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
It depends on what you mean by cheap. APS offers a complete XILINX development kit which will take you from synthesis through routing and simulation and into an actual test board for as little as $650.00!!!! That's alot of bang for $650.00. See the details at http://www.erols.com/aaps Christos Dimitrakakis <mbge4cd1@afs.mcc.ac.uk> wrote in article <33785149.62F8@afs.mcc.ac.uk>... > Is there a cheap way to develop for FPGAs, suitable for the enthusiast? > > -- > Christos Dimitrakakis > --------------------- > mailto:mbge4cd1@fs4.eng.man.ac.uk > mailto:mbge4cd1@afs.mcc.ac.uk > http://www.man.ac.uk/~mbge4cd1 >Article: 6351
I've been thinking of using an FPGA as a trigger circuit in a (not so high-end) digital scope. Functions I need are high speed comparison of two 10-bit digital values against each other and against predetermined values. Output would be an indication if to trig or not. I'm not familiar to FPGA's or other 'programmable logic' so I would like to know if this is possible to do, before I go any further. With high speed I mean about 10ns from value input to indication output. Suggestions are highly appreciated! ---------------------------------------- Svenn-Ivar Svendsen, Student Dept. of Electrical Engeneering sis@hekta.iet.hist.no http://hekta.iet.hist.no/~sisArticle: 6352
Ralf, In response to PCI interface within an FPGA, you must be careful. I have looked as Altera, Xilinx, and Lucent (Orca), and specs can be misleading. After doin some digging, I found the ORCA to be the only one that was Fully Compliant. They have a pretty good PCI kit and their architecture is very conducent to PCI implementations. I have alot of experience with Xilinx, and ORCA and have found ORCA and the tools in general to be better and easier to use. TonyArticle: 6353
Ok.. yes everyone counts gates differently at least until a year or so ago. Altera wrote a white paper (it is on their site) on how they count gates. Well sinc then Xilinx(at times) and Lucent have been using the same counting techniques. The counting assumes the following: 70% logic, 30% ram, then they add 4 gates per ram bit and count it all up to get the total gate count. The latest data books from vendors have been showing this. For Example: Lucents Orca book shows a range of gates for a given device assuming no ram, to 30% ram. Thsie 2C40A device which is tauted to have 40k gates, now has an advertised range of 38k to 98k gates. The second killer in FPGA's is the routing resources. Even if an FPGA has tons of gates, if you can't get to them its useless, so when looking at gates, or even counting up flops, you MUST also consider the routing resources. Many of us remember the key number of 65% utilization is an certian FPGA!!! Now as far as ASIC versus FPGA gates, I have a general rule of thumb that an FPGA will require about 2.5 x Asic gates. This is very dependent upon tge FPGA architecture and the design. Some FPGA;s have narrow LUT functions, which can increase the required gates.. TonyArticle: 6354
I have used Lucent;s ORCA in a PCI application with very few problems. Both the Bus Initiator/Target and stand alone Target kits were pretty easy to get working. And this kit is the ONLY one that is "really" PCI compliant. Caution.. Read the specs carefully and compare teh data books.Article: 6355
Svenn-Ivar wrote: > I need are high speed comparison of > two 10-bit digital values against each other and against predetermined > values. I mean about 10ns from value input to indication output. Look-up-table based FPGA are ideally suited to perform this function very efficiently, but 10 ns is hardly possible without pipelining. But I think your application allows pipelining without any penalty. With pipelining this is very easy ( "piece-of-cake). I would suggest the fastest flavor of the XC4000E. The smallest part ( XC4003E ) is more than big enough for your function. Hope to see you at our seminar in Oslo, June 6. Peter Alfke, Xilinx ApplicationsArticle: 6356
