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 7175

Article: 7175
Subject: Re: Price of Serial EEPROM is Outrageous
From: s_clubb@netcomuk.co.uk (Stuart Clubb)
Date: Sun, 10 Aug 1997 20:14:09 GMT
Links: << >>  << T >>  << A >>
On Sat, 09 Aug 1997 08:07:35 GMT, peter299@maroon.tc.umn.edu (Wade D.
Peterson) wrote:

>If you are using XILINX serial proms I agree.  XILINX is very
>preditory on their pricing.  So much so that it's probably illegal,
>although I don't know anybody who has sued them yet.  They play some
>very nasty sales tactics.

That's quite a statement! Would you care to elaborate?

Stuart.
Article: 7176
Subject: Re: Download FLEX10K over the LPT port
From: !(no@spam)@vcd.hp.com (J.Szamosfalvi)
Date: 11 Aug 1997 03:23:22 GMT
Links: << >>  << T >>  << A >>
: Thomas Berndt (berndt@eas.iis.fhg.de) wrote:
: : 
: : Hi,
: : 
: : does anybody knows how I can download designs which are compiled with MAXPLUS2
: : over the LPT port?

You need a byteblaster between the parallel port and a 10 pin download header
in your circuit.   MAXPLUS2 will automatically generate the needed *.sof 
file for you.  You need a Windows PC to do a download.  Read the applicable 
application notes very thoroughly before design and pin assignment.
 
---
#include <Standard.Disclaimer> about this post being a personal opinion

Article: 7177
Subject: Re: Problems with Lattice pDS+
From: nick toop <nick@cortexco.demon.co.uk>
Date: Mon, 11 Aug 1997 11:46:08 +0100
Links: << >>  << T >>  << A >>
In article <t7wlo2l33l0.fsf@mandakini.eecs.umich.edu>, Kiran Nimmagadda
<kirann@eecs.umich.edu> writes
>hello everybody,
>
>i am facing some problems compiling my design in the 
>lattice pDS+ viewlogic software. the version of pDS+ i am using is 
>1.0 and lattice no longer provides support to it as it is too old(1993). 
>
>my design is very simple...it basically consists of two counters, three
>muxs and five user( i.e., me) defined macros. all the user defined macros
>have only and & or gates and compile fine. i am attaching a part of the
>report file and the message log...
>-------------------------------------------------------------------------
>excerpts from the report file....
>------------------------------------
>Fitter Parameters Used
>----------------------
>
>       AVG_GLB_IN:  16
>             AUTO:  OFF
>           EFFORT:  4
> IGNORE_FIXED_PIN:  OFF
>        MAX_DELAY:  2
>       MAX_GLB_IN:  16
>       PARAM_FILE:  
>             PART:  ispLSI1016-90LJ
>       TIMING_SIM:  timing
>              TRY:  
>       FAST_ROUTE:  OFF
>     STRONG_ROUTE:  FIXED
>
>Process Status
>--------------
>
>      Design Analysis:  complete
> Pre-Route Netlisting:  complete
>   Logic Partitioning:  complete
>      Place and Route:  incomplete
>           Post Route:  incomplete
> Fuse File Generation:  incomplete
>------------------------------------------------------------------------
>excerpts from the message log....
>------------------------------------
>dpm> pass1 elapsed time: 00:00:52
>dpm> formatting output...
>LAF2LIF Version 1.50  8/19/93
>Copyright (c) Lattice Semiconductor Corp. 1991-1992. All rights reserved.
>
>17018 INTERNAL ERROR: -    25: yacc: Syntax error, state 58.
>17018 INTERNAL ERROR: -    31: yacc: Syntax error, state 58.
>17018 INTERNAL ERROR: -    33: yacc: Syntax error, state 58.
>17018 INTERNAL ERROR: -    35: yacc: Syntax error, state 203.
>17018 INTERNAL ERROR: -    36: yacc: Syntax error, state 203.
>17018 INTERNAL ERROR: -    37: yacc: Syntax error, state 203.
>17003 INTERNAL ERROR: -   130: Invalid <sym_phys_loc>: '/'
>17018 INTERNAL ERROR: -   131: yacc: Syntax error, state 282.
>17004 INTERNAL ERROR: -   131: Illegal pin attribute.
>17003 INTERNAL ERROR: -   154: Invalid <sym_phys_loc>: '/'
>17018 INTERNAL ERROR: -   155: yacc: Syntax error, state 282.
>17004 INTERNAL ERROR: -   155: Illegal pin attribute.
>17003 INTERNAL ERROR: -   162: Invalid <sym_phys_loc>: '/'
>17018 INTERNAL ERROR: -   163: yacc: Syntax error, state 282.
>17004 INTERNAL ERROR: -   163: Illegal pin attribute.
>17016 INTERNAL ERROR: IOC on net RAS has no associated XPIN!
>17016 INTERNAL ERROR: IOC on net W has no associated XPIN!
>17016 INTERNAL ERROR: IOC on net CAS has no associated XPIN!
>dpm> creating reports...
>dpm> 25502 WARNING: could not open report file CASRASW.rp7
>dpm> abnormal exit. process stopped.
>dpm> total elapsed time: 00:02:08
>-----------------------------------------------------------------------------
>
>the error numbers above are not listed in the error index in the manual and
>i have no clue as to what the problem could be. if someone can give some
>insight into this situation, it would be great. 
>
>thanks a lot in advance and email replies are most welcome to
>kirann@eecs.umich.edu 
>
>regards,
>kiran nimmagadda.

