Site Home Archive Home FAQ Home How to search the Archive How to Navigate the Archive
Compare FPGA features and resources
Threads starting:
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
================================================================== CALL FOR PAPERS SPECIAL ISSUE ON RECENT ADVANCES IN FPGA TECHNOLOGY IEEE TRANSACTIONS ON VLSI SYSTEMS Field-Programmable Gate Arrays (FPGAs) have become an important technology for the implementation of VLSI circuits and systems due to the steady increase in FPGA density and speed. Aside from main-stream use in reasonably high volume production, the field- programmability and re-programmability of FPGAs enable many novel and promising applications, including rapid system prototyping, reconfigurable system designs, rapid system prototyping, circuit emulation, and reconfigurable custom computing. This special is- sue will be devoted to recent advances in all areas related to the FPGA technology. Topics of interest include, but not limited to: o Advances in FPGA architectures, including design of programm- able logic blocks, programmable interconnects, programmable I/Os, and development of new FPGAs and field-configurable memories. o Novel applications of FPGAs, including rapid prototyping, logic emulation, reconfigurable custom computing, and dynamically reconfigurable applications. o New CAD algorithms and tools for FPGAs, including new algo- rithms for sequential and combinational logic optimization, technology mapping, partitioning, placement, routing, and de- velopment of new FPGA synthesis or layout systems. o Advances in field-programmable technology, including new pro- cess and fabrication technologies, and field-programmable analog arrays. Authors of papers accepted for the 1997 ACM/SIGDA 5th Interna- tional Symposium on Field-Programmable Gate Arrays (FPGA'97) are especially encouraged to submit an extended version of their pa- pers for possible inclusion in the special issue. Submissions of relevant work not presented at FPGA'97 are also welcome. This special issue is being coordinated by Guest Editors Jason Cong of Univ. of California, Los Angeles (UCLA) and Carl Ebeling of Univ. of Washington. Prospective authors should submit postscript versions of their papers electronically using the www no later than March 31, 1997, and indicate they should be considered for this special issue: http://microsys6.engr.utk.edu/~tvlsi Alternatively, email may be used by contacting: tvlsi@microsys6.engr.utk.edu If neither of the electronic means is available to authors, trad- itional paper manuscripts should be sent to: IEEE Trans. on VLSI Systems c/o Prof. Bing Sheu Electrical Engineering Powell Hall, Room 604 University of Southern California Los Angeles, CA 90089-0271 ================================================================== +-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+ | 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: 5401
Yes, using an external RAM. Far cheaper, too. >I am currently designing 106 Bytes(848 bits)FIFO in XILINX 4000E. >Is there any other more efficient way rather than using RAM cell macros? Peter. Return address is invalid to help stop junk mail. E-mail replies to z80@digiserve.com.Article: 5402
>Doing such designs in schematic is a waste of designer's time. This is >so much easier accomplished in a HDL. Some people would differ. It may be easier for *you* (if you are a VHDL expert) but it sure as hell is harder for me if I am not a VHDL expert (and one does have to be quite good at VHDL to avoid getting into a right mess, which means using it a lot) or if I cannot afford to cough up $Xk for a new Xilinx FPGA design suite with VHDL capability. OTOH schematics are always readily translatable into any toolset, now or in 10 years' time. One just draws the circuit ("schematic capture" methodology, in the USA :) ) >Why would you want the 16450/16550 with all the overhead (and limited) >functionality? If you start with my compact UART, then you could easily >add a FIFO buffer if you need it, and your design will consume very >little FPGA space. Why does anyone use a 16550 with its "limited" functionality? The IBM PC also has a limited functionality, but for some reason seems to be quite popular nowadays. Let me tell you: a UART is very rarely useful without a microprocessor around. The micro needs to be able to set the baud rate. So one had a baud rate generator. It needs to get interrupts. So one has circuitry for that. And so on. I don't see that the 16550 has a particularly "limited" functionality. They should have provided IRQ capability from *all* the status bits, but again it would be easy to add that if one had a circuit. But as its free, one should not complain I suppose :) Peter. Return address is invalid to help stop junk mail. E-mail replies to z80@digiserve.com.Article: 5403