Karl E. Vinacco wrote: > > Hi, > The company that I work for is in the process of deciding on a new > simulation tool. Right now we use a schematic entry/ABEL-HDL tool, but > we want to upgrade to a VHDL or Verilog. I have been left with the task > of deciding which is better. I could use some input on the pros and > cons of each, maybe listing some of the major benefits or drawbacks. > Also what seems to be the most widely used. Any help I can get would be > greatly appreciated. Please reply to me as well as the newsgroup, and > no spam please. Thanks I use Model Technology's Vsystem Plus, (http://www.model.com) which supports both VHDL and Verilog, and even allows mixing the two. As I have no experience with a different package and am pretty new at this game anyway, I can't give you a comparative review of this software package versus others. But at least I thought that getting this package would eliminate having to decide between the two languages. As a computer scientist, I personally think that these languages are both pretty clunky, but I guess there isn't a better alternative that's widely accepted right now. The company claims they sold more copies of similar software than anyone else. So maybe that means that the software is OK? At least, it gives you a lot of different windows for looking at different things. Vitit Kantabutra Idaho State University vkantabu@howland.isu.eduArticle: 6357
zibby sobota wrote: > > Does anyone know if I can use xilinx xblox with orcad capture ver 7.00 ? > It seems that only Orcad Express can do the job but I am not sure . > Has anyone had any experience with that ? > > please help > > Zbigniew Sobota > Igt Australia XBLOX is being dumped by Xilinx, so avoid it until you know what, if anything, will replace it. Better yet, look into synthesis for those design components that you were thinking of using XBLOX for. -- ===================================================================== William Lenihan lenihan3we@earthlink.net "The greatest barrier to communication is the delusion that it has already occurred." -- Peter Cummings =====================================================================Article: 6358
"William E. Lenihan III" <lenihan3we@earthlink.net> writes: >XBLOX is being dumped by Xilinx, so avoid it until you know what, if >anything, will replace it. Better yet, look into synthesis for those >design components that you were thinking of using XBLOX for. I would hope that they're moving towards implementing a standard LPM library. Anyone know for sure? -- Paul Secinaro (pss1@christa.unh.edu) Synthetic Vision and Pattern Analysis Laboratory UNH Dept. of Electrical and Computer EngineeringArticle: 6359
The APS-X84-FB Lab Kit is available NOW with: 1) Schematic Capture Synthesis 2) Simulator 3) Router Software 4) Serial Xchecker cables 5) Parallel Xchecker Cable 6) 84 pin PLCC X84 board with ISA BUS connector on board 8255 etc. 7) XABEL PLD software 8) One years XILINX Foundation maint. and upgrades (this would include M1) 9) Examples and Lab Book including not only FPGA/PLD design, but C control code for using the design in a real world control environment (PC). Purchasing this kit will give you all you need to start designing for the soon to be released XILINX Lab Kits (which do not include any hardware or software upgrades). Get a jump on designing and using these kits NOW with the APS-X84-FS kits. Special rates are available for accademic users! XIlinx is really working hard (and smart) to provide low cost solutions to get starting engineers into FPGA kits. APS is pleased to offer our low cost solutions to engineers, instructors, and students which enable real cost effective tools for the industry leading XILINX FPGAs and PLDs at reasonable rates. Check out the APS family of kits for both low cost development and for more serious VHDL and full FPGA chip capable solutions at: http://www.erols.com/aapsArticle: 6360
Bharat Kurani (Bharat.Kurani@add.ssw.abbott.com) wrote: : I need address/phone/fax list of all semiconductor companines Then you should be looking in the Thomas Register. Check your local library. : Thank you : bharat@antrix.com : Bharat.Kurani@add.ssw.abbott.com -- #===================================================================# | John Lundgren - Elec Tech - Info Tech Svcs. | spamfree@mail. | | Rancho Santiago Community College District | rancho.cc.ca.us | | 17th St at Bristol \ Santa Ana, CA 92706 | http://www.rancho| | My opinions are my own, and not my employer's. | .cc.ca.us | ! You MAY NOT use my email address for unsolicited Email or lists! ! #======P=G=P==k=e=y==a=v=a=i=l=a=b=l=e==u=p=o=n==r=e=q=u=e=s=t======#Article: 6361