I have seen no replies.  I am no expert and apologise if I am stating
the obvious. 

"yacc" is a standard parsing program and in this case it is unable to
analyse your input file...it doesn't think it sees proper syntax. It
would appear that it objects to '/'...is this a valid input token?

Are you running Windows NT 4.0? There are problems with it and some
Lattice programs...intermediate files seem to get scrambled in some way. 

As a last resort I would suggest you keep simplifying your input file
till things work and then start putting things back to find out what
input line makes it fail.

I recommend the Lattice/Synario free CD most strongly.

Regards,
Nick Toop

-- 
nick toop
Article: 7178
Subject: Splash-2 et al.?
From: <dnk@io.com>
Date: 11 Aug 1997 15:26:27 GMT
Links: << >>  << T >>  << A >>
I have scanned through the Splash-2 book, and it mentions commercial
licensing.  However, I've looked at:

        http://www.super.org:8000/FPGA/splash2.html
        http://pequod.ee.vt.edu/
	http://www.optimagic.com/research.html
	http://www.io.com/~guccione/HW_list.html
	(Web searches turns up mostly Stanford's Splash-2, 
	which I'm sure isn't the same FPGA-based Splash-2).  

And I'm not able to find commercial boards, except Wildfire from
Annapolis Micro Systems.  Are there others?

Thanks in advance for any leads.

Danny Kumamoto

mailto:dnk@pobox.com   http://www.pobox.com/~dnk     TEL: +1 512-918-3640
Postal: 13492 Research Blvd., Suite 120-295, Austin, TX 78750-2252, U.S.A.
Article: 7179
Subject: VSI Meeting Notice
From: sbaker@best.com (Stan Baker)
Date: Mon, 11 Aug 1997 15:55:02 GMT
Links: << >>  << T >>  << A >>
	
NOTICE OF VSI MEETING:

THE WORLDWIDE MEMBERSHIP MEETING OF
THE VIRTUAL CHIP INTERFACE ALLIANCE

Friday, October 3, 1997 (after the Embedded Systems Conference)
Silicon Valley
Free for representatives from VSIA member companies.
$150 for non-members.
	
PROGRAM 
The fall membership meeting will break some new ground.  It will not
only present the progress of the Working Groups (DWGs) but will begin
the discussion of how this work will benefit the engineering of
system-level ICs.  We are inviting non-members (for a fee) so we can
begin to discuss the application of VSI specifications with the
industry beyond the Alliance.

We will also have an important keynote speaker at lunch and updates on
the progress of the Alliance itself.

The program will begin at 9:00 and conclude at 4:00 pm when parallel
sessions for each of the six DWGs will begin, ending at 5:30pm.  A
cocktail reception for all attendees follows.

You are invited to send as many representatives as you wish. The DWG
break-out sessions at 4:00pm will include:
--Design Implementation and Verification
--High-level System Design
--IP Protection 
--Mixed-Signal
--On-chip Buses
--Manufacturing Related Test

The VSI website (www.vsi.org) will maintain a current description of
the meeting as it develops further.   

DETAILS
9:00-5:30 pm  (Reception at 5:30)
Check-in and breakfast: 7:30-9:00 am

TO REGISTER
Please register in advance, by contacting the VSI Alliance:
15495 Los Gatos Boulevard, Los Gatos, CA 95032, USA
Phone: 1-408-356-8800
Fax: 1-408-356-9018
Email: conference@vsi.org
Registration info on the web.

VENUE
Santa Clara Marriott
2700 Mission College Blvd., Santa Clara, Calif.
By special arrangement, a limited number of hotel rooms are available
at the Marriott at a rate of $149, single or double occupancy per
night.  
For reservations, call 1-800-228-9290 or 1-408-327-6808.  Mention the
"VSI Meeting".  
The special priced rooms are available on the nights of October 2nd
and 3rd.  You must make reservations by September 11 to get this rate.
Individuals re responsible for their own billing arrangements with the
hotel.

Please register as soon as possible and indicate which DWG breakout
session you expect to attend -- so we can make plans for a good
meeting.

Best regards,
Stan Baker
Executive Director,
VSI Alliance

Article: 7180
Subject: Re: free FPGA software from actel
From: dstewart@dmicros.com
Date: 11 Aug 1997 20:24:17 +0200
Links: << >>  << T >>  << A >>
On 10 Aug 1997 15:00:59 GMT, rstevew@armory.com (Richard Steven Walz)
wrote:

>In article <33ebabe0.228220832@news.m.iinet.net.au>,
>David R Brooks <daveb@iinet.net.au> wrote:
>>Sam Falaki <Falaki@uqtr.uquebec.ca> wrote:
>>
>>:> >> Has somebody tested the free actel software? Is it worth to download?
>>:> >> Is programmer support included? Is the ACTIVATOR needed or is there
>>:> >> a free design for a simple programming hardware.
>>:
>>:I managed to compile VHDL and target a very wide range of devices
>>:without even going through any tutorial or reading any "help"
>>:whatsoever.  It's really easy to use.  You can kind of guess your
>>:way along and next thing you know you're done.  I liked it.
>>
>> The software may be free (and good), but what else do you need to get
>>working silicon? A big-bucks programmer, maybe? The bottom line is the
>>cost of _all_ tools necessary.
>>
>> Please, will one of the SRAM-based FPGA vendors put out something
>>similar? So we (the great impoverished <g>) can actually put their
>>chips to work in real systems. 
>>
>>--  Dave Brooks <http://www.iinet.net.au/~daveb>
>>PGP public key: finger  daveb@opera.iinet.net.au
>>                servers daveb@iinet.net.au
>>    fingerprint 20 8F 95 22 96 D6 1C 0B  3D 4D C3 D4 50 A1 C4 34
>> What's all this? see http://www.iinet.net.au/~daveb/crypto.html
>---------------------------------------------
>Last I looked, Altera had bought Intel's FlexPLD's and called them FlashLOGIC
>or such and they are FPGA's that are SRAM and EEFLash programmable with a
>LPT-JTAG cable I have the plans for that uses just one '244. I don't do PLD
>that big, so I haven't bothered, but that sounds like what you just asked
>for. Yes? No? All you need to do is hunt down the command to dump the SRAM
>to EEFlash, (they want to sell you software to do that one tiny bit, but it
>is a well known code), and you got it!! I THINK I still have that info!
>Write or call them! www.altera.com .
>-Steve
>-- 
>-Steve Walz  rstevew@armory.com  ftp://ftp.armory.com:/pub/user/rstevew
>-Electronics Site!! 1000 Files/50 Dirs!! http://www.armory.com/~rstevew
>Europe:(Italy) ftp://ftp.cised.unina.it:/pub/electronics/ftp.armory.com
>Oz:.AU ftp://ftp.peninsula.apana.org.au:/pub/electronics/ftp.armory.com
Altera have announced that they are to discontinue the Flashlogic
family,  so I wouldn't design them in.  I think the last time buy for
these parts has already been done.  PLDShell,  which was the free
software,  was crap anyway.

Article: 7181
Subject: Re: Price of Serial EPROM is Outrageous - Better Explanation
From: jim granville <Jim.Granville@xtra.co.nz>
Date: Mon, 11 Aug 1997 13:07:59 -0700
Links: << >>  << T >>  << A >>
Tom Burgess wrote:
> 
> Bruce Nepple wrote:
> >
> > An example
> > Altera'a new EPF6016 is $22, decreasing to $12 (budgetary) in 1998
> > It takes a serial 260KBit EEPROM, priced at $18 (Yes, from Atmel).
> > I can buy a parallel EPROM for $3.  What gives?
> >
> 
> Last time I checked, (over a year ago) the Atmel parts were at rough
> price parity with the non-reprogrammable Xilinx 17xxx parts and
> therefore were not a bad deal, but both are still outrageous compared
> to jellybean 27Cxxx EPROMs. The reason? Limited sales volume
> (microscopic compared to EPROMs), captive markets, and not much
> competition. I believe that the vendors would say they are providing
> the serial PROMS as a service and are basically just recovering costs.
> Since there is usually a microprocessor in a system that uses FPGAs,
> most cost-conscious designers go for serial or parallel download,
> in effect using cheap EPROM (or disk) instead of expensive serial PROMs.
> 
>         regards, tom (tburgess@drao.nrc.ca)

It may also be that the FPGA vendors realise that most customers looking
at
FPGA do NOT include the SPROM prices when comparing brands of FPGA, and
also
when comparing FPGA - CPLD.

This makes it a good place to do some margin recovery :-)

 Another option may be the new DATAFLASH devices from ATMEL.
These are SPI interface, 4MBit serial FLASH devices, in larger packages
than
the 17xxx series, but at appx 20% lower cost than 4MB byte wide FLASH,
they
are targeted for answer phones ( == good volumes, and price fussy market
).

 I have not looked to see how they interface to FPGA bootload, but if
you were choosing
a F/EPROM + small uC, a DATAFLASH <- Small uC -> FPGA is an alternative. 


-- 
======= Manufacturers of Serious Design Tools for uC and PLD  =========
= Optimising Modula-2 Structured Text compilers for ALL 80X51 variants
= Reusable object modules, for i2c, SPI and SPL bus interfaces
= Safe, Readable & Fast code - Step up from Assembler and C
= Emulators / Programmers for ATMEL 89C1051, 2051, 89C51 89S8252 89C55
= *NEW* Bondout ICE for 89C51/89C52/89C55 
= for more info, Email : DesignTools@xtra.co.nz  Subject : c51Tools

Article: 7182
Subject: *Fast Fault Simulation*
From: rcstwks@aol.com (RCSTWKS)
Date: 11 Aug 1997 21:42:18 GMT
Links: << >>  << T >>  << A >>
Fault-Simulation - Turbo Fault

High Performance Fault Simulator:

TurboFault combines high performance, versatility and accuracy. It is
highly competitive with hardware accelerators for classical test
fault grading. It supports synchronous and asynchronous designs at the
gate-level, including tri-state gates, latches, flip-flops, single and
multi-port RAMS, complex bus resolution functions, and USER Defined
primitives (UDPs). TurboFault reads Verilog gate-level netlists, and will
also read Standard Delay Format (SDF) timing files.