Peter wrote: > >I am currently designing 106 Bytes(848 bits)FIFO in XILINX 4000E. > >Is there any other more efficient way rather than using RAM cell macros? > Yes, using an external RAM. Far cheaper, too. It's served me well, but I'm not sure where the breakpoint is. For two cells of FIFO, you may be better off using an external 'real' FIFO, to get the pin-count down. Some cunning multiplexing ought to get you down to 8 data pins and a handful of strobes. Small synchronous FIFOs (IDTs Supersync?) make life easier, too. 32K*8 12/15nS SRAM is incredibly cheap, but don't expect to be able to sustain cycles at 80MHz.... (16MHz is comfortable, though, in ALtera 8k series, YMMV) Hope that helps, SteveArticle: 5404
Duncan Davis wrote: > I am convinced I read a posting from someone regarding the topic of > simulating at the gate level on Quicksim, having syntheised from Galileo. > I seem to remember it involved EDIF. I think it appeared about a month or > two ago. > > Does this ring any bells with anyone or can anyone comment on methods for > going from a VHDL entered design via, Galileo/Leonardo and simulating the > result. yes, i recall that thread from back last month. someone wanted to create an eddm db for a design synth'd by galileo using its edif output. enread wasn't able to read the edif properly (lib element problems as i recall). someone suggested a workaround - resyn in al2, then run sg (eg. write eddm) to get the schematic (should work ok) which can be sim'd. you should be able to find the thread by using dejanews. -- Lance Gin "Off the keyboard, over the bridge Delco Systems-GM Hughes Electronics through the gateway, C43LYG@dso.hac.com nothing but NET!"Article: 5405
In article fsf@manocska.bendor.com.au, Zoltan Kocsi <zoltan@bendor.com.au> () writes: > Hi, > > It was very great news that Exemplar released Leonardo for Linux. > The benchmark posted here also showed very pleasant results. > So I rang the local Exemplar supplier. > Now he told me that for a mere AUD37,000 I could have it (basic stuff). > I'm not (yet) that rich, I'm afraid. <SNIP> ... > > So, I wonder if any of those family-specific cheap tools would ever be > released under Linux ? The Exemplar release should be indicative that there's > a potential market. Is it really that hard to port a HP or a Sun version > to Linux ? Or, is it shameful to have a Linux version ? I don't think it requires much effort to port the tools, but there's still cost involved in paying people to support the platform. The Exemplar release is really just testing the waters to see if Linux users are really willing/able to pay for their tools. From your comment above, I guess not. It's a catch-22. Software vendors don't want to support a platform with few customers and customers aren't likely to switch platforms without software support. Generally, it's the software vendors who take the lead. ---------------------------------------------------------------------- Jeff Hunsinger jeffh@oakhill-csic.sps.mot.comArticle: 5406
Hi All, I was wondering if someone could point out the advantages/disadvantages and reasons between choosing between Mealy or Moore state machines. How does it affect the performance of the state machine, and when synthesised, what are the effects on the resulting circuitry? Thanks in advance for any advice, Cheers, Phil.Article: 5407
Peeters wrote: >even pci@xilinx.com never replied. I don't recall seeing your e-mail come through, so I don't know what happened. We are trying to phase out the pci@xilinx.com alias. Any technical questions, including PCI questions, should be addressed to hotline@xilinx.com. To answer your question, I'm not aware of anyone who is marketing a PCI prototyping board with a Xilinx chip as the interface. ------------------------------------------------------------------------ / 7\'7 Jim McManus mailto:jim.mcmanus@xilinx.com \ \ ` Xilinx PCI Applications Engineer 1-800-255-7778 (toll-free) / / 2100 Logic Drive 1-408-879-4729 Fax \_\/.\ San Jose, California 95124 http://www.xilinx.com/products/logicore/lounge/pci/pci.htm ------------------------------------------------------------------------Article: 5408
