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 17725

Article: 17725
Subject: Re: Virtex dev boards
From: Sukandar Kartadinata <sk@zkm.de>
Date: Fri, 27 Aug 1999 21:09:01 +0200
Links: << >>  << T >>  << A >>
Daryl Bradley wrote:

> have a look at www.xess.com
>
> THey have a virtex board coming soon with memory on it
> Haven't looked at the specs but maybe suitable?

yes, there's quite a lot on that board, although not quite enough memory.
will ask them if they can fit it with more

thanks,
Sukandar


Article: 17726
Subject: Re: Virtex dev boards
From: Sukandar Kartadinata <sk@zkm.de>
Date: Fri, 27 Aug 1999 21:09:56 +0200
Links: << >>  << T >>  << A >>
Bill Blyth wrote:

> The Virtex boards from Alpha Data have separate memory interfaces with
> SRAM and SSRAM. Size might be a problem. RC1000 has 4 x 512k x 32 SRAM.
> ADM-XRC has 4 x 128k(256,512) x 36.
> Try www.alphadata.co.uk.

yes, nice boards.
I'm in contact with them

thanks,
Sukandar


Article: 17727
Subject: Re: looking for image processing hardware
From: Sukandar Kartadinata <sk@zkm.de>
Date: Fri, 27 Aug 1999 21:14:37 +0200
Links: << >>  << T >>  << A >>


Jonathan Feifarek wrote:

> Signalware (www.signalware.com) makes several plug-in boards

hmm, not much on their website.....I could only find a C50 based board
which is slightly under-powered for our needs.

also www.signalware.com redirects automatically to
www.qwest.net/ipcommerce/SNI/
(now it's www.sni.net/dsp/index.htm)

> (I assume 24 bits mean 8 bits per
> color channel),

yes

thanks,
Sukandar


Article: 17728
Subject: Re: microcontroller vs FPGA
From: Joshua Lamorie <jpl@xiphos.ca>
Date: Fri, 27 Aug 1999 15:40:07 -0400
Links: << >>  << T >>  << A >>
Daniel Figuerola Estrada wrote:
> Has anyone worked with those two technologies and could give his opinion
> of them?

Why look at them as mutually exclusive?  We develop a controller board
(very small, PCMCIA form-factor) that has both a microcontroller AND an
FPGA.  This allows for some amazing flexibility and utility.  Especially
because of all sorts of neat devices already on the die on the uC.

We run the FPGA at twice the clock speed of the uC, and operate
networking, flash/memory interface, ADC, and many more things in there. 
We run our Fuzzy and PID control in the uC.. it's perfect.

If you look at straight logic, you can also get amazing performance for
cost, weight, and power.  Any comparisons are strictly dependant on a
specific application, but for many of our trade-off studies.  A logic
only node can beat an embedded PowerPC!!

In the end though, it's apples and oranges, so why not have fruit salad?
=)

Joshua Lamorie
Systems Designer
Xiphos Technologies Inc.
Article: 17729
Subject: Re: Feasibility of 200 MHz, 12K design on FPGA
From: Ray Andraka <randraka@ids.net>
Date: Fri, 27 Aug 1999 15:46:58 -0400
Links: << >>  << T >>  << A >>
Depends on the design, your skills and the device.  There's a good
chance you'll need to instantiate at a low level to get the constructs
you need unless you know your synthesis tool inside out (and probably
even then).  You're also probably going to have to floorplan to keep
routes tight.  It is certainly possible to do a 200MHz design in the
current crop of fast parts, but you aren't going to do it by accident.





thiru1457@my-deja.com wrote:

> Hi,
>
>    I intend to implement a logic in FPGA (roughly
> 12K logic gates complexity), running at 200 MHz.
>
>    Is it possible to implement the above design,
> with the current FPGA technology?.
>
>    How difficult it would be (interms of manweeks)
>    for synthesis and "place and route" the above
> design.
>
> Thanks in advance for your response.
>
> -Regards
>
> Thiru
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17730
Subject: Re: constrain into one XC4000 CLB
From: Ray Andraka <randraka@ids.net>
Date: Fri, 27 Aug 1999 16:02:31 -0400
Links: << >>  << T >>  << A >>
The UCF will only let you use LOC constraints, not RLOCs and you won't be
able to nail down the OR gate there.  Instead, put an FMAP around the OR
gate (in parallel with it),  and attach RLOC=R0C0 constraints to both
flip-flops, and to the FMAP.  Read the chapter on constraints in the online
user's guides.

