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 2825

Article: 2825
Subject: Re: PIC16C71 CORE for XC4000 ?
From: Les Hughes <L.J.Hughes@gre.ac.uk>
Date: Tue, 13 Feb 1996 11:14:42 +0000
Links: << >>  << T >>  << A >>
Dorian Nawrath wrote:
> 
> Gerrit Telkamp (telkamp@eis.cs.tu-bs.de) wrote:
> : Hello,
> : where can I get a PIC16C71-core (schematic or VHDL) for a XILINX XC4000 design ?
> : Thank you for every comment,
> : Gerrit.
> 
> I'm interested too,
> Dorian

Me too.....

-- 
                   Les Hughes - Applications Group 
 Computing Services           | Phone: +44 (0)181 331 8390 / 8566
 University of Greenwich      |
 Woolwich,  London SE18 6PF   | E-Mail:   L.J.Hughes@greenwich.ac.uk


Article: 2826
Subject: ARM-based ASICs
From: Ian.Page@comlab.ox.ac.uk (Ian Page)
Date: Tue, 13 Feb 96 15:39:48 GMT
Links: << >>  << T >>  << A >>
The Hardware Compilation Group at Oxford University is pleased to
announce the formal start of a new collaboration between ourselves,
Advanced RISC Machines (ARM), and European Silicon Structures (ES2). 

The `Aspire' project intends to use Oxford's hardware/software
co-design tools and expertise to offer a design flow into ARM-based
ASICs starting from system descriptions in the form of C-like
programs. The flow will also support a rapid prototyping phase on a
new ARM-based reconfigurable computer.

There is now a vacancy on this project the details of which have been
posted on uk.jobs.offered. Alternatively the job description is at:
    http://www.comlab.ox.ac.uk/oucl/users/ian.page/hwcomp/aspire/aspire.html


==============================================================================
  Ian Page,  F.I.E.E.,  University Lecturer, 
  Head of Hardware Compilation Research Group,
  Oxford University Computing Laboratory,
  Wolfson Building, Parks Road, Oxford OX1 3QD, U.K.

  Tel: (+44/0) 1865 273853 direct, 273840 sec, 0378 058921 mobile, 273839 fax.

  www:  http://www.comlab.ox.ac.uk/oucl/people/ian.page.html
==============================================================================



Article: 2827
Subject: Re: FPGA density
From: Eric Ryherd <eric@vautomation.com>
Date: 13 Feb 1996 16:12:05 GMT
Links: << >>  << T >>  << A >>
baten@hermes1.econ.uni-hamburg.de (Miranda Baten) wrote:
>Dear Sir
>
>if today a best FPGA have some 100 k gater density
>and some 50 % duty,is it possibly to make  from
>one FPGA chip  R3000/r3010 or R3081 MP ?
>
>I would very pleased for some info .Please
>post me.

Very unlikely... Even the new high density FPGAs must be derated
in the gate capactiy when designing real circuits. In general, FPGAs
have way more DFFs than you can use. And the vendors compute their
"gate capacities" by assuming you will use ALL of the DFFs in the FPGA.

On the otherhand... THey are getting bigger all the time. 8 bit
micros fit quite nicely in 13,000 gate FPGAs (then synthesize to <4,000
ASIC gates) which are available today.

The Xilinx XC8100 family is real nice... check it out... It quotes
very nearly the same number of gates as an ASIC!

-- 
Eric Ryherd                eric@vautomation.com  
VAutomation Inc.           Synthesizable HDL Cores 
20 Trafalgar Square        http://www.vautomation.com
Suite 443 Nashua NH 03063  (603) 882-2282 FAX:882-1587




Article: 2828
Subject: FPGA software for <$1000
From: moby@kcbbs.gen.nz (Mike Diack)
Date: 13 Feb 96 18:40:04 GMT
Links: << >>  << T >>  << A >>
When i posed this question, I got only one reply - from an Altera person
who said there was an entry level package that included a programmer,
schematic entry and route software for $500. He said it supported the
lower end devices. This sounds ideal, but Altera are not represented
here.
2 Questions :
What devices does it support
What is it called (& where can it be got in the SF Bay area)
M


Article: 2829
Subject: Re: Xilinx is NOT specified MINIMUM delay -- is it right??
From: Ray Andraka <randraka@ids.net>
Date: 13 Feb 1996 18:47:22 GMT
Links: << >>  << T >>  << A >>
Tony Clark <tonyc@perth.DIALix.oz.au> wrote:
>
> On Sat, 10 Feb 1996, HIKIMA Toshio wrote:
> 
> > Hi FPGAers,
> > 
> > I am using Xilinx XC4013. By the databook it is not specified
> > minimum delay.
> > My rep said, " there is no specificaton about minimum delay". Is it right?
> > 
> > If true, how can I design DRAM I/F? DRAM's spec has many complex constraient
> > so that it should be used minimum delay value.
> > 
> > Thanks,
> > --
> > T.Hikima
> > 
> > 

The right way to do this is with a SYNCHRONOUS design clocked at a 
multiple of the bus speed.  To get the high performance you will need 
for the DRAM controller you will have to observe high performance 
design rules and will likely need to floorplan at least the I/O paths 
to control the delays.  It is really not that difficult to do, it just 
takes a little experience with the parts and the tools.

As far as specifying minimum delays goes, there are very few logic 
devices that have these specified, and for good reason.  If you are 
depending on min delays, what's going to happen when the parts are 
improved?  You guessed it, your design that depended on min delays 
doesn't work anymore.  Its a bitch going back into a design you 
haven't touched in a few years to chase down a problem that 
'mysteriously' appears in production.  If you're lucky, you've left 
the company by then and some other poor SOB has to figure it out.
In the words of Raytheon's Bob McGurrin: "Speed Kills".  Design for 
it now so you don't become the next victim.

