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
I have not seen anyone mention what is IMHO one major reason for avoiding the use of "reverse engineering protection" schemes. Perhaps other respondents live in a more perfect world than mine, where every device works as intended and process variations are unknown. I suspect the first person forced to deal with such a scheme will be some poor sucker in the product engineering department of the designer's own company. Or perhaps some quality engineer who receives the first returned material from a customer whose assembly line is shut down until some nebulous glitch is found and eradicated. To paraphrase/mangle Shakespeare "An engineer hoist on his own petard makes a lovely sight." Instead of concentrating on baffling the competition you might consider some additional effort put into design for failure analysis. Make the device easier to inspect and understand and you will undoubtedly receive much more benefit from than the competition. If you have the equivalent of a 741 op amp, which the entire world will rush to copy, then perhaps reverse engineering protection makes sense. But if you do use some such scheme you had better know how to defeat it yourself or be long gone by the time the failures roll in. -- Regards, R.E. York My employer's opinions tend to look like MILSPEC documents. My own opinions are easy to distinguish and seldom congruent.Article: 2626
Is this a question like "Where does a 900-pound gorilla sit?" ;-) >>>>> "Andrew" == Teknomage <3app@qlink.queensu.ca> writes: > I'm not sure if I know what an FPGA does and it's applications. FPGAs (Field-Programmable Gate Arrays) perform a variety of logic functions, like a gate array, but they're user-programmable (many are reprogrammable in the system). FPGAs generally contain matrices of combinatorial logic blocks, registers, and I/O blocks, which can be selectively interconnected by arrays of short and long lines which overlay these blocks. Because of this fine-grain architecture, FPGAs are very useful for integrating all sorts of oddball functions. The functions performed by an FPGA are described by some combination of schematics, state machines, boolean equations, high-level languages, lookup tables, etc.; but the designs entered in these formats eventually all get processed by vendor-supplied placement and routing tools, which determine the states of all the (proprietary) configuration bits needed to make the FPGA do what you want. So the answer is "Whatever you want it to do!" -- Roger Williams PGP key available from PGP public keyservers Coelacanth Engineering consulting & turnkey product development Middleborough, MA wireless * DSP-based instrumentation * ATE tel +1 508 947-8049 * fax +1 508 947-9118 * http://www.coelacanth.com/Article: 2627
---snip--- ¿ ¿ROMS, on the other hand, even implant ROMS, can be read with a micro-probe. not when the addressing scheme is frequency dependent. the pattern at any address may be discerned but its function in an application cannot. i guess it all depends on what you mean by reverse engineering, i.e., the chip or the usage. wray kephartArticle: 2628
Chuang Hsien-Ho (eea80593@maddux.EE.NCTU.edu.tw) ´£¨ì: : We have a project, "A Baseband Chip Set for Digital Cellular Phone", in : progress. We will soon finish the verilog coding. Now we want to prepare : for the emulation step. We have about tens of thousands of gates, targeting : on at least 6MHz. : Being a academic project, we might not able to afford a complete commercial : system. We plan to buy some RAM-based FPGAs(like XC4000 or ALTERA FLEX series), : maybe some FPICs(like Aptix). : Could you experienced persons give me some comments or suggestions? Or where : can I get the detail information of these vendors? Thanks a lot. Many thanks to those who gave me suggestion. It sounds that ALTERA will be better in rounting, but Xilinx will be easy for first user. I am really interested in the average utilization of XC4000 and FLEX8000(FLEX10K). Does anyone has experience? However, no one told me about the FPICs. The only vendor I know is Aptix, but I don't know if they have agent in Taiwan. Could anyone tell me? I also heard that the FPIC is expensive. If we don't use FPIC, is there any alternatives for the emulation board? -- =============================== Hsien-Ho Chuang eea80593@yankees.ee.nctu.edu.tw ===============================Article: 2629