Heinrich Fonfara wrote:

> Hi,
> can somebody tell me how to constrain two FFs that synchronize two
> different signals with the same clock followed by a simple or-gate to be
> implemented into one Xilinx XC4005  CLB.
> I use schematic entry from OrCAD and the usage of constraints is very
> similar to that of  Xilinx Foundation entry. I also use an UCF-file.
>
> Regards,
>
> Heinrich Fonfara



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17731
Subject: Re: multiplier Virtex
From: Ray Andraka <randraka@ids.net>
Date: Fri, 27 Aug 1999 16:08:12 -0400
Links: << >>  << T >>  << A >>
Depends on the size of the multiplier, obviously.  The efficient parallel
multiplier can do a 2xN partial product in N LUTs (2 per slice).  You need
to compute m/2 2xN partial products for an n x m multiplier.  Those
partial products are summed in an adder tree, which requires one LUT per
adder bit.  Some of the adders will need to have wired shifts in front of
them to handle the weighting of the individual partial products.  Please
refer to the multiplier page under the DSP section on my website for more
detail on multiplication in FPGAs.   For an area optimization, you could
go with a multiple of the data rate to share hardware over multiple
partial products, again refer to my multiplier page.

Pieter Op de Beeck wrote:

> Can anyone give me the total number of CLBs for a multiplier (area or
> performance optimized) in a Virtex ?  Is this the same as for the 4000
> family (why not?)?
>
> Thanks,
> Pieter



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17732
Subject: Re: microcontroller vs FPGA
From: Ray Andraka <randraka@ids.net>
Date: Fri, 27 Aug 1999 16:20:49 -0400
Links: << >>  << T >>  << A >>
For higher data rate stuff, no question about using FPGAs.  For lower data rates, bit serial designs in
FPGAs can make for a low power and very compact design.  FPGA design is made more difficult, but not
impossible, if there is lots of exception handling.

I have to argue with the chap who at the beginning of this thread says microcontrollers are always more
maintanable and cheaper for silicon and tools.  I recently looked at a microprocessor application, that
by the time we got the RTOS, the compilers, debuggers etc, we would have spent several times the cost
of basic FPGA tools, and even a bit more than a Synplicity synthesis suite with the RTL analyst.  With
proper FPGA design, there is no need to have to move pins.  I've done several hundred high speed FPGA
designs now, and I don't think I've ever had to move a pin.  The key is to pick the pin assignments
wisely (hint don't let the tool do it for you) and register at the I/O whenever possible.  BTW, the
Xilinx 3000 series, the Atmel 6K series, and Actel Act 1 families are all about 10 years old, and are
all still available today.

Dave Vanden Bout wrote:

> microcontrollers - used for low-speed I/O (KHz range) and algorithms with complicated control flow.
> FPGAs - used for high-speed I/O (MHz range) and algorithms with simple control flow.
>
> Daniel Figuerola Estrada wrote:
>
> > I am making a project in wich I compare advantages and disadvantages in
> > using microcontrollers and FPGA in the design of a digital system.
> >
> > Has anyone worked with those two technologies and could give his opinion
> > of them?
> >
> > ------------------------------------------------------------------------
> >  ////////////////////////////////\    ("`-''-/").___..--''"`-.__
> >  //  Marcel Figuerola Estrada   //      `6_ 6  )   `-.   (    ).`-.__.`)
> >  //       pfa@tinet.fut.es      //      (_Y_.)'  ._   )   `._ `.``-..-'
> >  //  Valls - Catalunya - Europe //    _..`--'_..-_/  /--'_.'  ,'
> >  \////////////////////////////////  (il),-''    (li),'  ((!.-'
> > ------------------------------------------------------------------------
>
>   ------------------------------------------------------------------------
>
>   Dave Vanden Bout <devb@xess.com>
>   FPGA Product Manager
>   XESS Corp.
>
>   Dave Vanden Bout
>   FPGA Product Manager  <devb@xess.com>
>   XESS Corp.
>   2608 Sweetgum Drive   Fax: (919) 387-1302
>   Apex                  Work: (919) 387-0076
>   NC                    Netscape Conference Address
>   27502
>   USA
>   Additional Information:
>   Last Name  Vanden Bout
>   First Name David
>   Version    2.1



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17733
Subject: Re: looking for image processing hardware
From: Ray Andraka <randraka@ids.net>
Date: Fri, 27 Aug 1999 16:22:19 -0400
Links: << >>  << T >>  << A >>
Try Visicom or Coreco (formerly dipix).  I think both have boards that might
meet your needs.