We have not had any direct experience with them yet, but they have been in to see us many times. The fine grained arch. which they use is appealing, but we don't really know if that will translate well into better designs. They have large relatively cheap chips which can be reprogrammed, but not in circuit without using one their programmers (which comes with the development kit). I understand they are working on getting BP and DATA IO on board as programmers, and that they are working on a true 5volt in circuit programmable chip. We also do not own a suitable synthesis tool which they support. They only supported Synopsis, and now work with Leanardo form Exemplar. We have Gallileo from Exemplar and would need to upgrade. Over all they have a unique solution for ASIC development in FPGAs, with their FLASH based chips, and sound like they are moving in the right directiion. They also expressed a willingness to implement big HDL designs for users for free on a case by case basis. This is probably the route we will take. Richard -- ______________________________________________ Richard D. Schwarz, President Associated Professional Systems (APS) FPGA Solutions/Test Boards/ EDA Software SIGTEK Spread Spectrum & Comm. Equipment 3003 Latrobe Court, Abingdon, Maryland 21009 Phone: 410-569-5897 Fax: 410-661-2760 Email: aaps@erols.com Web site: http://www.erols.com/aaps Terry Spoon <tspoon@pacbell.net> wrote in article <3377EA15.61BF@pacbell.net>... > Does anyone have design experience using Gatefield ProASIC family > design tools? VHDL->Synopsys->Gatefield->Quickhdl ? >Article: 6362
There may be an easier way than understanding the .bit file format. For example, you can create something called a "raw bits" format using the command line "makebits -r <filename>". This creates a 1's and 0's file formatted like it appears in the device. You can fairly easily see the pre-amble, length count, dummy bits, and frames. There is also a fairly useful utility on the Xilinx web site called 'makesrc' (make source). The makesrc program coverts an MCS file into a file with the format the way you want it. Its somewhere in their support section. I would provide the URL, but for some reason, my ISP provider is having problems reaching the Xilinx site at the moment. -- Steven Knapp OptiMagic(tm) Logic Design Solutions E-mail: optmagic@ix.netcom.com Programmable Logic Jump Station: http://www.netcom.com/~optmagic Archer Lawrence <archer@tanisys.com> wrote in article <337C7534.BBA805F@tanisys.com>... | I am looking for the file format for the '.bit' file produced from the | Xilinx makebit program. The documentation describes the configuration | stream format, but not this file. Thanks in advance | | Archer Lawrence | Sr. Engineer | Tanisys Technologies | |Article: 6363
| And this kit is the ONLY one that is "really" PCI compliant. | | Caution.. Read the specs carefully and compare teh data books. I hate to disagree, but I believe that the Xilinx LogiCORE interface for their XC4013E device is also fully PCI compliant, both electrically, and to the PCI protocol. I know that they had some problems with their 1.0 release, but the 1.1 release has resolved the problems (I used to work for Xilinx back then and helped create the 1.1 release VIEWsim testbench). Xilinx did two independent verifications. One with a commercially-available PCI bus model written in VHDL (http://www.vchips.com/products/pcitest.htm) and another with an independently-produced VIEWsim testbench. The results from both systems were cross-checked to avoid errors. Various customers also did their own testing with Synopsys PCI testbench (http://www.synopsys.com/products/bus/pcibus_ds.html) and had similar results. You can judge the compliance for yourself by downloading the PCI Protocol Compliance Checklist at 'http://www.xilinx.com/products/logicore/lounge/pcim/docs/pcicompl.pdf'. The interface is also shipping in production systems today. -- Steven Knapp OptiMagic(tm) Logic Design Solutions E-mail: optmagic@ix.netcom.com Programmable Logic Jump Station: http://www.netcom.com/~optmagicArticle: 6364
Electronics At Work, Exhibition, Conference & Workshops Melbourne Convention Centre Monday, 16th to Thursday, 19th of June 1997 Conference: Telecommunications: Forecasting the Future Workshops: 1. EMC Workshop/Update 2. Soldering Workshop 3. Software Workshop Exhibition: "Australia's only Exhibition dedicated to the Electronics Industry" Australia's leading electronics suppliers and manufacturers will be displaying their products. Free Admission. For further details and to register: http://www.aedc.com.auArticle: 6365