hbcsc388@csun.edu (eric schonning) wrote: >Rob-L (rob-l@superlink.net) wrote: > >> When someone buys something, they should expect proper operation for some >> time, but not forever. If it can be repaired, great, but it can provide >> value even if it fails later on. That's not cheating the purchaser. > >> With electronic components/assemblies, the manufacturer has configured >> some materials for you, in order to perform some function you desire. >> You pay them for the product, and you get that function for at least the >> warranty period or some reasonable time for the type of device. > >> So you make a protected chip, and make it to last some number of hours >> minimum. If it fails before then, it gets replaced free. If it lasts >> beyond its expected lifetime, that's free use of a product, which is a >> benefit to the purchaser. If a product is not used as intended and it >> fails, or if it's tampered with and self-destructs, the purchaser eats >> the loss. That's the way it's always been. > >Please warn all the members of these newgroups of any possible commerical >or any other products that may have your design in them so we don't by the >unrepairable crap you make. I only hope that all who think like you are >miserably put out of business. Hey Eric, Rob's talking about a warranty. Have you ever heard of a warranty? You know, that thing that says if it breaks within a certain amount of time, it's repaired or fixed free. What do you expect after that? Oh, I know! Warrenties should last forever! They should defy physics. Whether it's repairable or not after is nothing more than ecomnomics. If it's cheaper to buy a new one than to repair it, then buy it. That's why when your Walkman breaks you buy another one, not pay $100 in labor plus $20 in parts to fix it. That's why all IC's are "unrepairable crap" as you so eloquently put it. It's not worth it to open the device up, find out what the problem is, and then through some very expensive equipment, effect the repairs IF they can done. But, hey, YOU go right ahead and design a completely repairable chip (and I do mean completely - not just bit fixing). Yes, a designer could have a chip shut down after a certain number of cycles but let's be reasonable. New ways of reverse engineering are always being developed. I don't believe anything can always be kept from prying eyes (codes can be broken). Beyond that, people are always changing loyalties, and let's face it, people talk. If word ever got out that a corp. was purposely killing it's chips after a certain amount of time, that company would be DEAD. Who would want to do business with them? Customers all the way up the production chain would want to know if parts from *that* company were in the design. No company's going to risk it. If they do, they're probably small, definitely stupid, and far and few between. Do some thinking and homework before you again advertise to the group how you have no understanding of what you're talking about and send an inane flame, jumping to wild accusatory conclusions. Mike S. "Opinions are mine"Article: 2630
In article <4debg1$84k@newsgate.sps.mot.com>, Mike Saltmarsh <r27863@email.sps.mot.com> wrote: > If word ever got out that a corp. was purposely > killing it's chips after a certain amount of time, that company would be > DEAD. Who would want to do business with them? Customers all the way up the > production chain would want to know if parts from *that* company were in the > design. No company's going to risk it. If they do, they're probably small, > definitely stupid, and far and few between. I don't know specifically about their _chips_, but automobile manufacturers are famous for designing parts that _have_ to be replaced in order to give their dealers something to do when new cars aren't selling. Back before the EPA stuff that essentially forced a relatively long-term warranty, the auto companies were loath to put in transistorized breakerless ignitions because 'tuneups' were a major source of revenue for the dealers. After recently paying $500-800 each for a the 'brains' (chip) of an air conditioning system and for the little chip that tells when the catalytic converter is acting up (after warranty, sigh...), I fear that the auto companies may be back to their old tricks. --- BTW, the newest engine management chips on some cars are starting to store all sorts of data on max revs, time between services, etc., so that if you overrev the engine & break something, the auto company won't pay up. Wonder how long it will take the DA's & lawyers to start subpoenaing these chips after an accident... -- www/ftp directory: ftp://ftp.netcom.com/pub/hb/hbaker/home.html Copyright (c) 1996 by Henry G. Baker. All rights reserved. ** Warning: Due to its censorship, CompuServe and its subscribers ** ** are expressly prohibited from storing or copying this document ** ** in any form. **Article: 2631