Sukandar Kartadinata wrote:

> Hi everybody,
> I'm looking for a DSP/FPGA development system to do the following:
>
> Capture the RGB output (1280x1024, 75Hz, 24bit) of a standard PC (Mac,
> SGI, etc.), do some processing, then output RGB again.
> (btw, there is a reason why I don't do the processing within the PC)
>
> Processing: 2D spatial transformations like zooms and shifts but also
> more complex according to displacement maps. Plus 4pt. bilinear
> interpolation
>
> In particular I couldn't find any frame grabbers that can handle the
> high pixelrates of almost 100Mpixel/s. Usually they're targeted at video
> capturing with max. 40Mpixel/s. Also, most frame grabbers are PCI cards
> - I'm looking for an embedded system solution.
>
> Another problem is memory bandwitdh. Apart from accessing the image
> buffer at least 4 times (for the interpolation) I also have to pipe in
> the displacement and interpolation data.
>
> What I did find is an ADC from AD (AD9884) that might at least be suited
> for the converting, although it's mainly targeted at LCD monitors.
> However, I'd really like to avoid messing too much with the hardware.
>
> In general I'd prefer a "one-supplier-solution" rather than buying
> components from different companies. I'd also be willing to spent some
> money on a nice software environmant rather than doing everything from
> scratch in assembler.
>
> Finally, does anyone know a newsgroup that specializes on image
> processing (like the music-dsp list does for sound) ??
>
> Thanks,
> Sukandar
> sk@zkm.de



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17734
Subject: test!
From: "George" <g_roberts75@hotmail.com>
Date: Fri, 27 Aug 1999 22:34:10 +0100
Links: << >>  << T >>  << A >>
test!


Article: 17735
Subject: PLL cascading in VIRTEX
From: "George" <g_roberts75@hotmail.com>
Date: Fri, 27 Aug 1999 22:41:06 +0100
Links: << >>  << T >>  << A >>
Hi,

Is it wise to cascade two frequency multipliers in the form of PLLs in
VIRTEX chips in order to get a high multiplication factor (within the
frequency limits of course), e.g. cascade PLLX4 and PLLX8 to obtain a
multipication factor of 32=4x8. Is there any limitation other than the
maximum PLL output frequency?

Thanx,

George


Article: 17736
Subject: Re: looking for image processing hardware
From: Sukandar Kartadinata <sk@zkm.de>
Date: Sat, 28 Aug 1999 02:41:47 +0200
Links: << >>  << T >>  << A >>
Ray Andraka wrote:

> Try Visicom or Coreco (formerly dipix).  I think both have boards that might
> meet your needs.

Thanks, but as far as I can see all their products max out at 40MHz for
capture, whereas we'd need 1280 x 1024 x 24bit x 75Hz = 100MHz.....

Sukandar

Article: 17737
Subject: Re: looking for image processing hardware
From: Sukandar Kartadinata <sk@zkm.de>
Date: Sat, 28 Aug 1999 02:42:07 +0200
Links: << >>  << T >>  << A >>
Ray Andraka wrote:

> Try Visicom or Coreco (formerly dipix).  I think both have boards that might
> meet your needs.

Thanks, but as far as I can see all their products max out at 40MHz for
capture, whereas we'd need 1280 x 1024 x 24bit x 75Hz = 100MHz.....

Sukandar

Article: 17738
Subject: Re: VHDL to debounce & latch input from a switch
From: "Clyde R. Shappee" <clydes@world.std.com>
Date: Sat, 28 Aug 1999 03:10:40 GMT
Links: << >>  << T >>  << A >>
Kevin,

Do you have high frequency clock available?  Say 10MHz or some such
frequency?

An alternative is to us a momentary switch that is single pole and two
throw.  It is less common, but the debounce curcuits are a pair of cross
coupled NAND or NOR gates.

