Webpack which is free, and they also have very large 
: capacities (100k-200k for XC2S200). Two boards I have found 
: so far with these are

: 1. XESS XSA-100 (100k gate, $279) from xess.com

: 2. Digilab XLA 2 (XC2S200 Spartan II - 200k gates, $99) from 
: digilentinc.com
There is another, similar $99 board out there, new with a spartan
2E, otherwise not to different.
http://www.nuhorizons.com/EngineeringServices/SpartanIIEBoard.html

: The digilent board seems to have better features at a lower 
: price, but XESS appears more popular overall. Are there any 
: other compelling reasons for using one or the other?

The external SDRAM on the XESS board could be nice.  There is a core
availible on their website to interface this to FPGA designs.

If you are demonstrating DSP stuff, this could be usefully, as you
can store a large ammount of data to be processed on the board in
advance, then the people using the boards can have the satisfaction
of seeing just how fast they can do neat things....  Combine that
with the primitive onboard VGA (theres some good examples out there
on generating VGA timings in FPGAs with VHDL / Schematic) and you
might be surprised at what comes out of the labs...

On the other hand, there is more embeded 'BlockRAM' on the larger
SpartanIIE chip on the Digilent board, so I guess the answer depends
on the size of the data chunks you want to process vs. the number of
chunks.

The XESS XSV Boards are probably very good for DSP/FPGA classes,
having audio in and out, composite video in and VGA out
functionality - XESS links to some examples of their usage from the
website.  There a bit pricey though.  

---
cds


Article: 40746
Subject: Re: Synthesis tools comparison?
From: "S. Ramirez" <sramirez@cfl.rr.com>
Date: Thu, 14 Mar 2002 17:53:36 GMT
Links: << >>  << T >>  << A >>
"Tim" <tim@rockylogic.com.nooospam.com> wrote in message
news:1016124118.3610.0.nnrp-12.9e9832fa@news.demon.co.uk...
> S. Ramirez wrote
>
> >    You will be hard pressed to find these comparisons, since the license
> > agreements of these vendors say something close to "Licensee shall not
..
> > disclose the results of any benchmarking of the SOFTWARE, or use such
> > results for its own competing software development activities, without
the
> > prior written permission of Blah Blah."
>
> You could fall back on the legal principle that "common abuse is not
> actionable".  From what we hear about some of the tools, common abuse
> would be a fair summary :-)

I could, but I don't have an attorney right now and would rather not get one
anyhow to experiment legally.  If you can spare an extra $100K-200K, please
send it along and let's try it! :)
Simon Ramirez, Consultant
Synchronous Design, Inc.
Oviedo, FL  USA



Article: 40747
Subject: Re: Proto boards for labs
From: "Noddy" <g9731642@campus.ru.ac.za>
Date: Thu, 14 Mar 2002 20:25:26 +0200
Links: << >>  << T >>  << A >>
Hi,

Haven't really looked at the DigiLab XLA 2 with the spartan II on, but we
got the Digilab XLA some time ago, with Spartan XL on, and I didn't like it
that much. I found that too many peripherals such as LED's and different BNC
ports took up valuable I/O pin connections. On top of that, they had
engineered there own programming dongle on to the board which didn't like
the Xilinx Foundation programmers (Hardware Debugger and JTAG) at all.
Instead, had to use some little utility I found on there website to program.

I would recomend having a look at the Burch Ed. board (www.burched.com). I
think the price is about $200 (AUS), has a Spartan II 200k gates (XC2S200)
and easy access to all the pins. There is a picture of the board on their
webpage. I have been using it for the past 9 months and it has worked
perfectly as a prototyping board. I am a university student and learnt to
work with it by myself.

Adrian



> Hello,
>
> We are trying to decide on boards for use in a course in the
> Electrical Engg. department at our university. The
> requirements we have come up with so far are as follows:
>
> 1. Easy to get started and working with it.
> 2. Relatively inexpensive
> 3. Can be used to implement reasonably large designs,
> especially DSP filters, FFTs etc. Here  we would like to
> implement the designs without having to go to too much low-
> level optimization (bit-serial implementations etc.) unless
> absolutely necessary, as this would change the focus from
> learning about FPGA implementation to low-level design.
> 4. Can be used with existing tools (we have Synopsys DC and
> FPGA compiler II, so we would prefer not to have to purchase
> Foundation etc. unless necessary)
>
> It seems that the Xilinx Spartan chips can be used with the
> Webpack which is free, and they also have very large
> capacities (100k-200k for XC2S200). Two boards I have found
> so far with these are
>
> 1. XESS XSA-100 (100k gate, $279) from xess.com
> 2. Digilab XLA 2 (XC2S200 Spartan II - 200k gates, $99) from
> digilentinc.com
>
> The digilent board seems to have better features at a lower
> price, but XESS appears more popular overall. Are there any
> other compelling reasons for using one or the other?
>
> Alternately, are there any other boards that are especially
> suited to classroom use? Apart from basic use in a classroom
> setting, we are also interested in boards that can be used
> for rapid reconfiguration etc., for use in research into
> reconfigurable computing techniques.
>
> Any replies much appreciated.
>
> Thank you,
>
> Nitin
>
>



Article: 40748
Subject: WTB: Coolrunner
From: "Tom" <yhocrash@yahoo.com>
Date: Thu, 14 Mar 2002 12:27:32 -0600
Links: << >>  << T >>  << A >>
Hi All,
Apologies if this is too OT.
I need a few parts for a project
Easy enough to source, but thought someone
might have excess they'd like to be rid of.

Req. 6 or so, any of the following:
xcr3128xl-12-TQ144
xcr3256xl-12-TQ144
xcr3384xl-12-TQ144
Sockets would be nice too.

BTW: anyone care to recommend a fav source
 for small orders like this ?

Many thx,
Tom
Private replies to:
Buyer@REMOVEsomeisp.net
Pls remove REMOVE.



Article: 40749
Subject: Re: Difference between Virtex-II(E) und Virtex-E
From: "Falk Brunner" <Falk.Brunner@gmx.de>
Date: Thu, 14 Mar 2002 19:38:15 +0100
Links: << >>  << T >>  << A >>
"Austin Lesea" <austin.lesea@xilinx.com> schrieb im Newsbeitrag
news:3C90C799.E64C72C2@xilinx.com...

> The software got twice as fast, and many more powerful features were
> added to make designs easier.  The new power and speed of Virtex II

Please, dont speakt with this marketing attitude. The first release of ISE
4.1 was buggy like HELL, and many bugs very much looked like software guy
"wasting" no time with testing. . . . .

--
MfG
Falk







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
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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