-Ray Andraka
Chairman, the Andraka Consulting Group
401/884-7930   FAX 401/884-7950
mailto:randraka@ids.net
http://www.ids.net/~randraka/
 
The Andraka Consulting Group is a digital hardware design firm 
specializing in high performance FPGA designs.  Services include 
complete design, development, simulation, and integration of these 
devices and the surrounding circuits.  We also evaluate,troubleshoot, 
and improve existing designs. Please call or write for a free 
brochure.





Article: 2830
Subject: Re: Xilinx FPGA's with Mentor Tools?
From: Lance Gin <c43lyg@dso.hac.com>
Date: 13 Feb 1996 20:27:32 GMT
Links: << >>  << T >>  << A >>
Les Hughes <L.J.Hughes@gre.ac.uk> wrote: 
>Hi,
>We teach digital systems using the Mentor / Xilinx s/ware. This consists 
>of a basic schematic capture - simulation - bit file production - 
>download to a test rig using xchecker, exercise followed by
>similar tasks using VHDL - QuickSim - Autologic - XACT - XChecker
>As for QuickVHDL and A/logic II, this was only released for our 
>OS last month so I haven't installed it yet.
>
>We have a direct support agreement with MGC and I have found 
>there help excellent. Our Xilinx support is through an academic 
>scheme and is not as good.
>
>Just for the record, we're using the A3F release and XACT 5.1.1 on 
>Sun Solaris 2.5. Generally, once bashed into shape by my mate vi ;-) 
>the system works well. 

hi les,

thanks much for your input ... is XCHECKER really available to you
under XACT 5.1.1 on solaris? we were told by our xilinx rep last week
that this debug capability is not available under XACT 5.1 on *any* 
unix platform (we're an HP house) and that its only available under
the new XACT 6.0 release for PC's (windows). we're about to embark
on our first xilinx design, a large XC4025E, so i'm interested in 
running XACT on unix where we have adequate cpu/memory/disk resources.
however, we anticipate a feature like XCHECKER will be very useful,
indeed, enough to sway our platform decision. perhaps XCHECKER has
been unbundled for sale in the UK?

based on responses to my posting, xilinx and mentor tools appear to 
be popular teaching vehicles. i tend to be methodology-oriented and 
was wondering if you might have any class/lab notes describing your 
design process which you'd be willing to share? i suppose this would
only be feasible if your docs were in electronic form.

thanks again for your comments,

-- 
____________________________________________________________________________

Lance Gin                                              "off the keyboard
Delco Systems - GM Hughes Electronics                   over the bridge,
OFC: 805.961.7737  FAX: 805.961.7329                    through the gateway,
C43LYG@dso.hac.com                                      nothing but NET!"
____________________________________________________________________________




Article: 2831
Subject: Re: Help: Xilinx behavior if Power down
From: paulfr@ix.netcom.com(* )
Date: 14 Feb 1996 03:37:41 GMT
Links: << >>  << T >>  << A >>
In <DMI1r0.GyB@maz-hh.de> pwu@maz-hh.de (Peter Wurbs) writes: 
>
>
>Hi Xilinx-Freaks,
>
>I use an output of a XC3195A as a Chip-Select for a SRAM.
>The CS-Signal has a pullup to a battery-buffered voltage to
>maintain RAM data if the power is down. The FPGA is supplied
>by the unbuffered power supply.
>
>It is the normal behavior of the IOB, that it is high impedance
>if the power voltage is below a certain level.
>But I could measure that the CS-signal is pulled to Low by the
>FPGA if the power voltage is near 0.8V. For less than 0.8 V it is
>o.k. again.
>But the precondition for the pullup to Vbatt is, that the FPGA output
>is high impedance over the full range of power supply.
>
>
>
>                        | Vbatt (1.8V if power down)
>                        |
>                        |
>                        |
>                        -
>                       | |
>                       | |
>                       |_|
>|---------------|       |           |------|
>|               |       |       CS  |      |
>|   FPGA        |-------------------| RAM  |
>|               |                   |      |
>|---------------|                   |------|
>       |                               |
>       |                               |  
>      VCC                             Vbatt
>
>
>Is this behavior a property of Xilinx-FPGA's ?
>Or is it the problem, that I pull the output to Vbatt while VCC is low
?
>Can I avoid this problem ?
>
>Thanks in advance for your help .
>
>Bye,
>
>Peter.
>
>---------------------------------------------------------
>Peter Wurbs (MAZ Hamburg GmbH, Dep. Broadband Communication)
>Phone:  ++40-76629 1771
>Fax:    ++40-76629 199
>e-Mail: pwu@maz-hh.de
>---------------------------------------------------------
>
>

Peter,

We had a problem similar to this ... except it was with EEProm.
We were getting data corruption of our calibration tables 
during power up/down.

First, a small cap combined with the 5 nH/inch lead inductance
can filter out spikes on the CS line.

Next, using a power supervisor such as the TI7705B will give
you a signal to gate the CS line off when power dips below
10 percent of nominal voltage. Be sure to use the R_ output
so the default condition is RESET. 

Finally, the FPGA will be more predictable if you use the
supervisor output to RESET the FPGA.  This will give several
hundred mSecs delay before the following recover/restart. 

The power on/off sequence gets a bit involved but you may
have to go through it.  We did. And EEProms have software 
data lockout protection built in along with some hardware
protections.

Paul Freda






Article: 2832
Subject: Crosspoint Solutions FPGA
From: SnowBin <Snowbin@popd.ix.netcom.com>
Date: 14 Feb 1996 08:11:29 GMT
Links: << >>  << T >>  << A >>
     Has anyone out there used FPGA devices from Crosspoint Solutions?  
I'd would like to know if there are things that I must be aware of.

     Due to the fine grain technology that Crosspoint has, I'd like to 
use top down tools to target Crosspoint FPGAs. I am most concerned in the 
layout and routing phases of the design process--what's the chance for 
first time success, the accuracy of timing parameters from back 
annotation etc... 