Clyde


Kevin Toliver wrote:

> Can anyone tell me where to find vhdl code to debounce and latch a input
> from a momentary switch?
>
> Regards,
>
> Kevin

Article: 17739
Subject: Re: map hang
From: fliptron@netcom.com (Philip Freidin)
Date: 28 Aug 1999 03:26:59 GMT
Links: << >>  << T >>  << A >>

The following was posted June 11  th of this year, and is almost
certainly an answer to your problem:


==============================
A friend and I have just been through exactly this problem: Student 
edition F1.5 hangs in the router. We also had crashes in the mapper.

The solution to both was to apply BOTH the F1.5 service pack, and the 
version 2 hot fix to that service pack. Do not apply F1.5i service packs 
to version F1.5, it wont work.

The main service pack is at
	ftp://ftp.xilinx.com/pub/swhelp/M1.5_updates/15_service_pack1_nt.zip

and the increment is at
	ftp://ftp.xilinx.com/pub/swhelp/M1.5_updates/15_sp1_ftp2_nt.zip

For ref, see this for warnings
	http://www.xilinx.com/techdocs/4897.htm


Some of the problems seem to be related to having visual C++ V 6.0 on
your system.

Philip Freidin

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

You wrote:

In article <37C68FD6.45617301@ini.phys.ethz.ch>,
Tobi Delbruck  <tobi@ini.phys.ethz.ch> wrote:
>i got xilinx student edition v1.5, installed it on my windows NT machine,
>and it immediately
>started bombing at the "map" stage of implementation, with access violation
>at address 0x78001799.
>this happens even after i installed the service pack for 1.5.  also, i do
>NOT have microsoft
>visual studio installed.  it could be some other application installed a
>bad microsoft DLL.
>but i read in the xilinx tech notes that xilinx decided they did NOT think
>it was due
>to this DLL.  anyhow, it is all very confusing and unsatisfying because i
>cannot
>even compile a simple AND gate to my nice, brand new 4010 XS40 XESS board.
>
>does anyone know the real fix to this, besides buying for $500 the
>university edition
>of 2.1? (assuming that fixed it).
>
>thanks!
>-tobi delbruck
>
>Bob Pearson wrote:
>
>> Hi,
>>
>> Has any one encountered a situation where M1.5i
>> map get hung apparently in an infinite loop?  And
>> if so what was the cause?
>>
>> Thanks for any help
>>
>> Bob Pearson
>> bpearson@pmr.com
>
>
>
>--
>Tobi Delbruck
>Institute of Neuroinformatics, UniZ/ETHZ
>Winterthurerstr. 190
>8057 Zurich, Switzerland
>phone +41 1/635 30 38 fax +41 1/635 30 53
>tobi@ini.phys.ethz.ch
><http://www.ini.unizh.ch/~tobi>
>
>


Article: 17740
Subject: Re: Feasibility of 200 MHz, 12K design on FPGA
From: Phil Hays <spampostmaster@sprynet.com>
Date: Fri, 27 Aug 1999 23:49:04 -0700
Links: << >>  << T >>  << A >>
thiru1457@my-deja.com wrote:

>    I intend to implement a logic in FPGA (roughly
> 12K logic gates complexity), running at 200 MHz.
> 
>    Is it possible to implement the above design,
> with the current FPGA technology?.


It is impossible to estimate practicality or effort needed without any
more information.

The size is no problem.

The speed might be a problem.  The path from flip flop, through any
routing and/or logic, to setup into the next flip flop needs to be 5 ns
or less for all paths (with a few exceptions) for a design running at
200 MHz.  To route a signal a significant distance may well take several
ns.  To get a design to work at such speeds there can be at most only a
few levels of logic between each flip flop, and not much routing, or
perhaps a longer route (or a route with large fanout), and a single
4in-LUT (four input look up table).

The layout of the design also matters because longer routes take more
time.  If the design is regular and needs to communicate only to a few
of the nearest neighbor logic cells then it might be made to run very
fast.  If communication needs to be very general, it might be very hard
to make run fast.  If the design is regular and rectangular it might be
made to run very fast.  Very irregular logic is very difficult to place.
Some non rectangular logic is hard to place, and some is fairly easy.

If your design is already structured into simple logic between pipeline
stages that can be easily floorplanned into a regular rectangular
structure, a design like this might be done in 2 to 6 person weeks by an
experienced FPGA designer.

It also might take a few person years of effort to fit the design into
the next generation of FGPAs, defined for this statement as twice as
fast as the current generation.