I need a little help from you FPGA users out there to settle some important questions. And I'll compensate you for the effort. Would you please answer a very short questionnaire for me?? It's on the web at "http://www.best.com/~sbaker/work/quest.htm". Only 10 short questions. I’m inquiring about what engineers who have used FPGAs think about these devices in competition with gate arrays, DSPs, MCUs, MPUs, co-processors, etc. To repay for your time spent, I’ll share the results with you when I get it sorted out. That emailed data report will be useful to you, to see what other engineers think about this issue. The data will come only from users of programmable logic devices, and not from vendors. If you work for a vendor of programmable devices or other ICs, please disregard this notice. I promise, the time it takes should be very short. There's no paperwork to fill out and no stamp required. It’s fun. The questionnaire will be available for the next couple of weeks. Sincerely, Stan Baker Contributing editor, EE Times. www.best.com/~sbaker/ sbaker@best.com p.s. if you can't access the web, send a reply by email and I'll forward an email version of the questionnaire to you.Article: 2632
READING WOKSHOPS ON PARALLEL COMPUTING The Parallel, Emergent, & Distributed Architectures Laboratory (PEDAL) in conjunction with the Dept of Computer Science at Reading University announces a series of workshops on the broad theme of parallel computing. The intention of these workshops is to provide a forum for both academia and industry to meet and discuss key issues in the area of Parallel & Disributed processing and the related field of emergent computing structures which often relies on implicit recognition or direct use of parallelism. The first workshop is planned for 28-29 Mar 96 in Reading (see call for papers below). All papers will be reviewed and will appear in the proceedings to be distributed at the workshop. In addition selected papers on particular themes will be considered for publication in a book series on advances in parallel computing. Finally there will be an award for the best paper (in the series) by a postgraduate student or Research Assistant registred for a Higher Degree. For more information or if you are interested in sponsoring an event please contact :- Prof G.M. Megson (G.M.Megson@reading.ac.uk) or Dr Shirley Williams (Shirley.Williams@reading.ac.uk) ****************************************************************************** Best Paper Award Rules : (1) The authors of the paper must registered for a Higher Degree (e.g. PhD) at a recognized academic institution (either full-time or part-time). where a paper is has multiple authors all authors should be registered (2) A finalist will be selected from each of the Workshops in the series and the winner selected from this shortlist. (3) The selection will be made on quality, timeliness, and perceived impact on the field. Review or survey papers will not be considered. (4) The decision will be made by an independent panel appointed by the workshop organisers and their decision will be final. (5) The award will consist of a certificate citing the authors and the paper title (for each finalist) and a contribution towards attending an overseas conference to present a paper for the overall winner. (6) Any papers published in (5) must give due acknowledgement to the support of the Workshop programme. ******************************************************************************* 1st READING WORKSHOP ON PARALLEL, EMERGENT, AND DISTRIBUTED COMPUTING PERFORMANCE MONITORING AND EVALUATION OF HIGH PERFORMANCE SYSTEMS Announcement and first call for papers March 28-29 96, Reading University The problem of monitoring and evaluating is essential for our understanding of the ways in which components of high performance systems interact and hence how we can exploit them effectively. Papers are invited on all aspects of performance monitoring and evaluation including, but not limited to * Benchmarking * Software support for effective measurement * problems in measurement * dynamic systems * real-time systems * process control * emergent (monitoring techniques) * reliability (performance degradation) * scalability DEADLINES Extended Abstract (3 A4 pages max) 25 January 1996 Authors notified by 15 Feburary 1996 Full paper required (10 pages, 10pt, A4 Max) 15 March 1996 SUBMISSION Extended abstract should consist of a summary including a short description of the problem and its significant with references to previous work. A submission should consist of the extended abstract together with title, authors names and affiliations, addresses contact numbers and email addresses. Abstracts