Richard Dungan wrote: > > Hello everyone, > > I'm thinking about buying the Actel "Designer Series" FPGA design > package. Can anyone provide any comments or details of practical > experience with this? I have prior experience with CUPL up to > moderately complex devices (Atmel ATV2500) but I'll be a newbie to > very large FPGAa and also to VHDL. > The intended first application is _very_ register-rich, hence the > choice of Actel. > > Thanks in advance, > > Richard > > ------------Richard Dungan------------- > Radix Electronic Designs, Orpington, UK > Email> Richard.Radix@BTinternet.com > --------------------------------------- Hi, I've been using Viewlogic/Actel for about 4 1/2 years so have some experience with this combination. Note that there are many good chips out there and many good s/w packages for many of the vendors each with their own strengths and weaknesses; because of some rather specific requirements, most of the work that i do is in Actel. As for register rich, if you select Actel, you might want to look at the newer stuff they put out like the 32300dx; this one has their maximum # of storage elements: ~2,000 dedicated flip-flops; ~2,000 more combinational modules which can be used for storage; (1 c-module for a latch, 2 c-modules for a edge-triggerred flip-flop); 250 i/o modules with an input and output latch in each module; they can be used either individually or combined (via a h/w macro) with a single c-module into a flip-flop). ~3,000 bits of SRAM which can be organized in a variety of ways (e.g. FIFO, dual-port RAM, etc.) the densest device that i have hands on experience with is their a32200dx which is one size smaller. as for the software, here are some comments: the older version had a lot of trouble keeping the pins fixed; you had to remember how to run the s/w in non-obvious ways - this was a real pain and is finally fixed. some older versions of s/w wouldn't lock out the probe pins and forced manual assigning. this is typical of the types of bugs seen most often. actel removed their 'chip edit' program which allowed you to observe and change module placement with a gui interface; it's back in the latest version with some of the features removed (and which should be back in in future versions). most tightly coupled to the viewlogic s/w - a lot of people swear at it, i've heard few swear by it. personally, i think it's ok, has some real nice features, licensing is a pain in the butt. you can get the whole system bundled together. they use a database system in the newer versions which is a pain to install and get running correctly and sometimes crashes; i believe this database is history in the soon to come out upgrade version. they give you a free vhdl compiler; this seems 'ok' at best. my vhdl is a bit limited and i am using it more in my design flow; the actel stuff is supported by other synthesis tools - you may want to look at them and have heard some good things about third party synthesizers. it takes a pretty good sized 'puter to run their s/w - i would recommend a 166 mhz pentium with 32 meg of ram. can't talk much about the unix requirements (both actel and viewlogic are on unix). the place and route s/w is FANTASTIC. s/w always routes even for dense designs; for 'production' designs i've used it up to > 90%; for evaluation circuits i use it close to 100%. for many 'production' circuits i am often hit with changes in requirements which entail significant logic redesigns; i have never had a problem keeping the pinouts fixed and feel quite confident that it will complete successfully with no significant changes in timing. i consider this to be a critical asset for fpga's since requests often come in too late for redoing boards and being able to reprogram without modifying the boards is extremely valuable. they also have incremental place and route which is handy for small changes. additionally, there is an option for timing driven layout which will run based on user input requirements. another good package they have is a macro generator. this has a wide variety of macros for standard building blocks with a number of configurations to optimize for speed or area or functionality or structure. the designs are very tight and it will be difficult for a human to beat it; in any event, it would take a lot of design time to beat their generator for at best a very small improvement. this package has been a very large contributor to getting designs done faster and more accurately than manual schematic entry and allows me to concentrate on architecture and algorithms. a nice bonus is (at least w/ viewlogic) automagic symbol generation so blocks can be generated and then placed down on a top-level, road-map schematic very