     Another issue that concerns me is that I'm using COMPASS synthesis 
tools and COMPASS did not have Crosspoint's library.  I heard a rumor 
that COMPASS is coming out with a synthesis library to support 
Crosspoint's FPGAs.  Can anyone confirm this? Is anyone using COMPASS 
tools to target Crosspoint? Any major problems using COMPASS tools with 
Crosspoint FPGAs?

     I'll certainly appreciate your comments/suggestions/viewpoints on 
the above.

--Bin



Article: 2833
Subject: Re: Xilinx is NOT specified MINIMUM delay -
From: pwu@maz-hh.de (Peter Wurbs)
Date: Wed, 14 Feb 1996 08:17:07 GMT
Links: << >>  << T >>  << A >>
In article 100000@perth.DIALix.oz.au, Tony Clark <tonyc@perth.DIALix.oz.au> () writes:
>On Sat, 10 Feb 1996, HIKIMA Toshio wrote:
>
>> Hi FPGAers,
>> 
>> I am using Xilinx XC4013. By the databook it is not specified
>> minimum delay.
>> My rep said, " there is no specificaton about minimum delay". Is it right?
>> 
>> If true, how can I design DRAM I/F? DRAM's spec has many complex constraient
>> so that it should be used minimum delay value.
>> 
>> Thanks,
>> --
>> T.Hikima
>> 
>> 
>You really need to simulate after working out the delays due to routing 
>etc.  I'm not a Xilinx user yet, but thats the impression I get from 
>looking at the Databook...
>


Even the post layout simulation does not include Min-Delays.
If you generate a SDF-File from the LCA-File it only includes
Max-Delays.


Bye,

Peter.

---------------------------------------------------------
Peter Wurbs (MAZ Hamburg GmbH, Dep. Broadband Communication)
e-Mail: pwu@maz-hh.de
---------------------------------------------------------



Article: 2834
Subject: Re: ARM-based ASICs
From: billms@nixon.icsl.ucla.edu (Bill Mangione-Smith)
Date: Wed, 14 Feb 1996 16:43:11 GMT
Links: << >>  << T >>  << A >>
In article <9602131539.AA06279@peridot.comlab.ox.ac.uk> Ian.Page@comlab.ox.ac.uk (Ian Page) writes:

   The `Aspire' project intends to use Oxford's hardware/software
   co-design tools and expertise to offer a design flow into ARM-based
   ASICs starting from system descriptions in the form of C-like
   programs. The flow will also support a rapid prototyping phase on a
   new ARM-based reconfigurable computer.

Ian, can you give some more information on the reconfigurable computer?
I wasn't able to find a dns entry for your web page.

Bill


Article: 2835
Subject: Re: Xilinx FPGA's with Mentor Tools?
From: Lance Gin <c43lyg@dso.hac.com>
Date: 14 Feb 1996 21:02:09 GMT
Links: << >>  << T >>  << A >>
john,

you pose some interesting issues, and of course, i have questions ...

>We are using Mentor's Top down tools to target Xilinx devices using XACT and NeoCAD. 

what versions of mentor and XACT are/were you running? is XACT running
on HP as well?

>1/ XACT did not install correctly on our HP-UX host

was this problem recognized by xilinx? has it been fixed?

>2/ Mentor creates ModelName_s; Xilinx does not like _s. We couldn't create a .xnf 
>file for months

was this problem recognized by xilinx/mentor? has it been fixed?

>3/ Xilinx recommend XBlox. XBlocks models don't have timing characteristics.

are you referring to models for use by the XACT XDELAY static TA tool?
or perhaps you're referring to QuickPath? does xilinx have any plans to
supply the models you need?

>4/ TimSim8 (backannotation of routed design timing to "Schematic") does not modify 
>the source schematic (EDDM database) via a viewpoint. It generates another design 
>database in a separate directory. This presents a rather serious configuration 
>control problem.

heard something about this. understand VHDL VITAL libraries are not available
with the mentor/XACT kit :-( therefore, must be a problem with schematic design
method. possibly related to XACT performing an additional level of optimization
on the XNF created by mentor? if true, the new XNF may not have 1:1 structural
correspondance to your original schematics. thus the need to generate a new
set of schematics. right? or off-base?

>5/ An EDIF read licence is required, at the unreasonable price of ~=A32000.

how is this used in the mentor/XACT flow?

>6/ XNFBA still does not work.

we're new to the xilinx world. what's XNFBA?

>Mentor tools work well together and deliver good results. Xilinx tools are good and 
>deliver well. The integration, intertool communications and design control are 
>absolutely lousy. It is particularly annoying given that Mentor's Framework 
>Initiative promised to solve these problems! (Oh yes, its Xilinx's problem for not 
>integrating......)

a big concern of mine, esp if i start having problems. here in the US,
i have to deal with *many* interfaces:

1. a distributor who sells the mentor XACT kit. mentor is not allowed to sell
   the kit to me (so where is mentor's $$ motivation to support my new xilinx
   flow?). the distributor has an FAE, but he's experienced with viewlogic on
   PC's, not mentor.

2. a xilinx FAE. he too, is not that familiar with mentor (but good on other
   stuff).

3. the xilinx tech support hotline. i asked if xilinx supported VITAL. they
   said "yes" and cited their preferred 3rd-party partners. i talked to
   mentor. no VITAL. good advice when it comes to device questions though.

4. our local mentor FAE. finally, someone who might be able to help. he's a
   sharp guy, but overworked and supporting sales. (see also 1. above).

5. mentor tech support. ok, haven't tried them yet. my experience tells me
   they're more helpful in detailed, tool-specific questions as opposed to
   multi-vendor, multi-tool flow problems. we'll see.

as a side note, i wonder how this issue will be handled once ANTARES gets
going. imagine a flow with mentor, model tech, exemplar, meta, co-soft,
xilinx, ... how many people will i need to talk to then?