Hi there, I'm currently installing Designer 3.1 in a networked license and networked environment for ACTEL FPGAs on Sun-Workstations running Solaris. It seems to me, installation is a nightmare. Thus I would like to know if there is anybody out there having the same experience and may be some hints or workarounds. Copying data from CD-ROM is quite a nice job, but starting the license server is terrible. Any suggestions will be welcome. Thanx in advance. ----------------------------------------------------------------- name: Michael Koch __ ____ depart: Design Tools | \ |_______ company: IMS - \_________| adress: Allmandring 30a ____ 70569 Stuttgart | /\ /\ Germany | / \ / \ phone: xx49-711-685 5897 _|___/ _\/ _\ fax: xx49-711-685 5930 e-mail: koch@mikro.uni-stuttgart.de -----------------------------------------------------------------Article: 5409
CHRISTOS, CHECK OUT THE LINEAR RECURSIVE SEQUENCES TUTORIAL IN PDF FORMAT IN THE SPREAD SPECTRUM SECTION OF MY SITE AT: http://www.erols.com/aaps I wrote it several years ago for an article which appeared in RF Design magazine. It tells you the basics on generating the sequences and gives auto correlation trade offs for maximal and non maximal length codes. Also you might want to check out the PN SIM program which is an older DOS application based in Turbo Vision, but it has tables of at up to 100 different TAP weight settings for each family (register length) of sequences. http://www.erols.com/aaps/spsp.htm This takes you to the Spread Spectrum page. Also I am completing a VHDL example program which will run in the APS-X84 test FPGA board which will have a detailed analysis of writing PN/LRS/RNGs. It can be found on the same web server at: http://www.erols.com/aaps/prog.htm Then click the choice for the X-Files (X84 examples) I think you will enjoy the motiff as well as the up coming examples. I would be glad to send you some preliminary stuff if you are in a bind. Good Luck, Richard Schwarz John L. Smith <jsmith@univision.com> wrote in article <3300E873.6AEF@univision.com>... > Christos Dimitrakakis wrote: > > > > Anyone got any info on RNGs? > > Since I'm only generating a 6-bit number with it > > I could just use a simple counter scheme for it that runs @8Mhz > > while using one of the other clock outputs provided in the chip, > > say the 490Hz one, for the rest of the circuitry. > > Will that reduce the decorrelate the timing of the RNG from the > > rest of the chip? > > > > Is there any better way that does not take up a lot of chip area? > > [I considered using a Fibonacci series, but it takes up too much space > > if sufficient randomness is required] > > > > Peter Alfke at Xilinx put together a good app note on compact > long linear feedback shift registers for the 4000 family, > available in pdf @ > > http://www.xilinx.com/apps/counter.htm > > Use a long enough one, and you probably don't have to > worry about correlation. > > > > -- > John L. Smith > Univision Technologies, Inc. > 6 Fortune Drive > Billerica, MA 01821-3917 > jsmith@univision.com >Article: 5410
Peter wrote: > > >Doing such designs in schematic is a waste of designer's time. This is > >so much easier accomplished in a HDL. > > Some people would differ. > > It may be easier for *you* (if you are a VHDL expert) but it sure as > hell is harder for me if I am not a VHDL expert (and one does have to > be quite good at VHDL to avoid getting into a right mess, which means > using it a lot) or if I cannot afford to cough up $Xk for a new Xilinx > FPGA design suite with VHDL capability. I'm not an VHDL expert. I'm just learning how to exploit another language, the Altera Hardware Description Language (AHDL). I was also, until recently, a schematic addict, but it is only now I start to realize how much more power I have in such a language, it is extremely easy to learn and understand!. Also, I would not have been able to complete the design in schematics, because I don't have all the parameterizable features there. In AHDL, I can generate a fancy kind of register of any length based on a parameter input, so easy, so readable, so powerful! The Altera PLS-ES package shouldn't set you back by