Advanced Cached-Concurrent Algorithm:

Turbofault utilizes a new algorithm optimed for today's computer
hardware that maximizes the simulation power of workstations. Syntest
Cached-Concurrent algorithm eliminates needless operations and with new
Fast Queque technology combines the best of unit delay and cycle-based
capabilities. No other fault simulator, hardware or software, matches the
performance of TurboFault.

TurboFault makes fault simulation an integral design tool for generating
a quality manufacturing test set. TurboFault supports single timing delay
for simulation accuracy and flexibility, without sacrificing speed.

TurboFault is the fastest concurrent fault simulator based on the latest
advances in cycle-based simulation technology. It simulates even *faster*
than existing expensive hardware accelerated fault simulators. Fault
simulation also consumes memory very quickly, so memory management is
critical. TurboFault combines efficient memory management with special
fault handling resulting in low memory consumption. 

To hear more of Turbo-Fault, please send an e-mail to
Suzanne@world.std.com, please send us your company name, your name 
and fax #. 

Thanks!

The staff at Syntest Technology.

*ATPG*
*Boundry Scan*
*RAM BIST*
*Design for test services*
*Training*
*TurboFault Simulation*
*IDDQ*
*Testability Analysis*

-------------------------------------------------------/-/-/-/-/-/-

ps. Ask about our 50% trade up offer!
Article: 7183
Subject: FPGA'98 Call For Papers
From: hauck@ece.nwu.edu (Scott A. Hauck)
Date: Mon, 11 Aug 1997 19:54:03 -0600
Links: << >>  << T >>  << A >>
                                 FPGA `98
                             Call for Papers

                1998 Sixth ACM International Symposium on
                     Field-Programmable Gate Arrays

                  DoubleTree Hotel, Monterey, California
                          February 22-25, 1998

                   http://www.ece.nwu.edu/~hauck/fpga98

        ==========================================================
As Field-Programmable Gate Arrays become more essential to the design of
digital systems there is increased pressure to improve their performance,
density and automatic design. This symposium follows the largest ever
gathering of this kind last year in Monterey at FPGA `97. For FPGA `98,
we are once again soliciting submissions describing novel research and
development in one or more of the following (or similar related) areas
of interest:

FPGA architecture: logic block & routing architectures, I/O structures
                   and circuits, new commercial architectures.
CAD for FPGAs:     placement, routing, logic optimization, technology
                   mapping, system level partitioning, testing and
                   verification.
Interactions:      between CAD, architecture, applications, and
                   programming technology.
Fast prototyping:  for System level design, Multi-Chip Modules.
Applications:      use of FPGAs in novel circuits, as emulators and
                   compiled accelerators.
Field-programmable interconnect chips and devices (FPIC/FPID.)
FPGA-based compute engines.
Field-programmable analog arrays.

        ==========================================================
Authors should submit 20 copies of their paper (12 pages maximum) by
September 26, 1997. Notification of acceptance will be sent by December 1,
1997. The authors of the accepted papers will be required to submit the
final camera ready copy by December 15, 1997. A proceedings of the accepted
papers will be published by ACM, and included in the ACM/SIGDA CD-ROM
publications. All submissions should be sent to:

                           Sinan Kaptanoglu
                               FPGA `98
                          Actel Corporation
                       955 East Arques Avenue,
                       Sunnyvale, CA 94086 USA
                        e-mail:sinan@actel.com
                        phone: (408) 522-4319
                          fax: (408) 522-8041

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

General   Chair: Jason Cong,       UCLA,
Financial Chair: Carl Ebeling,     U. of Washington,
Program   Chair: Sinan Kaptanoglu, Actel,
Publicity Chair: Scott Hauck,      Northwestern U.

============================
Technical Program Committee:
============================

Michael Butts,             Quickturn
Jason Cong,                UCLA
Eugene Ding,               Lucent
Carl Ebeling,              U. of Washington
Scott Hauck,               Northwestern U.
Dwight Hill,               Synopsys
Brad Hutchings,            BYU
Sinan Kaptanoglu,          Actel
David Lewis,               U. of Toronto
Fabrizio Lombardi,         Texas A&M
Jonathan Rose,             U. of Toronto
Rob Rutenbar,              CMU
Malgorzata Marek-Sadowska, UCSB
Gabriele Saucier,          IMAG
Martine Schlag,            UCSC
Tim Southgate,             Altera
Steve Trimberger,          Xilinx
John Wawrzynek,            UCB
Martin Wong,               UT at Austin

============================================================================
Sponsored by ACM SIGDA, with support from Actel, Xilinx, Altera, and Lucent.
============================================================================
+-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+
|               Scott A. Hauck, Assistant Professor                         |
|  Dept. of ECE                        Voice: (847) 467-1849                |
|  Northwestern University             FAX: (847) 467-4144                  |
|  2145 Sheridan Road                  Email: hauck@ece.nwu.edu             |
|  Evanston, IL  60208                 WWW: http://www.ece.nwu.edu/~hauck   |
+-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+
Article: 7184
Subject: Re: Are 2 PCs better than One?
From: "George Tzanatos" <gtzanatos@philips.oz.au>
Date: 12 Aug 97 02:49:50 GMT
Links: << >>  << T >>  << A >>


Wen-King Su <wen-king@myri.com> wrote in article
<5s2uji$js8@neptune.myri.com>...
> In a previous article mush@jps.net (David Decker) writes:
> 
> ;2. I've seen a switch that can switch a monitor, keyboard,
> :and mouse. among up to 4 live PC boxes. 
> ;(OmniView Belkin F1D066 for ~$300.00.)
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
A 4-way switchbox (monitor/keyboard/mouse) costs about $50 ex tax in
Australia. 
Cables are a couple of dollars each. The only thing to watch out for is 
the monitor cables if you are running >800x600 @ 256 colors (buy the
'better' 
monitor cable that they sell - it the one made using coax rather than
hookup wire).