>Neither Mentor nor Xilinx could provide us with an overall process diagram and 
>process description.

hope this has changed. actel documents their mentor flow in their docs
(although i found actel hotline's mentor experience to be lacking as well -
seems they have lots of viewlogic expereience).

>We have, somehow, managed to knife and fork a usable process.

hope this changes too. as fpga's get bigger, design flows will look more
like asic flows - right up mentor's alley.

last question: is the XChecker debug utility avail in your XACT kit
(apparently a newer utility, may not be in your XACT version on your platform).

"opinions expressed are my own and not my company's"

-- 
____________________________________________________________________________

Lance Gin                                              "off the keyboard
Delco Systems - GM Hughes Electronics                   over the bridge,
OFC: 805.961.7737  FAX: 805.961.7329                    through the gateway,
C43LYG@dso.hac.com                                      nothing but NET!"
____________________________________________________________________________




Article: 2836
Subject: re-routing with locked pinout
From: Joe Troxel <jtroxel@ball.com>
Date: 14 Feb 1996 23:26:25 GMT
Links: << >>  << T >>  << A >>
i recently received a pile of literature from a local xilinx rep which
goes to great length to point out the deficiencies in altera fpga
routing resources.  while i can easily see the limitations on paper,
i'm curious as to what people have experienced in reality.  

so how about it?  anyone care to share their real life success stories
and/or tales of woe relating to altera and xilinx routing issues?

later.  joe.



Article: 2837
Subject: Re: Xilinx FPGA's with Mentor Tools?
From: bobarker@ix.netcom.com(Robert Barker )
Date: 15 Feb 1996 04:52:08 GMT
Links: << >>  << T >>  << A >>
In <4ftikh$sig@hacgate2.hac.com> Lance Gin <c43lyg@dso.hac.com> writes:
In answer to issue 5 below, Antares is up and running. The flow you are
looking for is available now. The Exemplar Galileo environment is a
fully, interactive environment with synthesis, back annotation, timing
analysis, and verification. The Verification is provided by Model
Technology. The flow is fully VITAL compliant and Exemplar maintains
the VITAL libraries. For Xilinx Exemplar supports both EDIF and XNF
netlist formats. They even extract timing from the Xilinx netlist after
PPR and overlay it to the VITAL models. The whole flow makes use of the
VITAL acceleration capability of the MTI simulator. The flow is
identical on PCs and Workstations. The entire flow is supported by
Exemplar, except for PPR, but they are also quite knowledgeable about
the vendor place and route tools.

Bob
>
>john,
>
>you pose some interesting issues, and of course, i have questions ...
>
>>We are using Mentor's Top down tools to target Xilinx devices using
XACT and NeoCAD. 
>
>what versions of mentor and XACT are/were you running? is XACT running
>on HP as well?
>
>>1/ XACT did not install correctly on our HP-UX host
>
>was this problem recognized by xilinx? has it been fixed?
>
>>2/ Mentor creates ModelName_s; Xilinx does not like _s. We couldn't
create a .xnf 
>>file for months
>
>was this problem recognized by xilinx/mentor? has it been fixed?
>
>>3/ Xilinx recommend XBlox. XBlocks models don't have timing
characteristics.
>
>are you referring to models for use by the XACT XDELAY static TA tool?
>or perhaps you're referring to QuickPath? does xilinx have any plans
to
>supply the models you need?
>
>>4/ TimSim8 (backannotation of routed design timing to "Schematic")
does not modify 
>>the source schematic (EDDM database) via a viewpoint. It generates
another design 
>>database in a separate directory. This presents a rather serious
configuration 
>>control problem.
>
>heard something about this. understand VHDL VITAL libraries are not
available
>with the mentor/XACT kit :-( therefore, must be a problem with
schematic design
>method. possibly related to XACT performing an additional level of
optimization
>on the XNF created by mentor? if true, the new XNF may not have 1:1
structural
>correspondance to your original schematics. thus the need to generate
a new
>set of schematics. right? or off-base?
>
>>5/ An EDIF read licence is required, at the unreasonable price of
~=A32000.
>
>how is this used in the mentor/XACT flow?
>
>>6/ XNFBA still does not work.
>
>we're new to the xilinx world. what's XNFBA?
>
>>Mentor tools work well together and deliver good results. Xilinx
tools are good and 
>>deliver well. The integration, intertool communications and design
control are 
>>absolutely lousy. It is particularly annoying given that Mentor's
Framework 
>>Initiative promised to solve these problems! (Oh yes, its Xilinx's
problem for not 
>>integrating......)
>
>a big concern of mine, esp if i start having problems. here in the US,
>i have to deal with *many* interfaces:
>
>1. a distributor who sells the mentor XACT kit. mentor is not allowed
to sell
>   the kit to me (so where is mentor's $$ motivation to support my new
xilinx
>   flow?). the distributor has an FAE, but he's experienced with
viewlogic on
>   PC's, not mentor.
>
>2. a xilinx FAE. he too, is not that familiar with mentor (but good on
other
>   stuff).
>
>3. the xilinx tech support hotline. i asked if xilinx supported VITAL.
they
>   said "yes" and cited their preferred 3rd-party partners. i talked
to
>   mentor. no VITAL. good advice when it comes to device questions
though.
>
>4. our local mentor FAE. finally, someone who might be able to help.
he's a
>   sharp guy, but overworked and supporting sales. (see also 1.
above).
>
>5. mentor tech support. ok, haven't tried them yet. my experience
tells me
>   they're more helpful in detailed, tool-specific questions as
opposed to
>   multi-vendor, multi-tool flow problems. we'll see.
>
>as a side note, i wonder how this issue will be handled once ANTARES
gets
>going. imagine a flow with mentor, model tech, exemplar, meta,
co-soft,
>xilinx, ... how many people will i need to talk to then?
>
>>Neither Mentor nor Xilinx could provide us with an overall process
diagram and 
>>process description.
>
>hope this has changed. actel documents their mentor flow in their docs
>(although i found actel hotline's mentor experience to be lacking as
well -
>seems they have lots of viewlogic expereience).
>
>>We have, somehow, managed to knife and fork a usable process.
>
>hope this changes too. as fpga's get bigger, design flows will look
more
>like asic flows - right up mentor's alley.
>
>last question: is the XChecker debug utility avail in your XACT kit
>(apparently a newer utility, may not be in your XACT version on your
platform).
>
>"opinions expressed are my own and not my company's"
>
>-- 
>______________________________________________________________________
_____
>
>Lance Gin                                              "off the
keyboard
>Delco Systems - GM Hughes Electronics                   over the
bridge,
>OFC: 805.961.7737  FAX: 805.961.7329                    through the
gateway,
>C43LYG@dso.hac.com                                      nothing but
NET!"
>______________________________________________________________________
_____
>
>