quickly. static timing analyzer: this tool is a big labor saver in timing analysis and eliminates the need to generate worst-case stimulus vectors for back-annotated timing analysis. also, it can automagically compute the setup and hold times for external signals. you can dial in whatever conditions you want (i.e., temp, process, voltage, rad levels, etc.). action probe: this s/w allows you to bring off chip any two internal signals in a running system. you can either probe the signals using a scope, counter, logic analyzer, etc., or display values or vectors directly on the screen. this observation doesn't require any reprogramming and operates w/out interfering with the running logic (assuming that you didn't assign any signals to the probe pins - s/w will lock them out for you). debugger: this tool allows you to either manually or from a file stimulate a chip on the programming head and make observations from either at the chip's outputs or any internal node in either a bit basis or a vector's. just a few thoughts, hope it helps, rkArticle: 6366
Qualis Design Corporation has released the Fall schedule for our many hands-on, application-focused courses in Verilog- and VHDL-based design. Our courses are like no other -- just take a look at our lineup: Verilog System Design --------------------- Introductory: High Level Design Using Verilog (5 days) System Verification Using Verilog (5 days) Verilog for Board-Level Design (5 days) Elite: ASIC Synthesis and Verification Strategies Using Verilog (5 days) Advanced Techniques Using Verilog (3 days) Verilog Synthesis ----------------- Introductory: Verilog for Synthesis: A Solid Foundation (5 days) Elite: ASIC Synthesis Strategies Using Verilog (3 days) Behavioral Synthesis Strategies Using Verilog (3 days) For more info on our suite of HDL classes, to review our Fall schedule, or if you're interested in an on-site class, check out our web site at http://www.qualis.com, or send us e-mail at mailto:hot@qualis.com . You can also call for more info at +1.503.644.9700 . Qualis Design Corporation 8705 SW Nimbus Suite 118 Beaverton OR 97008 USA Ph: +1.503.644.9700 Fax: +1.503.643.1583 Copyright (c) 1997 Qualis Design CorporationArticle: 6367
Qualis Design Corporation has released the Fall schedule for our many hands-on, application-focused courses in VHDL- and Verilog-based design. Our courses are like no other -- just take a look at our lineup: VHDL System Design ------------------ Introductory: High Level Design Using VHDL (5 days) System Verification Using VHDL (5 days) VHDL for Board-Level Design (5 days) Elite: ASIC Synthesis and Verification Strategies Using VHDL (5 days) Advanced Techniques Using VHDL (3 days) VHDL Synthesis -------------- Introductory: VHDL for Synthesis: A Solid Foundation (5 days) Elite: ASIC Synthesis Strategies Using VHDL (3 days) Behavioral Synthesis Strategies Using VHDL (3 days) For more info on our suite of HDL classes, to review our Fall schedule, or if you're interested in an on-site class, check out our web site at http://www.qualis.com, or send us e-mail at hot@qualis.com. You can also call for more info at +1.503.644.9700 Qualis Design Corporation 8705 SW Nimbus Suite 118 Beaverton OR 97008 USA Ph: +1.503.644.9700 Fax: +1.503.643.1583 Copyright (c) 1997 Qualis Design CorporationArticle: 6368
I think, that you have a problem with 10K programming memories. EPC-1 PROM is available only in DIP or PLCC -packages (too thick to fit Type II frame). I've used FLEX8282(TQFP100) and EPC1064 (TQFP32) programming memory. This concept has been working fine. ------------------------------------------------------------- Markku Vahataini email: markku@fincitec.fi Fincitec Oy PO.BOX 11 PH: +358 16 221490 FIN-94601 KEMI FAX: +358 16 221561 FINLAND http://www.fincitec.com Ivan Rossi wrote: > > I want to insert a PCMCIA in an Altera FLEX10K. Anybody has had a positive > experience about that and could show me any valid partners that could > supply me the core in vhdl description? Thank you. > Ivan Rossi, Mantova Italy > rossivan@tecna.itArticle: 6369
Christos Dimitrakakis <mbge4cd1@afs.mcc.ac.uk> wrote: >Is there a cheap way to develop for FPGAs, suitable for the enthusiast? > >-- Christos Cheapest I know of is Warp 2 from Cypress. You should be able to pick up a copy for £69 over here. See http://www.cypress.com/cypress/warp2/page2.htm for details. Good luck Julian -- --------------------------------------------------------------------- Julian Cox CoxJA@augustsl.demon.co.uk error: smartass.sig not found Hardware development eng. August Systems Ltd ---------------------------------------------------------------------Article: 6370