more than about $500. > Why does anyone use a 16550 with its "limited" functionality? The IBM > PC also has a limited functionality, but for some reason seems to be > quite popular nowadays. > > Let me tell you: a UART is very rarely useful without a microprocessor > around. The micro needs to be able to set the baud rate. So one had a > baud rate generator. It needs to get interrupts. So one has circuitry > for that. And so on. > > I don't see that the 16550 has a particularly "limited" functionality. > They should have provided IRQ capability from *all* the status bits, > but again it would be easy to add that if one had a circuit. The limitations of the 16550 is this: 1) You can't make an 11-bit UART. How do you do that? With programmable logic, of course! 2) Why would I want all the 16550 fancy registers to consume valuable space in my FPGA, when a basic functionality UART will do everything I need at a fraction of the space requirements? 3) By using the 16550, the interface is designed for a microcontroller. This make it harder to use if you want to interface with programmable logic instead. Of course, with a basic functionality UART you may have to twiddle some bits to get the interrupt conditions you want, but this really PEANUTS! You are working with programmable logic, so what's the problem?!? > But as its free, one should not complain I suppose :) I guess :) -- Rune Baeverrud <r@acte.no> Field Application Engineer |\ _,,,---,,_ ScandComp, ACTE NC Norway AS ZZzz /,`.-'`' -. ;-;;,_ P.O. Box 190, N-2020 Skedsmokorset, Norway |,4- ) )-,_..;\ ( `'-' Tel: +47 6389 8969 Fax: +47 6389 8979 '---''(_/--' `-'\_)Article: 5411
In case you didn't know, the latest version (9.0x) of Lucent (Orca) Foundry has the following minor bug: Epic (EditLCA equivalent :-) ) won't display the guts of the edited device if your Windows display is configured for more than 256 colours. This is a minor annoyance, easily circumvented. But you might spin your wheels lots looking for the answer (hint: it's not on the support website yet!). -- Bob **************************************************************** Bob Elkind mailto:eteam@aracnet.com 7118 SW Lee Road part-time fax number:503.357.9001 Gaston, OR 97119 cell:503.709.1985 home:503.359.4903 ****** Video processing, R&D, ASIC, FPGA design consulting *****Article: 5412
Steve Wiseman wrote: > > Peter wrote: > > > >I am currently designing 106 Bytes(848 bits)FIFO in XILINX 4000E. > > >Is there any other more efficient way rather than using RAM cell macros? > > > Yes, using an external RAM. Far cheaper, too. > > It's served me well, but I'm not sure where the breakpoint is. For two > cells of FIFO, you may be better off using an external 'real' FIFO, to > get the pin-count down. Some cunning multiplexing ought to get you down > to 8 data pins and a handful of strobes. Small synchronous FIFOs (IDTs > Supersync?) make life easier, too. 32K*8 12/15nS SRAM is incredibly > cheap, but don't expect to be able to sustain cycles at 80MHz.... (16MHz > is comfortable, though, in ALtera 8k series, YMMV) A vague recollection of the 7200 FIFO family from 1993: 7200 (256 x 9 @ 10MHz) = $3. @30MHz = $6. Regards, ScottArticle: 5413
1997 IEEE/ACM International Workshop on Logic Synthesis http://www.ee.princeton.edu/iwls97.html Granlibakken Resort, Lake Tahoe, California May 18-21 , 1997 Call for Participation Submission Deadline extended to March 1, 1997 (Papers submitted to IWLS are eligible for submission to ICCAD.) Contents 1. Synopsis 2. Benchmarks 3. About IWLS 4. About Granlibakken 5. Executive Committee 6. Technical Program Committee 7. Sponsored by... Synopsis Logic Synthesis has traditionally focused on optimization techniques for combinational and sequential circuits through the manipulation of Boolean equations and state machines. IWLS '97, the sixth workshop in this series, seeks presentations both on these topics and on new directions in synthesis-based design methodology. Topics of interest include: Area, timing, power optimization Logic synthesis systems CMOS, ECL, GaAS Optimization Designer Experiences with Synthesis Two-level Logic Optimization Interaction with physical design Multi-level Logic Optimization Incremental Synthesis/ECO Support FSM Optimization Asynchronous Logic Synthesis Sequential Circuit Optimization Formal Verification Retiming and resynthesis Optimization at the RTL Level Technology