Article: 2838
Subject: Re: re-routing with locked pinout
From: Andreas Hofmann <ftdanho@ftd.ericsson.se>
Date: Thu, 15 Feb 1996 09:02:44 +0100
Links: << >>  << T >>  << A >>
Rerouting with locked pinout is possible with ALTERA if you have done
the
pinout well. That means that you have to do something like a
floorplanning to
determine your pinout. This is even reasonable if you don't want to
change
the pinout.
 
To do this floorplanning, you just have to use a copy of the footprint
of your
choosen device and write down the ROW/COLUMN information you will find
in the
device & adapter section of the ALTERA online help. This should help to
assign pins
belonging to a logical block close to each other. Tis is the reason why
I call it
floorplanning, because you define the area where the logic block is
placed.

Even, it is a good practice to put the outputs on the rows and the
inputs on
columns.
-- 
     _/ _/ _/
    _/ _/ _/  EED/E/X/A/ Andreas Hofmann          Phone:
+49-5121-707-378
   _/ _/ _/   Ericsson Eurolab Deutschland GmbH   FAX  :
+49-5121-707-333
  _/ _/ _/    Daimlerring 9                       Memo : EED.EEDANHO
 _/ _/ _/     31135 Hildesheim - Germany          Email:
ftdanho@ftd.ericsson.se
_/ _/ _/


Article: 2839
Subject: Abel mode for Emacs?
From: holmes@chrysal.com (Christopher G. Holmes)
Date: Thu, 15 Feb 1996 15:26:07 GMT
Links: << >>  << T >>  << A >>

Anyone know if there's an Abel mode for Emacs?  Know where I can get it?

Chris

-- 
=============================================================================
Christopher G. Holmes			Me:		holmes@chrysal.com
Chrysalis Research Corporation		Inquiries:	info@chrysal.com
52 Domino Drive				Voice:		508-371-9115
Concord, MA 01742-2817			Fax:		508-371-9175

-- 
=============================================================================
Christopher G. Holmes			Me:		holmes@chrysal.com
Chrysalis Research Corporation		Inquiries:	info@chrysal.com
52 Domino Drive				Voice:		508-371-9115


Article: 2840
Subject: Re: New Reconfigurable Computing Threads.
From: fliptron@netcom.com (Philip Freidin)
Date: Thu, 15 Feb 1996 16:27:06 GMT
Links: << >>  << T >>  << A >>

In article <311bcc22.274923224@news.jf.intel.com> you write:
>>>Is floating point important or because current FPGAs don't have floating point
>>>structures do we just through up our hands and keep to integer math?
>Add an FPU co-processor slave, and save the FPGA real-estate for more
>important stuff.

Absolutely.

>
>> What will it take to get reconfigurable computing off the ground?
>The reconfigurable FPGA JAVA processor.  Say, what about modifying
>Phil Friedin's small RISC into a JAVA interpreter?

Where do I get a spec so I can start on this. (at least half serious).
Maybe just a sw interpretor running on the existing R16 would be fine.
Do Java interpretors tend to be big or small (i.e. lines of C).

>
>More importantly Steve, and the whole point of this followup, what
>kind of beer are you going to bring to FCCM'96 this year?  And what
>shall I bring down from Portland?  :-)
>
>--Richard Vireday

I have every intention of being there this year too. I will bring
some iced tea.

Philip Freidin.







Article: 2841
Subject: Re: re-routing with locked pinout
From: greg_peek@ccm.jf.intel.com (Greg Peek)
Date: Thu, 15 Feb 1996 16:56:09 GMT
Links: << >>  << T >>  << A >>
Andreas Hofmann <ftdanho@ftd.ericsson.se> wrote:

>Rerouting with locked pinout is possible with ALTERA if you have done the
>pinout well. That means that you have to do something like a
>floorplanning to
>determine your pinout. This is even reasonable if you don't want to
>change the pinout.

Using Adreas's methodology is also very important with Xilinx devices.
Floorplanning not only improves the ability to do rerouting, but it
also improves the routed timing of the device.

I have found that Actel devices seem to have more robust routing
resources than Xilinx and Altera.  The last time I used their tools,
Actel didn't even provide the user the ability to do floorplanning or
hand routing. Yet  I was still able to get >90% utilization with
multiple reroutes for small design changes.  I let the tool set the
pinout only the first time I routed.

Next time, I plan to let the tool show me where it thinks the pins
should go.  Then I will try adjusting the pinout slightly for a more
PCB-layout friendly pin ordering and see if the tools can reroute with
my pinout locked in.
Greg Peek                     |  The opinions expressed are
greg_peek@ccm.jf.intel.com    |  my own, who else would claim them?