The switchboxes work well, although if you are verry fussy and running at
higher 
resolutions,  you may notice minor ghosting. Also, some VGA cards powerup 
in MONO mode if they don't see a monitor when the PC is powered on.

> :Is it any good?
> ;
> :2a. There is also the possibility of controlling multiple PCs on a
> ;network, remotely, from one PC. Carbon Copy, I think. Is this a better
> :way?
> 
> Oh the shackles that microsoft made us wear.  It is sad to see a whole
> lot of us getting driven to these contraptions in order to stay
productive.
> The most attractive solution is to get UNIX boxes.   With a site-wide
> licence, and a shell interface, you can start any number of runs on any
> number of machines from your machine at the office, at home, or anywhere.
> We do that routinely with ORCA FPGA software, which is a close kin to
> Xilinx M1.  I can start them, put them in the back ground, go away and
> log in some some where else to check on them and continue to work on
them.
> I do not currently have that luxury with the tool for another FPGA, and
> I am now sitting in my office waiting for it to finish so I can proceed
> with the next step, when I can be instead be waiting remotely at home,
> perhaps next to my pool. 
> 
Article: 7185
Subject: Any one getting 125MHz out of XILINX CPLDs?
From: Kayvon Irani <kirani@cinenet.net>
Date: Mon, 11 Aug 1997 22:57:28 -0700
Links: << >>  << T >>  << A >>
Hi:
	I'm looking at Xilinx CPLDs for a new design and trying to match the 
	maximum clock frequency number	in the data book for XC95108-7. The 
	maximum clock frequency on a 16bit counter I get is 83 MHz their number
	is 125 MHz. Can any one duplicate their results?

	Regards,
	Kayvon Irani
	Los Angeles, Ca
Article: 7186
Subject: Re: PCI Interface
From: "Austin Franklin" <d6arkroom@ix.netcom.com>
Date: 12 Aug 1997 12:42:33 GMT
Links: << >>  << T >>  << A >>
Steve,

I believe no matter what you ask the x86 to do...it will only do 4 DWORD
transfers per address phase....is that not true?

Austin Franklin
darkroom@ix.netcom.com


Steve Casselman <sc@vcc.com> wrote in article
<33EE1F9A.2BCA69F8@vcc.com>...
> > > I want to design a card with a PCI interface.
> > >  1. Can I make use of the DMAC of any 486/Pentium machines for data
> > >     transfers ? Or do I have to use a DMAC on my card and develop
> > >     an arbiter as well ?
> >
> 
> This code will generate a burst write from the
> cpu to a pci board if the board can handle
> bursts.
> /*
> ** This function talks to the on board RAM
> ** On the HOT Works board.
> */
> 
> void
> Pci6200::writeRAM(unsigned long addr, unsigned long * data, unsigned
> long count){
> 
>  unsigned long addrRAM = addr + membase;
> /* addr and membase come from the
>     driver and repersent the memory
>     mapped location of the board
> */
>         __asm
>         {
>                 push edi
>                 push ecx
>                 push esi
>                 mov esi, data
>                 mov edi, addrRAM
>                 mov ecx, count
>                 cld
>                 rep movsd
>                 pop esi
>                 pop ecx
>                 pop edi
>         }
> 
> 
>       }  /* end Pci6200::writeRAM() */
> 
> --
> Steve Casselman, President
> Virtual Computer Corporation
> http://www.vcc.com
> 
> 
> 

Article: 7187
Subject: Re: Price of Serial EEPROM is Outrageous
From: peter299@maroon.tc.umn.edu (Wade D. Peterson)
Date: Tue, 12 Aug 1997 15:06:03 GMT
Links: << >>  << T >>  << A >>
> Have you looked at the Atmel AT17Cxxx parts? They replace the Xilinx
>serials, and Atmel are usually pretty keen on pricing. You get
>in-circuit programmablility as an extra. The ISP algorithm? Its just
>I2C.

The real problem is that if you use XILINX FPGA, but don't buy their
PROMs, then XILINX starts jacking up their prices on everything you
buy from them.  XILINX has probably the most preditory pricing that
I've ever seen.  [I think it's downright illegal, but then again I'm
not a lawyer].

There are two companies (that I know of) here in town that are stuck
buying all XILINX FPGA and PROMs.  If they buy even one part from
somebody else like Altera, then XILINX will come in and increase their
costs by 40% or more.  