can be submitted in hardcopy form or as postscript files to the address below. Further details of style for the final papers will be issued at the time of acceptance. Submission should be made to the Prof G.M. Megson, PEDAL, Dept Computer Science, The University of Reading, PO Box 225, Whiteknights, Reading, Berkshire, RG6 6AY. Email G.M. Megson@reading.ac.uk, FAX +44 1734 751994, TEL +44 1734 318600. URLS: The Parallel, Emergent and Distributed Systems Laboratory is at http://www.cs.reading.ac.uk/cs/research/pedal/index.html General Information and Calls for Papers for the Reading Workshops on Parallel Computing are based at: http://www.cs.reading.ac.uk/cs/research/pedal/rwpc/rwpc.html ***********************************************************************Article: 2633
Henry Baker (hbaker@netcom.com) wrote: : After recently paying $500-800 each for a the 'brains' (chip) of an air : conditioning system and for the little chip that tells when the catalytic : converter is acting up (after warranty, sigh...), I fear that the auto companies : may be back to their old tricks. : --- : BTW, the newest engine management chips on some cars are starting to store : all sorts of data on max revs, time between services, etc., so that if you : overrev the engine & break something, the auto company won't pay up. : Wonder how long it will take the DA's & lawyers to start subpoenaing these chips : after an accident... Does anyone have any more details on this? How, for example, could the ECM even be aware when you get your oil changed. Unless the service procedure calls for hooking up a diagnostic connector and informing the computer that the oil was changed. Also, I wonder how well they are able to determine the true mileage on a vehicle from data in the computer. Most computers have some sort of vehicle speed sensor. The computer could integrate this speed and keep a pretty good estimate of the mileage. And if someone disconnects the speedometer, the input to the v.s.s. might also be disconnected. This would certainly cause some type of malfunction to be logged in the computer. I can see that the dealerships themselves have a vested self-interest in knowing the true mileage, since the temptation to disconnect is quite high nowadays. Both to extend the length of the warranty and to avoid stiff mileage penalties on leased vehicles. I'm sure that if there are means to determine mileage from computer data, that they are kept secret by the mfr. For example, such data would not benefit someone buying a used car or a third-party leasing company. It would only be used in cases where the dealer itself has an interest.Article: 2634
I have a study which was done by a researcher at University of Toronto which does a full comparison of routing in the Xilinx 4000 versus the Altera 8000 (Flex) family. The summary you may find interesting, and if you would like a faxed copy of it, please send your fax number to me (my email is jason.feinsmith@xilinx.com). The conclusion implies that Xilinx 4000 routes better than does the Altera Flex 8000. I'll quote 2 summary items from the paper: "3.1 Results for the Xilinx XC4000 FPGAs... It is interesting to note that for all circuits used, none of them becomes un-routable even under the worst pin constraints. This was true even for the circuits that were very tightly packed, in terms of percentage of available CLBs and I/O pins used." and "3.2 Results for the Altera FLEX 8000 FPGAs... The Altera FLEX 8000 FPGAs seem to be slightly susceptible to routing failures uner random pin constraints in cases where the I/O pin or logic element utilization is close to 100%. ... It seems that system designers, when implementig a circuit using FLEX 8000 FPGAs, should leave about 20% of the logic eements and I/O pins free to avoid routability problems due to pin constraints."Article: 2635
In article <hbaker-1501961604490001@10.0.2.15>, hbaker@netcom.com (Henry Baker) wrote: ---snip snip--- ¿BTW, the newest engine management chips on some cars are starting to store ¿all sorts of data on max revs, time between services, etc. ---more snips--- which cars? wray kephartArticle: 2636