Article: 2842
Subject: CFP: Hawaii Intl Conf: ENGINEERING COMPLEX COMPUTER SYSTEMS Track
From: Alberto Broggi <broggi@CE.UniPR.IT>
Date: Thu, 15 Feb 1996 18:32:55 +0100
Links: << >>  << T >>  << A >>


                                 CALL FOR PAPERS

                     * ENGINEERING COMPLEX COMPUTER SYSTEMS *

                                Thirtieth Annual
               HAWAII INTERNATIONAL CONFERENCE ON SYSTEMS SCIENCES 
                                   HICSS - 30
                        Maui, Hawaii, January 7-10, 1997


Papers are invited for the  Minitrack on  ENGINEERING COMPLEX COMPUTER SYSTEMS  
as  part  of  the  Advanced  Technology  track  at  the  Hawaii  International 
Conference on System Sciences (HICSS).


 1. PURPOSES

    Modern  computer  systems   and   applications   embody   many   different  
characteristics and properties that  are  currently  addressed,  studied,  and 
optimized independently. Nevertheless, although it is of basic  importance  to 
focus on these aspects independently, as a whole these  properties  feature  a 
complex interrelationship, and  thus  a  higher-level  view  of  the  complete 
project becomes mandatory.

    While perhaps some of the earlier computer  systems  could  be  described, 
designed and implemented with a particular focus on  one  objective  (such  as 
fault-tolerance or timeliness), or using a single method (such  as  Structured 
Programming),  it  is  very  questionable  whether  such  modern  and   future 
applications can be. Nowadays almost all electronic products are becoming more 
and more software based: complex computer systems are becoming common in  many 
sectors,  such  as  manufacturing,  communications,  defense,  transportation, 
aerospace, hazardous environments, energy, health  care,  etc.  These  systems 
feature a number of different characteristics (such as distributed processing, 
heterogeneous computational paradigms, high speed networks, novel bus systems,
or  special-purpose  hardware  enhancements  in  general)   and    performance 
requirements  (such  as  real-time  behavior,  fault   tolerance,    security, 
adaptability, development time and cost,  long life concerns).  The concurrent 
satisfaction of the systems requirements have a  considerable  impact  on  the 
hardware characteristics and vice-versa. The analysis of the complete project, 
as a whole, is a major point in the design of the computer system  itself  and
plays a basic role throughout the entire system life.       

    The  ECCS  Minitrack  will  bring  together  industrial,   academic,   and
government experts from  these  various  disciplines,  to  determine  how  the 
disciplines' problems  and  solution  techniques  interact  within  the  whole 
system.  Researchers, practitioners, tool developers and users, and technology 
transition experts are all welcome.
              

 2. ADDRESSED TOPICS:

    Papers are solicited on all major aspects of  ECCS  including  specifying, 
designing,  prototyping,  building,  testing,  operating,  maintaining,    and 
evolving of complex computer systems, including:

        *  Software engineering, re-engineering, reverse engineering
        *  Complex real-time architectures, tools, environments and languages
        *  AI and intelligent systems
        *  Database and data management
        *  Dependable real-time systems
        *  Virtual reality, multimedia, real-time imaging
        *  Algorithms, optimization and analysis
        *  Analytical techniques
        *  Megaprogramming, visual programming
        *  Performance estimation, prediction and optimization
        *  Prototyping and testing techniques
        *  Formal methods and formal specification techniques
        *  Hardware/software co-design
        *  Communications, networking, mobile computing
        *  Highly heterogeneous, distributed and parallel platforms
        *  Case studies and project reports


 3. MINITRACK COORDINATORS

    Alberto Broggi                       Alexander D. Stoyenko
      Dip. Ingegneria dell'Informazione    Real-Time Computing Laboratory, CIS
      Universita` di Parma                 New Jersey Institute of Technology
      I-43100 Parma, Italy                 Newark, New Jersey 07102 USA
      Fax: +39 - 521 905723                Fax: (201) 596-5777
      Email: broggi@CE.UniPR.IT            Email: alex@vulcan.njit.edu


    Papers should be submitted to:

                        Alberto Broggi
                          HICSS'97 ECCS Coordinator
                          Dipartimento di Ingegneria dell'Informazione
                          Universita` di Parma, Viale delle Scienze
                          I-43100 Parma, Italy


 4. FURTHER AND UP-TO-DATE INFORMATION:

    Further  information  about  the  ECCS  Minitrack  are  available  at  the
following WWW address: 
		   http://WWW.CE.UniPR.IT/hicss/eccs  




  * INSTRUCTIONS FOR SUBMITTING PAPERS:

1.  Submit 6 (six) copies of the  full  paper,  consisting  of  20 - 25  pages
    double-spaced including title  page,  abstract,  references  and  diagrams
    directly to the minitrack coordinator.

2.  Do not submit the paper to more than one  minitrack.    The  paper  should
    contain original material and not be  previously  published  or  currently
    submitted for consideration elsewhere.

3.  Each paper must have a tile page which includes the title,  full  name  of 
    all  authors,  and  their  complete  addresses  including  affiliation(s),
    telephone number(s) and e-mail address(es).

4.  The first page of the paper  should  include  the  title  and  a  300-word 
    abstract.



  * DEADLINES:

March 15, 1996:    Abstracts submitted to track coordinators for guidance and 
    indication of appropriate content. Authors unfamiliar with HICSS or those 
    who wish additional guidance are encouraged to contact any coordinator to 
    discuss potential papers.

June 1, 1996:      Full  papers  submitted  to  the  appropriate  track,   or 
    minitrack coordinator.

August 31, 1996:   Notification of accepted papers mailed to authors.

October 1, 1996:   Accepted manuscripts,  camera-ready ,  sent  to  minitrack 
    coordinators; one author from each paper  must  register  by  this  time.

November 15, 1996: All other registrations  must be  received.  Registrations 
    received after this deadline may not be accepted due to space limitation.


  
  * CONFERENCE PROCEEDINGS:

    The conference Proceedings are published and distributed by IEEE  Computer 
Society.