-----------------------------------------------------
Wade D. Peterson         | TEL: 612.722.3815
Consultant to Industry   | FAX: 612.722.5841
3525 E. 27th St. No. 301 |---------- EMAIL ----------
Minneapolis, MN  55406   | peter299@maroon.tc.umn.edu
---------------- Committed to Quality ---------------

Article: 7188
Subject: Job announcement?
From: Martin Maierhofer <m.maierhofer@tees.ac.uk>
Date: Tue, 12 Aug 1997 17:31:01 +0000
Links: << >>  << T >>  << A >>
Hello everybody.

Since I am new to this newsgroup, I wanted to check first if
it is ok to post a job announcement hre. If not, could
you please point me to a more reasonable newsgroup?

BTW, our project involves FPGA, VHDL, PCI -- and obviously we
are located in the UK.

Thanks for your patience,
  martin

-- 
Martin Maierhofer       | "History will teach us nothing."
m.maierhofer@tees.ac.uk |                     - Sting
Article: 7189
Subject: Re: Any one getting 125MHz out of XILINX CPLDs?
From: terry.harris@dial.pipex.com (Terry Harris)
Date: Tue, 12 Aug 1997 17:33:43 GMT
Links: << >>  << T >>  << A >>
Kayvon Irani <kirani@cinenet.net> wrote:

>Hi:
>	I'm looking at Xilinx CPLDs for a new design and trying to match the 
>	maximum clock frequency number	in the data book for XC95108-7. The 
>	maximum clock frequency on a 16bit counter I get is 83 MHz their number
>	is 125 MHz. Can any one duplicate their results?

I expect you are using normal rather than local or pin feedback which
introduces another 4ns delay. 

You can only use local feedback within a function block and pin
feedback from actual output pins. This means the maximum speed of
state machines and counters depends on the fit and impacts Xilinx's
claims about pin locking ability of the 9500 series if you are trying
to go fast. 

The 9536 doesn't have local feedback at all so a slower 9572 can
actually be faster for some designs. 

I had one design where 7 bits of a 16bit bus arrived 6.5ns before the
rest because only those bits could take advantage of local feedback  -
yuk. The design ended up in a Lattice 3000 series which worked out
much better all round. 

Cheers Terry...
Article: 7190
Subject: Re: Price of Serial EEPROM is Outrageous
From: s_clubb@netcomuk.co.uk (Stuart Clubb)
Date: Tue, 12 Aug 1997 20:29:41 GMT
Links: << >>  << T >>  << A >>
On Tue, 12 Aug 1997 15:06:03 GMT, peter299@maroon.tc.umn.edu (Wade D.
Peterson) wrote:

>There are two companies (that I know of) here in town that are stuck
>buying all XILINX FPGA and PROMs.  If they buy even one part from
>somebody else like Altera, then XILINX will come in and increase their
>costs by 40% or more.  

So tell them that you have redesigned and are going to program the
FPGA from a micro. If they jack up the price, then design them out,
and (please) provide documentary evidence to this newsgroup. I'm sure
somebody would be happy to stick a few K of GIF's on a web site to
prove the point.

If you've got any documentary evidence, then why don't they make it
public. If it is true, you have nothing to fear, and Xilinx has
EVERYTHING to lose. The bad publicity would cost them a lot more than
it would cost you. If the right spin was put on it, it could cripple
them.

I personally would cease designing with any manufacturers product who
attempted to blackmail me in such a way. It's not like Xilinx is the
only vendor out there.

Maybe the perceived threat was from a desparate salesman? I still
can't believe it's Xilinx policy to blackmail customers with massive
price rises.

I know designing people out is pretty extreme, but as you assert that
Xilinx will put the price up, I assume you refer to direct business,
in which case the value should be high, in which case I'm sure it
would be economical to do so.

Stuart
Article: 7191
Subject: Re: Price of Serial EEPROM is Outrageous
From: daveb@iinet.net.au (David R Brooks)
Date: Tue, 12 Aug 1997 23:35:37 GMT
Links: << >>  << T >>  << A >>
peter299@maroon.tc.umn.edu (Wade D. Peterson) wrote:

[snip]
:The real problem is that if you use XILINX FPGA, but don't buy their
:PROMs, then XILINX starts jacking up their prices on everything you
:buy from them.  XILINX has probably the most preditory pricing that
:I've ever seen.  [I think it's downright illegal, but then again I'm
:not a lawyer].

Hmm, I have done several systems using Xilinx FPGAs, that didn't use
their ROMs, for the simple reason the designs included a CPU, and
FPGAs were programmed as slaves by the CPU. So Xilinx didn't lose a
EEPROM sale there: there never was one. I'm not aware we were gouged
on price for it (maybe the Australian distributors are more
reasonable?).
 Now, suppose I have FPGAs on the shelf (maybe over-bought for some
such project), and decide to use them in another project which _just
happens_ to use Atmel EEPROMs, what the **** business of Xilinx is it?

 All you need do (if they ask) is tell them the project uses
slave-mode configuration. I doubt they'll bother to reverse-engineer
your product to check up.