Mapping Timing Verification FPGA and PLD Synthesis Testing and Synthesis for test Don't-Cares and Boolean Relations Interaction with module generators Symbolic Synthesis Use of synthesis in new applications Synthesis in FPGA-Based Emulation Applications of SAT Authors may submit extended abstracts for their proposed presentation. These must be no less than 1000 words and no more than 2500 words. These abstracts are not intended to be complete papers, but rather should convey the main ideas of the proposed presentation. We encourage submissions in the early stages of research which may highlight important new problems without necessarily providing complete solutions. The abstracts may be submitted by e-mailing self-contained Postscript files to iwls-submit@ee.princeton.edu by March 1, 1997. Acceptance notices will be sent by March 31, 1997. A set of workshop notes will be distributed at the conference. There will be no published proceedings. Benchmarks A benchmark set is being assembled by the CAD Benchmarking Laboratory. To contribute new benchmarks, or to obtain information about the existing suite, please write: benchmarks@cbl.ncsu.edu. About IWLS IWLS '93 and IWLS '95 introduced a number of format changes from previous workshops, which the committee tentatively intends to maintain for IWLS '97. These include an open program with high acceptance rate, heavy use of posters and short talks for presentation, and large amounts of time in the schedule for poster presentations. In addition, IWLS '97 will emphasize open discussions and ongoing research which are not provided by the traditional conference format. About Granlibakken The Granlibakken Conference Center is located in Tahoe City on the west shore of Lake Tahoe, 180 miles east of San Francisco. It boasts 160 rooms, clustered into two- and three-bedroom condominiums. Each bedroom is an attractive hotel room with private bath. Many of the clusters share a kitchen, living room and dining room -- a miniature lobby for private meetings. Organizations sending several people to the workshop may wish to rent entire two- and three-bedroom townhouses. The Granlibakken management has reserved space on Thursday, May 22 for organizations who wish to hold private, one-day workshops immediately preceding IWLS, and have agreed to charge organizations the IWLS conference rate for these meetings. Contact Mary Brown at Granlibakken sales (1-800-552-4494) for details. Granlibakken is within 10 minutes' drive of the West's premier ski resorts: Alpine Meadows and Squaw Valley USA. When California enjoys high snowfall, both areas remain open until Memorial Day. A wealth of hiking trails snake through the area. Weather permitting, Granlibakken's tennis courts and pool will be open for use. The weather in late May is variable; warm, sunny days and cool clear nights are the rule. Getting There Granlibakken is easily reached from either the San Francisco Bay Area or Reno, NV. Take Interstate 80 to Truckee. From there, follow State Route 89 south to Tahoe City. Turn right at the stop light in Tahoe City. After 1/4 mile, turn right on Granlibakken road and proceed to the end. Contacts/Executive Committee General Chair Rick Cadence mcgeer@cadence.com (408) McGeer Berkeley Labs 428-5325 Tech. Program Sharad Princeton (609) Chair Malik University sharad@ee.princeton.edu 258-4625 Benchmark Franc (919) Chair Brglez NCSU brglez@cbl.ncsu.edu 248-1925 Conference Kris Cadence (408) Coordinator Lamanno Berkeley Labs krisl@cadence.com 894-2479 Technical Program Committee Pranav Ashar NEC Michel Berkelaar TU-Eindhoven Robert K. Brayton UC Berkeley Franc Brglez NCSU Giovanni de Micheli Stanford Srinivas Devadas MIT Ewald Detjens Mentor Graphics Antun Domic Cadence Masahiro Fujita Fujitsu Laboratories of America Wolfgang Kunz University of Potsdam Luciano Lavagno Politecnico di Torino/Cadence Berkeley Labs Ken McElvain Synplicity Rick McGeer Cadence Berkeley Labs Sharad Malik (chair) Princeton University Shin-ichi Minato NTT Massoud Pedram USC Richard Rudell Synopsys Tsutomu Sasao Kysushu Institute of Technology Gabriele Saucier INPG Ellen Sentovich Cadence Berkeley Labs Fabio Somenzi University of Colorado Leon Stok IBM TJ Watson Research Center Sponsor Sponsored by the IEEE Computer Society, Technical Committee on VLSI. In co-operation sponsoship by ACM/SIGDA is being sought. -- Sharad Malik sharad@ee.princeton.edu Associate Profesor 609-258-4625 Dept. of Electrical Engineering 609-258-3745 Fax Princeton University http://www.ee.princeton.edu/~sharadArticle: 5414