In article <4dheje$16b@cloner2.ix.netcom.com>, kephart@ix.netcom.com (kephart) wrote: > In article <hbaker-1501961604490001@10.0.2.15>, > hbaker@netcom.com (Henry Baker) wrote: > ---snip snip--- > ¿BTW, the newest engine management chips on some cars are starting to store > ¿all sorts of data on max revs, time between services, etc. > ---more snips--- > > which cars? Certain foreign cars, for sure, according to a factory-trained mechanic of one of those car companies. He also mentioned that the engine diagnosis systems were becoming more standardized, so that you wouldn't have to buy a different computer&software for every different car you worked on. Since the chips on the American cars are probably more sophisticated, they may be doing more than their European counterparts. -- www/ftp directory: ftp://ftp.netcom.com/pub/hb/hbaker/home.html Copyright (c) 1996 by Henry G. Baker. All rights reserved. ** Warning: Due to its censorship, CompuServe and its subscribers ** ** are expressly prohibited from storing or copying this document ** ** in any form. **Article: 2637
>>>>> "JF" == jasonf <jasonf> writes: In article <4dgsqu$q59@mailman.xilinx> <jasonf> writes: JF> I have a study which was done by a researcher at University of JF> Toronto which does a full comparison of routing in the Xilinx JF> 4000 versus the Altera 8000 (Flex) family. The summary you JF> may find interesting, and if you would like a faxed copy of JF> it, please send your fax number to me (my email is JF> jason.feinsmith@xilinx.com). The conclusion implies that JF> Xilinx 4000 routes better than does the Altera Flex 8000. <blah blah blah blah ...much Xilinx shilling deleted> Sigh. I suppose it had to happen sooner or later. Unfortunately, this was posted my the Xilinx University Program Coordinator (who ultimately provides me with my software) but I will flame anyway. FLAME MODE ON Hey Jason, take this to comp.advocacy.fpga (or anywhere else)! The original reason that this newsgroup was setup was to study FPGAs in the context of configurable computing. We all know that there have been a whopping 10 messages regarding configurable computing since the inception of this newsgroup. Although relatively content-free, this newsgroup has at least been civil. Now, with you (JASON FEINSMITH) inciting a flame war over which FPGA is better, what little information that was being imparted will be replaced with shrill marketing. Thanks a bunch. I was getting ready to unsubscribe. Crap like this will push me over the edge. Sheesh and double sheesh. FLAME MODE OFF Brad Hutchings Reconfigurable Logic Lab Brigham Young University -- Brad L. Hutchings - (801) 378-2667 - hutch@ee.byu.edu Brigham Young University - Electrical Eng. Dept. - 459 CB - Provo, UT 84602 Reconfigurable Logic LaboratoryArticle: 2638
>>>>> "foo" == <jasonf@no.home.but.xilinx> writes: > The conclusion implies that Xilinx 4000 routes better than does > the Altera Flex 8000. Despite your return address ;-) I have to agree with you on this. We tested both of them for our DSP work, and ended up sending back Max+Plus in disgust. (Sorry, Altera.) > ... It seems that system > designers, when implementig a circuit using FLEX 8000 FPGAs, > should leave about 20% ... And the Max+Plus software explicitly reminds you of this. We found that 70% was a better goal. This, coupled with the *bizarre*, proprietary Max+Plus software, and Flex 8000s' high prices and poor availability made our decision easy. -- Roger Williams PGP key available from PGP public keyservers Coelacanth Engineering consulting & turnkey product development Middleborough, MA wireless * DSP-based instrumentation * ATE tel +1 508 947-8049 * fax +1 508 947-9118 * http://www.coelacanth.com/Article: 2639
Hi, I have the vendor specific design tools for both Actel/TI & Xilinx FPGA's. Both of these include a copy of PRO Capture. Does anyone know if I need to install both copies of PRO series capture, or can I just use one copy with two libraries? TIA - MarkArticle: 2640
Mark- I'm no expert on this, but I believe a single install will work....BUT it is the license files that will mess you up. Each of the FPGA libraries is probably secured to a particular license file. Save a copy of your original license files away someplace safe. Trying combining the two into a single license.wv file and see what happens? Otherwise it will probably work if you swap in the appropriate vendor's license file for that session. Let me know what works, either by this group or via the below address. I'm very interested. Doug Shade rxjf20@email.sps.mot.comArticle: 2641
SUBJECT:***KEYBOARD RECORDERS******** ALSO KNOWN AS:Keyboard Grabber, Keyboard Key Logger, Keyboard Monitor, Keyboard Recorder. PURPOSE: Captures keystrokes and sends & saves them to a hidden file. Now you can keep a record of any keyboard activity on your computer. Monitor your computer at home or office. My private collection of keyboard recorders is yours for only $9.95. You will receive 18 different programs on a 3 1/2 disk. You'll get:KEYCOPY,KEYFAKE,KEYREAD,KEYTRAP,KEYREC,KEYLOGWN(Windows), HACKKEY,BAGKEYS,GETIT,PLAYBACK,ROBOKEY,RECORD,ENCORE, KCAP10,PTM229N,QWERTMAN,GKG,DEPL. Just send $9.95 plus $1.00 for shipping and handling to: HLD PLUBLISHING COMPANY 1680 N. VINE ST. #1103 LOS ANGELES, CA. 90028Article: 2642