--  Dave Brooks <http://www.iinet.net.au/~daveb>
PGP public key: finger  daveb@opera.iinet.net.au
                servers daveb@iinet.net.au
    fingerprint 20 8F 95 22 96 D6 1C 0B  3D 4D C3 D4 50 A1 C4 34
 What's all this? see http://www.iinet.net.au/~daveb/crypto.html
Article: 7192
Subject: Re: Any one getting 125MHz out of XILINX CPLDs?
From: Graeme Robertson <graemer@rie.net.au>
Date: Wed, 13 Aug 1997 10:20:09 +1000
Links: << >>  << T >>  << A >>
Kayvon Irani wrote:
> 
> Hi:
>         I'm looking at Xilinx CPLDs for a new design and trying to match the
>         maximum clock frequency number  in the data book for XC95108-7. The
>         maximum clock frequency on a 16bit counter I get is 83 MHz their number
>         is 125 MHz. Can any one duplicate their results?
> 
>         Regards,
>         Kayvon Irani
>         Los Angeles, Ca


Hello,
I have recently had the same problem, although with a 10 bit counter in
an XC9572-7. I solved it by forcing the use of local feedback within a
function block. 
This can be done using the 'PARTITION' property if using ABEL HDL code
as described in XAPP072 or on P2-15 of the May 97 'ISP Application
Guide'.
Alternately an "experimental" product called Chipviewer can be found on
the Xilinx FTP site. This product allows you to produce a constraint
file for use in M1 that will place the counter elements physically where
ever you want it in the target device. (Buggy, but does work.)

Regards,
Graeme Robertson,
Melbourne, Australia
Article: 7193
Subject: Re: Any one getting 125MHz out of XILINX CPLDs?
From: Kayvon Irani <kirani@cinenet.net>
Date: Tue, 12 Aug 1997 22:05:17 -0700
Links: << >>  << T >>  << A >>
Terry Harris wrote:
> 

> 
> You can only use local feedback within a function block and pin
> feedback from actual output pins. 

	The Xilinx CPLD literature p2-9 shows a timing model of the device where
	the local feedback is shown to originate from the Q output of the flip-flop
	rather than output pins. Is it perhaps incorrect?

	Regards,
	Kayvon Irani
	Lear Astronics
Article: 7194
Subject: Re: Download FLEX10K over the LPT port
From: Steve Wiseman <steve@steves-house.org.uk>
Date: Wed, 13 Aug 1997 08:36:29 +0100
Links: << >>  << T >>  << A >>
no@spam J.Szamosfalvi wrote:
> 
> : Thomas Berndt (berndt@eas.iis.fhg.de) wrote:

> : : does anybody knows how I can download designs which are compiled with MAXPLUS2
> : : over the LPT port?
> 
> You need a byteblaster between the parallel port and a 10 pin download header
> in your circuit.   MAXPLUS2 will automatically generate the needed *.sof
> file for you.  You need a Windows PC to do a download.  Read the applicable
> application notes very thoroughly before design and pin assignment.

After needing to do this on a client site, I have a bit of Qbasic that
will 
launch a .rbf file through LPT1. It clashes with the Altera dongle and
Iomega 
parallel port ZIP drive driver, though. It's also slow, but if anyone
wants it, 
they're welcome to mail me. (It uses passive serial mode, and uses 4
connections 
plus ground. It seems reliable, but there's no guarantee. 
(If anyone knows how to get / to round _down_ to an integer result, or a
shift right 
operator in Qbasic, I'd be interested...)

Steve
Article: 7195
Subject: Re: Any one getting 125MHz out of XILINX CPLDs?
From: terry.harris@dial.pipex.com (Terry Harris)
Date: Wed, 13 Aug 1997 10:55:34 GMT
Links: << >>  << T >>  << A >>
Kayvon Irani <kirani@cinenet.net> wrote:

>Terry Harris wrote:
>> 
>
>> 
>> You can only use local feedback within a function block and pin
>> feedback from actual output pins. 
>
>	The Xilinx CPLD literature p2-9 shows a timing model of the device where
>	the local feedback is shown to originate from the Q output of the flip-flop
>	rather than output pins. Is it perhaps incorrect?

No thats right but local feedback only routes within that function
block which restricts fitting if you must use it. 

I presume pin feedback comes from output pins using the path that
would be used if the pin was an input. As far as I know pin feedback
has the same timing as input pins which is significantly faster than
non-local feedback. 

Cheers Terry...

Article: 7196
Subject: FPGA power consumption
From: David Bokaie <dbokaie@ix.netcom.com>
Date: Wed, 13 Aug 1997 09:01:10 -0700
Links: << >>  << T >>  << A >>
In our latest design we have notice that our Altera 7K FPGA is drawing
roughly 100ma of current.  Stopping the clock to the chip only reduced
the power consumption by 30%.  I was wondering if there is any study out
there that 

	1) explains the high power consumption of FPGA
	2) compare power consumption between different FPGA vendors
	   and different technology family.