Hello, If anyone is using Lucent Orcas with a download cable or with EPROMS could you please answer a couple of question. 1) The pin-out of the download cable is not consistent between the manual and the actual device. Any suggestions ? 2) What should the baud rate be ? Regards John Smith University of SussexArticle: 5415
Hello, I was annoyed by the price of the bitblaster but he had no other choice. However, I think, my bitblaster is dying so I would prefer to use your program that being ripped-off twice. Can I have a copy please. Why do you load the *.TTF file and not the *.SBF file ? Thanks Iakovos Stamoulis University of SussexArticle: 5416
Lance Gin wrote: > > Duncan Davis wrote: > > > I am convinced I read a posting from someone regarding the topic of > > simulating at the gate level on Quicksim, having syntheised from Galileo. > > I seem to remember it involved EDIF. I think it appeared about a month or > > two ago. > > > > Does this ring any bells with anyone or can anyone comment on methods for > > going from a VHDL entered design via, Galileo/Leonardo and simulating the > > result. > > yes, i recall that thread from back last month. someone wanted to create an > eddm db for a design synth'd by galileo using its edif output. enread wasn't > able to read the edif properly (lib element problems as i recall). someone > suggested a workaround - resyn in al2, then run sg (eg. write eddm) to get > the schematic (should work ok) which can be sim'd. > > you should be able to find the thread by using dejanews. > > -- > > Lance Gin "Off the keyboard, over the bridge > Delco Systems-GM Hughes Electronics through the gateway, > C43LYG@dso.hac.com nothing but NET!" We got dumped with Galileo by Mentor when they took away FPGA capability with Autologic. After checking Galileo out we decided that it was unacceptable. A few of the main capabilities we had with Autologic were not available in their (or Exemplar's) Galileo tool. No hierarchy, etc. So we decided to go with Synopsys FPGA tools. If you plan on using the latest Mentor software (B.3) then you can't use Autologic to resyn because FPGAs are no longer supported. The problem i saw with Galileo EDIF output was that the component names are uppercase. You end up with something like cell INV I believe a quick perl script could lowercase the component name and after setting a link to the xilinx libs the EDIF will read in successfully. Before the link is created i get can't find library xi5. Set the link ln -s /home/xxx/xilinx_5.2.1/xc5200 xi5 (replace /home/xxx/... with correct path, i am using 5200 library) Run the attached perl script on your EDIF file and save it away to a new file. edif_fix yourEDIFfile > newEDIFfile Then run enread on the new file. setenv MGC_WD `pwd` enread newEDIFfile This will create an eddm database which you can run schematic generator on to create the schematics and then you can invoke quicksim on the design. James Cleary CAD Application Support Wizard >>>>>>>>>>>> filename="edif_fix" #! /usr/local/bin/perl $edif_file = shift; open(INFILE,$edif_file); while (<INFILE>) { $line = $_; @temp = split(/\(/,$line); foreach (@temp) { if(grep(/cellRef/,$_)) { $_ =~ s/cellRef//g; $_ =~ s/\s//g; $_ =~ tr/A-Z/a-z/; $line =~ s/cellRef \w+/cellRef $_/; } elsif (grep(/cell /,$_)) { $_ =~ s/cell//g; $_ =~ s/\s//g; $_ =~ tr/A-Z/a-z/; $line =~ s/cell \w+/cell $_/; } } print $line; } close INFILE;Article: 5417
Actually, you're wrong. I already posted info from Intel's FlexLogic data sheet that stated their FlexLogic parts were FLASH-based. Altera just bought them that way. If you have data to the contrary, please post it. Wayne In article <01bc16d5$34c4f760$97e12399@default>, "Rhondalee Rohleder" <Rhondalee_Rohleder@msn.com> wrote: >Actually, you're *both* right. The devices available in the Intel >FLEXlogic line at the time it was sold were EPROM-based. But Intel had >planned from the outset to convert the line to FLASH, beginning with the >introduction of subsequent devices -- Altera just followed through on >Intel's plans after acquiring the line. At the same time, the line was >renamed FLASHlogic to avoid confusion with FLEX. > >Rhondalee Rohleder >Pace Technologies (Scottsdale, AZ) > >Dean Brown <deankb@popd.ix.netcom.com> wrote in article ><32F2B2A7.568F@popd.ix.netcom.com>... >> Wayne Turner wrote: >> > >> > In article <5cs4oq$cf5@borg.svpal.org>, garyk@svpal.svpal.org (George >Noten) wrote: >> > >Wayne Turner (waynet@goodnet.com) wrote: >> >> > 3. Converted? What the hell are you talking about? It was FLASH when >Intel >> > owned it, it was just called FlexLogic. When Altera bought it they had >to >> > change the name because they already had the FLEX 8000 family and >didn't want >> > it to be confused. So they named the part according to what it really >was: >> > FLASH. Hence the name FlashLogic. >> >> The Intel part was EPROM based, Altera re-engineered it to Flash, then >> dropped it less than a year later. The Flash based EPX880 was a direct >> drop in replacement for the IFX780, so customer support shouldn't be >> based on the 880 part but on the 780 as well which was available for 4-5 >> years. >>Article: 5418
In a previous article John Smith <tepa1@solx1.susx.ac.uk> writes: : ;Hello, : ;If anyone is using Lucent Orcas with a download cable or :with EPROMS could you please answer a couple of question. ; :1) The pin-out of the download cable is not consistent ; between the manual and the actual device. Any : suggestions ? My download cable works. Pinout is: 1 Vdd 2 CCLK 3 DATA 4 DONE 5 PGM- 6 GND 7 NC :2) What should the baud rate be ? 9600 works just fine.Article: 5419
Can anyone answer some questions I have about programming a Xilinx XC4003A FPGA board? I'm working on a project that would allow students to program a Xilinx board remotely over a network. Without going into the gory details of the design we're using, we have come to the conclusion that we cannot use the Xilinx XChecker cable or XChecker (the download program) to do the downloading of bit files to the board. We were able to get some xchecker replacement code that works with XC3000 and XC2000 series FPGA boards written by Scott Kroeger. Unfortunately we have been unsuccessful in programming XC4000 boards with that code. In the rather meager documentation we have we've uncovered some information about the cable connections and programming cycle...but not nearly enough. Does anyone know of in depth documentation on programming xilinx boards, making your own download cables or anything else even remotely related to what's in this post? Thanks in advance, -David Hirschfield PS: The web page for the project is at: http://srv.res.cmu.edu/xilinx/ +-===========================================================================-+ | --== e-mail ==-- | | _/_/_/ _/_/ _/ _/ _/_/_/ dch@andrew.cmu.edu | | _/ _/ _/ _/ _/ _/ _/ or | | _/ _/ _/_/_/_/ _/ _/ _/_/ dhirschfield@giss.nasa.gov | | _/ _/ _/ _/ _/_/ _/ | | _/_/_/ _/ _/ _/ _/_/_/ --== WWW ==-- | | http://srv.res.cmu.edu/~dch/ | +-===========================================================================-+Article: 5420
I'm trying to build an xchecker cable replacement and an xchecker program replacement for programming Xilinx XC4003A Demo boards, I have some documentation on the pin connections needed but I need some clarification. The documentation says that for the XC4000 the programming/downloading process is as follows: Raise the PROG line, wait some indeterminate time, and then lower it to begin programming the board. Toggle the CCLK line while feeding data to the DIN line until all the bits in the .bit file have been sent. If everything went OK, the DONE line should go high, signaling that the board got the program. Have I got this routine right? Are all my high/low signals correct? Does anyone have more information on this process? Specifically, does anyone know exactly what timings and settings the board requires? Thanks for any help, -David Hirschfield +-===========================================================================-+ | --== e-mail ==-- | | _/_/_/ _/_/ _/ _/ _/_/_/ dch@andrew.cmu.edu | | _/ _/ _/ _/ _/ _/ _/ or | | _/ _/ _/_/_/_/ _/ _/ _/_/ dhirschfield@giss.nasa.gov | | _/ _/ _/ _/ _/_/ _/ | | _/_/_/ _/ _/ _/ _/_/_/ --== WWW ==-- | | http://srv.res.cmu.edu/~dch/ | +-===========================================================================-+Article: 5421