The ENGINEERING COMPLEX COMPUTER SYSTEMS Minitrack is  part  of  the  Advanced 
Technology.   For more information on the  Advanced Technology Track  contact:

Ralph H. Sprague, Jr.
 E-mail: sprague@hawaii.edu
 Voice: (808) 956-7082
 Fax: (808) 956-9889



  * OTHER CONFERENCE TRACKS

There are three other majors tracks in  the  conference:    Software,  Digital 
Documents, and Information Systems.  The Information Systems Track has several 
minitracks that focus  on  a  variety  of  research  topics  in  Collaboration 
Technology, Decision Support and Knowledge-Based Systems,  and  Organizational 
Systems and Technology.   For  more information on the  other  tracks,  please
contact:

Software Technology Track:
 Hesham El-Rewini                  rewini@unocss.unomaha.edu

Digital Documents Track:
 M. Stuart Lynn                    msylnn@ucop.edu

Information Systems Track:
 Ralph H. Sprague, Jr.             sprague@hawaii.edu
 Jay F. Nunamaker, Jr.             nunamaker@bpa.arizona.edu
 Eileen Dennis (Track Assistant)   edennis@uga.cc.uga.edu

The purpose of HICSS is to provide a  forum  for  the  interchange  of  ideas, 
research results, development activities, and applications among  academicians 
and practitioners in computer-based systems sciences.  The conference consists 
of tutorials, advanced seminars, presentations of accepted papers, open forum, 
tasks forces, and  plenary and distinguished guest lectures.   There is a high 
degree of interaction and discussion among the conference participants because 
the conference is conducted in a workshop-like setting.

For  more  information  on  the  conference,  please  contact  the  conference
coordinator:

Barbara Edelstein
College of Business Administration
University of Hawai'i
2404 Maile Way
Honolulu, HI 96822
Voice: (808) 956-3251
Fax: (808) 956-9685
E-mail: hicss@hawaii.edu

or visit the World Wide Web page: http://www.cba.hawaii.edu/hicss



Article: 2843
Subject: Re: Xilinx is NOT specified MINIMUM delay -- is it right??
From: Bill Clark <wclark@duat.clark.com>
Date: Thu, 15 Feb 1996 19:40:31 -0700
Links: << >>  << T >>  << A >>
On Sat, 10 Feb 1996, HIKIMA Toshio wrote:
 
 > Hi FPGAers,
 >
 > I am using Xilinx XC4013. By the databook it is not specified
 > minimum delay.
 > My rep said, " there is no specificaton about minimum delay". Is it right?
 >
 > If true, how can I design DRAM I/F? DRAM's spec has many complex onstraient
 > so that it should be used minimum delay value.

No minimums.

If you have an asynchronous design which really requires min delays, try 
asking your Xilinx representative if a Speeds File exists for a very fast, 
possibly unnanounced, possibly hardwired part. It may well be safe to assume 
that the typical or max delays of such a part represent practical 3-4 
sigma minimums of parts to be offered for 2-3 years. Clearly, you will have to 
know a bit about Xilinx's plans to have confidence in this!

Very likely Xilinx prefers not to specify minimums to allow freedom for 
process improvements -- just like every other manufacturer of other than 
consumer Si.
--
Bill Clark, wclark@clark.com, Clark Associates, Inc. +1303 444 1890


Article: 2844
Subject: Re: Xilinx is NOT specified MINIMUM delay -
From: tartis@world.std.com (Tad B Artis)
Date: Fri, 16 Feb 1996 05:16:33 GMT
Links: << >>  << T >>  << A >>
I spoke with a rep & they said that even if you order a slow speed part you 
may very well receive a faster one if it's in inventory.  They also said that
process changes occur rapidly and that Xilinx only attempts to meet the
max specs & doesn't care about min specs.

Peter Wurbs (pwu@maz-hh.de) wrote:
: In article 100000@perth.DIALix.oz.au, Tony Clark <tonyc@perth.DIALix.oz.au> () writes:
: >On Sat, 10 Feb 1996, HIKIMA Toshio wrote:
: >
: >> Hi FPGAers,
: >> 
: >> I am using Xilinx XC4013. By the databook it is not specified
: >> minimum delay.
: >> My rep said, " there is no specificaton about minimum delay". Is it right?
: >> 
: >> If true, how can I design DRAM I/F? DRAM's spec has many complex constraient
: >> so that it should be used minimum delay value.
: >> 
: >> Thanks,
: >> --
: >> T.Hikima
: >> 
: >> 
: >You really need to simulate after working out the delays due to routing 
: >etc.  I'm not a Xilinx user yet, but thats the impression I get from 
: >looking at the Databook...
: >


: Even the post layout simulation does not include Min-Delays.
: If you generate a SDF-File from the LCA-File it only includes
: Max-Delays.


: Bye,

: Peter.

: ---------------------------------------------------------
: Peter Wurbs (MAZ Hamburg GmbH, Dep. Broadband Communication)
: e-Mail: pwu@maz-hh.de
: ---------------------------------------------------------



Article: 2845
Subject: Re: New Reconfigurable Computing Threads.
From: gratz@ite.inf.tu-dresden.de (Achim Gratz)
Date: 16 Feb 1996 08:43:42 +0100
Links: << >>  << T >>  << A >>
>>>>> "Philip" == Philip Freidin <fliptron@netcom.com> writes:

    Philip> In article <311bcc22.274923224@news.jf.intel.com> you
    Philip> write:
    >> The reconfigurable FPGA JAVA processor.  Say, what about
    >> modifying Phil Friedin's small RISC into a JAVA interpreter?

    Philip> Where do I get a spec so I can start on this. (at least
    Philip> half serious).  Maybe just a sw interpretor running on the
    Philip> existing R16 would be fine.  Do Java interpretors tend to
    Philip> be big or small (i.e. lines of C).

You'll want a processor for the Java bytecode.  BTW, Sun's already
working on that (not FPGA), supposedly to put them into their internet
terminals.  Specs can be ordered from Sun; if you're only half
serious, so you don't want to license Java, the stuff on their
http://java.sun.com/ might do.