It's also very possible that your design can't be done with existing
FPGAs at all, even assuming a factor of two speedup for the next
generation, regardless of effort, regardless of experience.

The questions you need to ask are:
1) can the design be pipelined "enough"?  (How much time will this
take?)
2) what is the "shape" of the parts of design?  How can they be
floorplanned to minimize route lengths and delays?

Best of luck.


-- 
Phil Hays
"Irritatingly,  science claims to set limits on what 
we can do,  even in principle."   Carl Sagan
Article: 17741
Subject: Re: PLL cascading in VIRTEX
From: Ray Andraka <randraka@ids.net>
Date: Sat, 28 Aug 1999 11:16:48 -0400
Links: << >>  << T >>  << A >>
Yes, there is a min/max frequency range for the DLLs.  I don't recall
offhand what the minimum reference frequency is, but IIRC, you can't get a
32x clock multiplication because you either violate the maximum DLL
frequency or the minimum reference.

George wrote:

> Hi,
>
> Is it wise to cascade two frequency multipliers in the form of PLLs in
> VIRTEX chips in order to get a high multiplication factor (within the
> frequency limits of course), e.g. cascade PLLX4 and PLLX8 to obtain a
> multipication factor of 32=4x8. Is there any limitation other than the
> maximum PLL output frequency?
>
> Thanx,
>
> George



--
-Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email randraka@ids.net
http://users.ids.net/~randraka


Article: 17742
Subject: Re: FPGA Express: Not enough storage...(etc.)
From: "Andy Peters" <apeters@noao.edu.nospam>
Date: Sat, 28 Aug 1999 20:19:09 -0700
Links: << >>  << T >>  << A >>
Joshua Lamorie wrote in message <37C6A794.EA82B0AA@xiphos.ca>...
>Gidday there,
>
> I'm currently climbing up the learning curve with FPGA Express, and
>porting my VHDL (isn't that a silly statement) from Cypress Warp.  I
>keep getting the following error on my code.
>
>Error L-1/C0:#0 Not enough storage is available to complete this
>operation.
>
> According to the Xilinx Answer #5301, this is due to underscores within
>binary strings.  However, though I am using binary strings, I do not
>have any underscores.  You can find the answer at:
>
>http://www.xilinx.com/techdocs/5301.htm
>
> Actually, it can't be anything with the strings, I just can't get past
>this step.  Is there a patch somewhere?  The software was installed by a
>Xilinx FAE, with about 30 CDs (exageration).  So, maybe he missed one.
>
> Any hints would be greatly appreciated.
>
>Joshua Lamorie
>Systems Designer
>Xiphos Technologies Inc.
>
>ps. Some specs
>
>PC: P-II 350 w/128MB RAM 270MB Virtual
>OS: NT4 SP5
>Foundation: F1.5 Build 3.1.140
>FPGA Express: 3.1.1.0w

Try downloading Service Pack 2 for F1.5.

-- a
----------------------------------------------------------------------------
--
Andy Peters
Sr Electrical Engineer
National Optical Astronomy Observatories
apeters (at) noao.edu

The Repulican Party: "We've upped our standards. Now, up yours!"



Article: 17743
Subject: Re: map hang
From: Tobi Delbruck <tobi@ini.phys.ethz.ch>
Date: Sun, 29 Aug 1999 16:56:28 +0200
Links: << >>  << T >>  << A >>
thank you thank you thank you!

that version 2 hot fix (which just replaces about 10 files in nt/bin, including
"map", did
in fact kill the crashing map problem.

the techdocs/4897.htm technote seems to have disappeared from the index and
trying to access it says it doesn't exist anymore.  so i don't know what it
says, but
so what - it runs now!

-tobi


Philip Freidin wrote:

> The following was posted June 11  th of this year, and is almost
> certainly an answer to your problem:
>
> ==============================
> A friend and I have just been through exactly this problem: Student
> edition F1.5 hangs in the router. We also had crashes in the mapper.
>
> The solution to both was to apply BOTH the F1.5 service pack, and the
> version 2 hot fix to that service pack. Do not apply F1.5i service packs
> to version F1.5, it wont work.
>
> The main service pack is at
>         ftp://ftp.xilinx.com/pub/swhelp/M1.5_updates/15_service_pack1_nt.zip
>
> and the increment is at
>         ftp://ftp.xilinx.com/pub/swhelp/M1.5_updates/15_sp1_ftp2_nt.zip
>
> For ref, see this for warnings
>         http://www.xilinx.com/techdocs/4897.htm
>
> Some of the problems seem to be related to having visual C++ V 6.0 on
> your system.
>
> Philip Freidin
>
> =====================================
>
> You wrote:
>
> In article <37C68FD6.45617301@ini.phys.ethz.ch>,
> Tobi Delbruck  <tobi@ini.phys.ethz.ch> wrote:
> >i got xilinx student edition v1.5, installed it on my windows NT machine,
> >and it immediately
> >started bombing at the "map" stage of implementation, with access violation
> >at address 0x78001799.
> >this happens even after i installed the service pack for 1.5.  also, i do
> >NOT have microsoft
> >visual studio installed.  it could be some other application installed a
> >bad microsoft DLL.
> >but i read in the xilinx tech notes that xilinx decided they did NOT think
> >it was due
> >to this DLL.  anyhow, it is all very confusing and unsatisfying because i
> >cannot
> >even compile a simple AND gate to my nice, brand new 4010 XS40 XESS board.
> >
> >does anyone know the real fix to this, besides buying for $500 the
> >university edition
> >of 2.1? (assuming that fixed it).
> >
> >thanks!
> >-tobi delbruck
> >
> >Bob Pearson wrote:
> >
> >> Hi,
> >>
> >> Has any one encountered a situation where M1.5i
> >> map get hung apparently in an infinite loop?  And
> >> if so what was the cause?
> >>
> >> Thanks for any help
> >>
> >> Bob Pearson
> >> bpearson@pmr.com
> >
> >
> >
> >--
> >Tobi Delbruck
> >Institute of Neuroinformatics, UniZ/ETHZ
> >Winterthurerstr. 190
> >8057 Zurich, Switzerland
> >phone +41 1/635 30 38 fax +41 1/635 30 53
> >tobi@ini.phys.ethz.ch
> ><http://www.ini.unizh.ch/~tobi>
> >
> >



--
Tobi Delbruck
Institute of Neuroinformatics, UniZ/ETHZ
Winterthurerstr. 190
8057 Zurich, Switzerland
phone +41 1/635 30 38 fax +41 1/635 30 53
tobi@ini.phys.ethz.ch
<http://www.ini.unizh.ch/~tobi>


Article: 17744
Subject: size of configuration data?
From: Asa Kalavade <kalavade@bell-labs.com>
Date: Sun, 29 Aug 1999 11:46:39 -0400
Links: << >>  << T >>  << A >>

Hi,
I am working with the 4085xla device and running xact m1.5.
The size of the configuration files (.rbt, .bit) generated
in my most recent run does not match the size from previous
runs. The bits printed in the .rbt file are still 1924992.
 
Should I be concerned about this size mismatch? (The design
does not work and I was wondering if this could be a reason.)

ls -lt foo*
all previous designs:
v1/foo1.rbt	1927897
v1/foo1.bit	240697

new design:
v2/foo1.rbt	1927901
v2/foo1.bit	240701

thanks,
asa
kalavade@lucent.com
Article: 17745
Subject: Re: size of configuration data?
From: fliptron@netcom.com (Philip Freidin)
Date: 29 Aug 1999 17:30:10 GMT
Links: << >>  << T >>  << A >>

Slight variations in length of the .RBT amd .BIT files is
normal. For instance, I have a XC4085XLA design with the
following lengths:

design.BIT  240699
design.RBT 1930623

The variation in length is due to the fact that both file
formats include a variable length header, prior to the
actual bit stream itself. You can check this easily by
looking at the header of the .RBT:

Xilinx ASCII Bitstream
Created by Bitstream M1.5.28i
Design name: 	design.ncd
Architecture:	xc4000xl
Part:        	4085xlbg560
Date:        	Fri Jul 09 01:01:24 1999
Bits:        	1924992              <<<< check this, it
                                     <<<< should not change.