If you are reading this now, you, like me, are badly in need of money. Believe me, I understand. If the United States had debtor's prison, I'd have been in it years ago. However, there is a way out. All you have to do is use the follow program. You can make HUNDREDS, if not THOUSANDS OF DOLLARS. Mail a single dollar bill to each of the following people, along with a note which says: "Please add me to your list. My name and address is (include your name and address)." (1) Ralph 7439 Hwy 70 S. #268 Nashville, Tn. 37221 (2) Bob Sullivan 4553 Atascadero Dr. Santa Barbara, Ca. 93110 (3) Bill Gray 8114 SW 104th Way Gainesville, FL 32608 (4) Y. Yuksel 1561 Mesa Dr. #123 Santa Ana, CA 92707 (5) J. M. Little Clemson University P.O. Box 3669 Clemson, SC 29632-3669 Then copy this message, and re-post it to as many newsgroups as you can, but with the following changes: Erase the first person's address. Move each of the remaining addresses up one number. (#2 becomes #1, #3 becomes #2, etc.) Add your name and address as the fifth one. What will happen is this. Others who, like you, are in need of money, will read your message, just as you are reading this one, and decide that they need the extra cash. They will send you their one dollar bill, and repost the message, making the same changes you did. You will still be on the list as #4. Someone will read their message, send money to everyone on the list, including you, and post another message, making the changes. You will still be on the list as #3. This will continue until you are off the list, but by the time you are off the list, AT LEAST five people will send you money. There's the five dollars you sent. I cannot confirm them, but there are continuous statements about how people have made hundreds, thousands, and even tens of thousands of dollars with this plan. Remember, the more places you put this, the more money you make. Does this seem unreasonable? It shouldn't. There are millions of people in the United States just like you, who need money, and are looking for a solution. Thousands of people are getting on the internet for the first time, every day, and each of them wants to make money. Is it so unbelievable that they, like you, will follow this plan, hoping for a big return? At most, it will cost you ten dollars: one dollar for each person, and five for the pen, stamps, and envelopes. Is it impossible for you to imagine there are ten people somewhere in America who want to make money? If there are, then you've only lost a little time. But in fact, there are MILLIONS of people who NEED the money, and they are looking for ANY chance to make some. You will more than make up for any time or money you spend. Is this program illegal? No. There's nothing illegal, or even shady, about this. Do you want proof? 1. The inclusion of the message "Add my name to your list." makes this a service, just like any TV advertisement or 1-900 number. Are those illegal? 2. Does the federal government police the internet? No. Do they open your mail? No. THAT would be illegal. 3. If this were illegal, do you really think I'd be dumb enough to give the police my address? Please, support this program. There is the potential for both you and I to make a LOT of money together. Thank you.Article: 5423
bergeon wrote: > > Can someone tell me where I can find a structure of a system which can do a > 1/T > (T is in BCD) with registers. > > Yves B. To provide both frequency and period readings from a counter, one can simply swap the count and gate inputs. For example, with an unknown input clock and a 1 second gate, you get frequency in Hz. With a fixed 1 MHz clock and an unknown gate (divided by two to get a constant high for the entire period) you get the reciprocal period (in microseconds). Perhaps your problem can be similarly restructured to avoid division? (the use of T for the quantity to be inverted seemed suggestive) regards, tomArticle: 5424
James Cleary wrote: > We got dumped with Galileo by Mentor when they took away FPGA capability > with Autologic. After checking Galileo out we decided that it was > unacceptable. A few of the main capabilities we had with Autologic were > not available in their (or Exemplar's) Galileo tool. No hierarchy, etc. > So we decided to go with Synopsys FPGA tools. Did you guys consider Leonardo? It does support hierachical designs and a lot more bells and whistles than Galelio. Regards, Kayvon Irani Los Angeles
Site Home Archive Home FAQ Home How to search the Archive How to Navigate the Archive
Compare FPGA features and resources
Threads starting:
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