--
Achim Gratz

--+<[ It's the small pleasures that make life so miserable. ]>+--

E-Mail: gratz@ite.inf.tu-dresden.de
Phone:  +49 351 4575 - 325


Article: 2846
Subject: ASIC& full-cust versus FPGA-future 5 jears
From: baten@hermes1.econ.uni-hamburg.de (Miranda Baten)
Date: Fri, 16 Feb 1996 11:37:42 GMT
Links: << >>  << T >>  << A >>
Dear Sir

I have suspect that in future 5 jears
ASIC and full cust. would have 
full advantage in field MP or DSP 
implantation (more as 100 000 gater)

Alsow a standard 64-bit MP -MIPS 4660 
have today  a price smaller as FPGA (28 $ )

In GaAs field ASIC have not competition from FPGA.

Some comment wilcome .Please post me.



Article: 2847
Subject: BDD Help
From: gomez66@eucmax.sim.ucm.es
Date: 16 Feb 96 12:05:57 GMT
Links: << >>  << T >>  << A >>
Hi!



I am strongly interested to find anyone who can answer me to this question:

Is there a C function to convert binary decision trees to reduced ordered

binary decision diagrams ? (In other words, ROBDD, BDD, ...).



Please, if you know it, send me an e-mail. Thank you very much.

==========================================================================

Juan A. Gomez Pulido          email: jangomez@ba.unex.es

Departamento Informatica      fax:   +34-27-257203

Universidad Extremadura       tel:   +34-27-257264

10071  Caceres  Spain

==========================================================================






Article: 2848
Subject: Re: Altera Simulation
From: Dean Fitzgerald <deanf>
Date: 16 Feb 1996 17:08:48 GMT
Links: << >>  << T >>  << A >>
neal@ctd.comsat.com (Neal Becker) wrote:
>I am just trying simulation on Altera for the first time.  I want to
>initialize some internal registers.  Is Altera's simulator really not
>capable of performing this basic function?  I can't seem to find any
>way to do this!

Hopefully, you have used the global set/reset signal to initialize
your registers/flip-flops.  It's an external pin.  It will make
your simulation task more simple.



Article: 2849
Subject: Re: New Reconfigurable Computing Threads. -- Java machines
From: jsgray@ix.netcom.com(Jan Gray)
Date: 16 Feb 1996 18:16:19 GMT
Links: << >>  << T >>  << A >>
In <fliptronDMtsD6.HGq@netcom.com> fliptron@netcom.com (Philip Freidin)
writes: 

>>> What will it take to get reconfigurable computing off the ground?
>>The reconfigurable FPGA JAVA processor.  Say, what about modifying
>>Phil Friedin's small RISC into a JAVA interpreter?

(Well, an FPGA based Java machine might not be dynamically
reconfigured, so I'm not sure how this helps, except to glamourize
processor implementations in FPGAs.)

>Where do I get a spec so I can start on this. (at least half serious).
>Maybe just a sw interpretor running on the existing R16 would be fine.
>Do Java interpretors tend to be big or small (i.e. lines of C).

The Java VM spec is at //java.sun.com/doc/programmer.html/...  A good
Java microprocessor should have a 32-bit datapath.  XC4013Es not
XC4005s...

Java bytecodes are a cross between the Smalltalk-80 virtual machine
bytecodes and Microsoft C compiler pcode.  That is, 32-bit oriented,
stack oriented, with a locals frame, constant pool, generically typed
object instructions, plus several varieties of explicitly typed numeric
opcodes (e.g. explicit iadd vs. ladd vs. fadd etc.).  Requires a
runtime object system providing object typing and garbage collection.

A simple Java interpreter plus object system would be a few thousand
lines of code.  A good one would be much more sophisticated.

(Folks interested in implementing Java VMs should go read the last
third of "Smalltalk-80: The Language and Its Implementation", and most
of "Smalltalk-80: Bits of History, Words of Advice".  Plus look at
Deutsch and Schiffman's Smalltalk-80 interpreter, Self, SOAR, and
possibly the various hardware LISP implementations for ideas.  And
don't forget the ACM Architectural Support for Programming Languages
and Operating Systems conferences' proceedings!)

Many of the instructions are easy to do in hardware "at speed".  On the
other hand, many of the instructions, such as new, invokevirtual,
putfield, athrow, or floating point, are more involved, and for those
you would want to drop down to "microcode" to emulate.  Which could
mean that a microarchitecture with unaligned ifetch hardware, stack
orientation, and possibly type tags (I'm still trying to understand if
tags help) together with a fast underlying RISC datapath could be a
good start.  Or it might be exactly the wrong way to go!

The fundamental design issue is how sophisticated is your download-time
translation pass over the bytecodes, in order to canonicalize or
regularize them?

For example, a register file can emulate a frame + stack, so a
translator which tracks stack contents can translate
  "iload local #1"
  "iload local #2"
  "iadd"
  "istore local #1"
into
  "add r29, r29, r30".

Too little canonicalization, and your microarchitecture is too complex.
Too much, and congratulations, you have just written a Java-to-RISC
optimizing compiler on top of a simple pipelined RISC.

For Java, the SOAR approach seems pretty good.  Translate to a
RISC-like instruction set, probably once, at code download time.  (I'm
not sure there is much value to *dynamic* Java bytecode translation,
and therefore it probably doesn't justify much/any hardware assist.  A
little dynamic in-line self-modifying code might be apropriate,
however!)  Then add additional hardware to accelerate tag checks,
putfield GC space testing (e.g. for remembered sets), frame management,
or whatever else takes up the time.

I used to worship the Alto and Dorado architects.  Now FPGAs put their
tools (and more) into my hands, our hands.  Who needs ECL and microwire
when we have LUTs and PIPs?  Ha ha ha ha ha ha ha!

Jan Gray
Redmond, WA





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