Good luck,
Philip Freidin


In article <37C955DF.61772FC3@bell-labs.com>,
Asa Kalavade  <kalavade@bell-labs.com> wrote:
>
>Hi,
>I am working with the 4085xla device and running xact m1.5.
>The size of the configuration files (.rbt, .bit) generated
>in my most recent run does not match the size from previous
>runs. The bits printed in the .rbt file are still 1924992.
> 
>Should I be concerned about this size mismatch? (The design
>does not work and I was wondering if this could be a reason.)
>
>ls -lt foo*
>all previous designs:
>v1/foo1.rbt	1927897
>v1/foo1.bit	240697
>
>new design:
>v2/foo1.rbt	1927901
>v2/foo1.bit	240701
>
>thanks,
>asa
>kalavade@lucent.com


Article: 17746
Subject: UK programmable logic companies
From: Tim Tyler <tt@cryogen.com>
Date: Sun, 29 Aug 1999 23:18:58 GMT
Links: << >>  << T >>  << A >>
I'm attempting to add some UK-oriented *companies* to my programmable
logic links (at http://www.alife.co.uk/links/hardware/).

So far I've come up with the following list:

http://www.embedded-solutions.ltd.uk/ Embedded solutions - boards
http://www.alphadata.co.uk/hardhome.html Alphadata - FPGA boards
http://www.nallatech.com/ Nallatech - FPGA boards
http://www.microcall.memec.com/xilinx/xilinx.htm Micro Call Ltd
http://www.entegra.co.uk/ EnTegra Ltd Home Page - in the DSP field...
http://www.doulos.co.uk/ VHDL/Verilog training and HDL-based FPGA designs
http://www.orcad.co.uk/europe/products/products_f.htm Orcad Express.
http://www.archaeoptryx.co.uk/ Xilinx Full Chip Design VHDL/Verilog
http://www.saros.co.uk/ VHDL and Verilog design tools
http://www.eda.co.uk/ - makers of "Protel for Windows"
http://www.nmi.co.uk/ip.htm NMI

If I've missed any prominent companies, please feel free to mail me their
URLs.
-- 
__________
 |im |yler  The Mandala Centre  http://www.mandala.co.uk/  tt@cryogen.com

Some things have to be believed to be seen.
Article: 17747
Subject: Re: Smallest Configurator for Xilinx
From: jim granville <Jim.Granville@DesignTools.co.nz>
Date: Mon, 30 Aug 1999 16:40:23 +1200
Links: << >>  << T >>  << A >>
Joshua Lamorie wrote:
> 
> Rickman wrote:
> 
<snip>
> We have decided to go with an Altera device, because it is the
> smallest.
> 
> The only question remaining is converting the output from the Xilinx software
> (some sort of bit file) into an SVF, or JAM file required by the Altera JTAG
> programmer.  Or, I hack together a JTAG programmer of my own. Any hints?

Ask Altera.
 
Another option, given the 'high' prices of these FPGA serial memories,
is
a uC and other memory.

Fairchild TSSOP8 uC ACE1101 + ATMEL TSSOP14 memory, choice of 3
protocals i2c : 24C256,  SPI : 25C256,  DataFlash :  AT45D011

and its a reprogrammable solution..

-jg


-- 
======= Manufacturers of Design Tools for uC and PLD  =====
* IceP2051 - Full Speed ICE, for 1K,2K,4K 20 Pin FLASH controllers
=> http://www.DesignTools.co.nz/winner51.htm  for highlights


Article: 17748
Subject: AMD Athlon CPU Speed at Simulates
From: jim granville <Jim.Granville@DesignTools.co.nz>
Date: Mon, 30 Aug 1999 16:42:49 +1200
Links: << >>  << T >>  << A >>
Hello,
According to the specs, the new AMD Athlon CPU is ~35% faster than
a equiv Pentium at real number work.

How does this translate into FPGA / VHDL Simulation performance speeds ?

Has anyone tried this CPU on FPGA tools yet ?

thanks - jg

Article: 17749
Subject: Re: looking for image processing hardware
From: walter <walter_daniel@my-deja.com>
Date: Mon, 30 Aug 1999 04:52:01 GMT
Links: << >>  << T >>  << A >>
Sorry but I don't read the original,
we build this boards, please send me by e-mail your needs

Walter.

In article <37C7305A.6A9F08FA@zkm.de>,
  sk@zkm.de wrote:
> Ray Andraka wrote:
>
> > Try Visicom or Coreco (formerly dipix).  I think both have boards that might
> > meet your needs.
>
> Thanks, but as far as I can see all their products max out at 40MHz for
> capture, whereas we'd need 1280 x 1024 x 24bit x 75Hz = 100MHz.....
>
> Sukandar
>
>

--
Walter D. Gallegos
Software & FPGA Services

Montevideo, MVD


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.


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