Svenn-Ivar wrote: > > I've been thinking of using an FPGA as a trigger circuit in a (not so > high-end) digital scope. Functions I need are high speed comparison of > two 10-bit digital values against each other and against predetermined > values. Output would be an indication if to trig or not. > I'm not familiar to FPGA's or other 'programmable logic' so I would like > to know if this is possible to do, before I go any further. With high > speed I mean about 10ns from value input to indication output. > Suggestions are highly appreciated! > For this relatively simple fixed function with the desired 10ns propagation delay, you'd be better off using a fast CPLD rather than an FPGA. -Ray Andraka, P.E. Chairman, the Andraka Consulting Group 401/884-7930 Fax 401/884-7950 email randraka@ids.net http://www.ids.net/~randrakaArticle: 6371
Okay, the site is back up. Here are the links to MAKESRC. Go to 'http://www.xilinx.com/support/techsup/ftp/htm_index/utils_prom.htm' and download either MAKESRC.ZIP for the PC or MAKESRC.TAR.Z for UNIX-based machines. -- Steven Knapp OptiMagic(tm) Logic Design Solutions E-mail: optmagic@ix.netcom.com Programmable Logic Jump Station: http://www.netcom.com/~optmagic Steven K. Knapp <optmagic@ix.netcom.com> wrote in article <01bc63d0$d7b2d4e0$a181b6c7@default>... | There may be an easier way than understanding the .bit file format. For | example, you can create something called a "raw bits" format using the | command line "makebits -r <filename>". This creates a 1's and 0's file | formatted like it appears in the device. You can fairly easily see the | pre-amble, length count, dummy bits, and frames. | | There is also a fairly useful utility on the Xilinx web site called | 'makesrc' (make source). The makesrc program coverts an MCS file into a | file with the format the way you want it. Its somewhere in their support | section. I would provide the URL, but for some reason, my ISP provider is | having problems reaching the Xilinx site at the moment. | -- | Steven Knapp | OptiMagic(tm) Logic Design Solutions | E-mail: optmagic@ix.netcom.com | Programmable Logic Jump Station: http://www.netcom.com/~optmagic | | Archer Lawrence <archer@tanisys.com> wrote in article | <337C7534.BBA805F@tanisys.com>... | | I am looking for the file format for the '.bit' file produced from the | | Xilinx makebit program. The documentation describes the configuration | | stream format, but not this file. Thanks in advance | | | | Archer Lawrence | | Sr. Engineer | | Tanisys Technologies | | | | |Article: 6372
Svenn-Ivar <sis@hekta.iet.hist.no> wrote: > I've been thinking of using an FPGA as a trigger circuit in a (not so > high-end) digital scope. Functions I need are high speed comparison of > two 10-bit digital values against each other and against predetermined > values. Output would be an indication if to trig or not. The Lucent ORCA chip (2Cxxa-4) can implement a 12-bit equaility comparator in 6.5 ns using the built-in carry chain (3.3 in-cout + 1.1 cin-cout + 2.1 cin-O4). This leaves you 3.5ns for routing data into and out of the circuit. It may be possible to meet 10ns, but I wouldn't bet the company on it. As an alternative, you can implement a 10 or 11-bit comparator in a single PFU by re-programming the lookup table for each trigger. This would give a trigger time of less than 3ns (PFU setup time to Clock) even using the slower -3 speed parts. -------------------==== Posted via Deja News ====----------------------- http://www.dejanews.com/ Search, Read, Post to UsenetArticle: 6373
If you want to develop for a reconfigurable SYSTEM you might want to look at http://www.vcc.com/products/pci6200.html for $995 you get all the software to program the XC6216 (16,000 gates) 2Meg of fast SRAM and lots of other features and software. All this on a PCI card for a "state-of-the-art" learning experiance. Steve Casselman, President Virtual Computer CorporationArticle: 6374
Xilinx isn't dropping the concept of XBLOX. XBLOX is being replaced by a better tool called LogiBLOX in the M1 software. It works the same way as XBLOX except you have to define the bus width. One of the major advantages is faster runtime with LogiBLOX. The other VHDL and Verilog simulation files if you instantiated LogiBLOX components in your HDL. Kate Xilinx FAE
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