Please send/or copy your responses to davidb@proxim.com

Thanks,

David
Article: 7197
Subject: Re: free FPGA software from actel
From: rstevew@armory.com (Richard Steven Walz)
Date: 13 Aug 1997 20:11:22 GMT
Links: << >>  << T >>  << A >>
In article <5snlch$47o@thorgal.et.tudelft.nl>,  <dstewart@dmicros.com> wrote:
>On 10 Aug 1997 15:00:59 GMT, rstevew@armory.com (Richard Steven Walz)
>wrote:
>
>>In article <33ebabe0.228220832@news.m.iinet.net.au>,
>>David R Brooks <daveb@iinet.net.au> wrote:
>>>Sam Falaki <Falaki@uqtr.uquebec.ca> wrote:
>>>
>>>:> >> Has somebody tested the free actel software? Is it worth to download?
>>>:> >> Is programmer support included? Is the ACTIVATOR needed or is there
>>>:> >> a free design for a simple programming hardware.
>>>:
>>>:I managed to compile VHDL and target a very wide range of devices
>>>:without even going through any tutorial or reading any "help"
>>>:whatsoever.  It's really easy to use.  You can kind of guess your
>>>:way along and next thing you know you're done.  I liked it.
>>>
>>> The software may be free (and good), but what else do you need to get
>>>working silicon? A big-bucks programmer, maybe? The bottom line is the
>>>cost of _all_ tools necessary.
>>>
>>> Please, will one of the SRAM-based FPGA vendors put out something
>>>similar? So we (the great impoverished <g>) can actually put their
>>>chips to work in real systems. 
>>>
>>>--  Dave Brooks <http://www.iinet.net.au/~daveb>
>>>PGP public key: finger  daveb@opera.iinet.net.au
>>>                servers daveb@iinet.net.au
>>>    fingerprint 20 8F 95 22 96 D6 1C 0B  3D 4D C3 D4 50 A1 C4 34
>>> What's all this? see http://www.iinet.net.au/~daveb/crypto.html
>>---------------------------------------------
>>Last I looked, Altera had bought Intel's FlexPLD's and called them FlashLOGIC
>>or such and they are FPGA's that are SRAM and EEFLash programmable with a
>>LPT-JTAG cable I have the plans for that uses just one '244. I don't do PLD
>>that big, so I haven't bothered, but that sounds like what you just asked
>>for. Yes? No? All you need to do is hunt down the command to dump the SRAM
>>to EEFlash, (they want to sell you software to do that one tiny bit, but it
>>is a well known code), and you got it!! I THINK I still have that info!
>>Write or call them! www.altera.com .
>>-Steve
>>-- 
>>-Steve Walz  rstevew@armory.com  ftp://ftp.armory.com:/pub/user/rstevew
>>-Electronics Site!! 1000 Files/50 Dirs!! http://www.armory.com/~rstevew
>>Europe:(Italy) ftp://ftp.cised.unina.it:/pub/electronics/ftp.armory.com
>>Oz:.AU ftp://ftp.peninsula.apana.org.au:/pub/electronics/ftp.armory.com
>Altera have announced that they are to discontinue the Flashlogic
>family,  so I wouldn't design them in.  I think the last time buy for
>these parts has already been done.  PLDShell,  which was the free
>software,  was crap anyway.
---------------------------
Ah! Thanks, important to know anyway. It seemed they were going to go with
it for a time. I figured they would get past rev 1 of software. I guess
they never did. Intel makes money again!
-Steve
-- 
-Steve Walz  rstevew@armory.com  ftp://ftp.armory.com:/pub/user/rstevew
-Electronics Site!! 1000 Files/50 Dirs!! http://www.armory.com/~rstevew
Europe:(Italy) ftp://ftp.cised.unina.it:/pub/electronics/ftp.armory.com
Oz:.AU ftp://ftp.peninsula.apana.org.au:/pub/electronics/ftp.armory.com
Article: 7198
Subject: Re: Should Xiling have more local clock nets?
From: lass@xilinx.com (Steve Lass)
Date: 13 Aug 1997 22:48:15 GMT
Links: << >>  << T >>  << A >>
: >With the older Xilinx devices, one could attach a L attribute to a
: >local clock net (forces it onto a long line), add a SC=1 (skew below
: >1ns) attribute, and that would do it. This was with the old (1991) APR
: >program. With the new devices (and using PPR) this doesn't work.

M1 does support the MAXSKEW constraint which can be used for clock nets.

Steve Lass
Xilinx Product Marketing
Article: 7199
Subject: Low-cost programming FPGAs (was: Re: free FPGA software from actel)
From: daveb@iinet.net.au (David R Brooks)
Date: Wed, 13 Aug 1997 23:30:59 GMT
Links: << >>  << T >>  << A >>
rstevew@armory.com (Richard Steven Walz) wrote:

[snip]
:Last I looked, Altera had bought Intel's FlexPLD's and called them FlashLOGIC
:or such and they are FPGA's that are SRAM and EEFLash programmable with a
:LPT-JTAG cable I have the plans for that uses just one '244. I don't do PLD
:that big, so I haven't bothered, but that sounds like what you just asked
:for. Yes? No? All you need to do is hunt down the command to dump the SRAM
:to EEFlash, (they want to sell you software to do that one tiny bit, but it
:is a well known code), and you got it!! I THINK I still have that info!
:Write or call them! www.altera.com .
:-Steve

 ISP has to be the way to go for this kind of thing. However Altera do
not (afaik) offer free design software. A colleague is using Altera
extensively, and confirms the "Bit Blaster" (ISP adapter) is just a
'244 sold at a high price.

 This seems to be the problem, everyone charges mega-$ either to
design, or to program the brutes.


--  Dave Brooks <http://www.iinet.net.au/~daveb>
PGP public key: finger  daveb@opera.iinet.net.au
                servers daveb@iinet.net.au
    fingerprint 20 8F 95 22 96 D6 1C 0B  3D 4D C3 D4 50 A1 C4 34
 What's all this? see http://www.iinet.net.au/~daveb/crypto.html


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