I believe the chrysler speedometer output comes from the computer. If you have the digital display, it comes directly. If you have the analog one, it comes from a D/A convertor and Analog meter. The digital odometer needs to be in the computer, the analog one I don't know. -- glenArticle: 2643
hbaker@netcom.com (Henry Baker) writes: >Certain foreign cars, for sure, according to a factory-trained mechanic of one >of those car companies. He also mentioned that the engine diagnosis systems >were becoming more standardized, so that you wouldn't have to buy a different >computer&software for every different car you worked on. >Since the chips on the American cars are probably more sophisticated, they may >be doing more than their European counterparts. Do I detect a certain snobbishness here? Bernie -- ============================================================================== I'm using an experimental version of "nn" to read news. So if anything seems strange about the headers of this mail or post, please inform me. Thanks!Article: 2644
In article <slworkDLA997.LAF@netcom.com>, slwork@netcom.com (Steve Work) wrote: ¿Henry Baker (hbaker@netcom.com) wrote: ---snip henry's bitching--- ¿: BTW, the newest engine management chips on some cars are starting to store ¿: all sorts of data on max revs, time between services, etc. ---more snip--- ¿Does anyone have any more details on this? How, for example, could the ¿ECM even be aware when you get your oil changed. ¿Also, I wonder how well they are able to determine the true mileage on a ¿vehicle from data in the computer. Most computers have some sort of ¿vehicle speed sensor. The computer could integrate this speed and keep a ¿pretty good estimate of the mileage. not necessary. use the (un-named mfg) multi-function odometer chip. ¿I can see that the dealerships themselves have a vested self-interest in ¿knowing the true mileage, since the temptation to disconnect is quite ¿high nowadays. Both to extend the length of the warranty and to avoid ¿stiff mileage penalties on leased vehicles. if you disconnect it the engine won't run (starting about '97'). ¿I'm sure that if there ¿are means to determine mileage from computer data, that they are kept ¿secret by the mfr. For example, such data would not benefit someone ¿buying a used car or a third-party leasing company. It would only be ¿used in cases where the dealer itself has an interest. can't be too secret, i have a bunch on my lab bench.Article: 2645
I'm working on a plan to do a minimal '486 PC chipset in a fpga, and for that I need to design a 8259 compatible interrupt controler. Does anybody have the logic/schematics/source for that, or know somebody who have, preferable for actel fpga, but I could translate from any schematics/description language. Best regards SorenArticle: 2646
In article <821893282.27348@rmcecl.demon.co.uk> MarkW@rmcecl.demon.co.uk "Mark Webster" writes: " "I have the vendor specific design tools for both Actel/TI & Xilinx "FPGA's. Both of these include a copy of PRO Capture. " "Does anyone know if I need to install both copies of PRO series "capture, or can I just use one copy with two libraries? Mark, Officially, you need to install both copies. However, it is possible to avoid a double installation. The workaround is to swap the security device and the licence.wv file from the Actel copy with those from the Xilinx copy when you want to load an Actel drwg, and vice versa. Merging the licence files won't work. Alternatively, you can purchase an unrestricted copy of PROCapture or ViewDRAW for UKP1200, which you can use for all FPGA and PLD vendors' devices, and also for your top-level schematics. -- David Pashley < ------------------------ < < < ---------- Email: david@fpga.demon.co.uk | Direct Insight Ltd < < < < > Tel: +44 1280 700262 | | *The EDA Source* < < < Fax: +44 1280 700577 | --------------------------- < ------------------------------------------Article: 2647
Does anyone know of a way to program an Actel FPGA in the circuit? I am aware that even if it is possible, it's not straight forward. Chris HersmanArticle: 2648
hi, I am looking for suggestions on good interview questions for a digital designer. Interesting verilog coding issues, synthesis questions, issues which shows the understanding of basic design fundemantals (meta-stability etc) are welcome. thanks muzo standard disclaimerArticle: 2649
Does anyone know if Steven Casselman's Virtual Computer Corporation is still alive? Thanks in advance, -Arrigo Benedetti
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