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
On Tue, 3 Oct 2000 02:17:24, bob_42690@my-deja.com wrote: > > > > > > P.S. I'm not threatened by being called an "amateur". I > > understood the comment. I am a relative newbie though (25+ years > > as a hardware designer, 1+ years doing this *neat* stuff). Just > > to be clear, I am also rather new to synthesis (good grief, I've > > been dreaming *in* VHDL recently), so treat me kindly. ;-) > > I had a dream that I was debugging a VHDL simulation! Now that's not > right. Must be some type of recursion thing. Well my dreams were a little more bizare than recursion (though as an engineer, recursion is bizare enough). This was more on the lines of: zz_zzz:process (clock, reset, dream) variable dreams: actresses; begin If reset = '1' -- If asynchronous reset then eyes <= open; -- then *Damn*! elsif falling_edge(clock) -- If clock broken or not morning yet then -- then Snore <= true; -- ;-) case dreams is -- Select a dream when Pammela => -- When interesting ... -- err, nevermind! ..well you get the idea. I was dreaming *IN* VHDL. Ack! ..and I'm not good at VHDL, either! ;-) ---- KeithArticle: 26076
"Keith R. Williams" wrote: > ..Amplify is another issue altogether. I'd like some information > from anyone with experience with it. I had written a longer response to this, and was cut off by Dr. Watson. Joy. I have used Amplify. I would recommend it if you writing HDL that pushes the silicon for speed. Amplify will allow for clock rate improvement by improvement of mapping and by placement information. I think it is a good tool. Amplify fits between a "push the button" type flow and a "I'll put every F-map in the exact right spot" type flow. It will get better results (meaning clock rate) than the first. It will NOT get better results than the second, however it takes rather less design effort. It is a spendy tool, however schedules and buying faster parts mean money as well. Questions, ask. -- Phil HaysArticle: 26077
Johan Petersson wrote: > > Hi Christophe, > > What simulator are you using? I have not used coregen RAM, but my wild > guess > would be that you have a bug in your simulator :) > > Good luck, > Johan P :) > > Christophe Heyert wrote: > > > > A very strange thing happened to me yesterday. I'm using a RAM block > > generated by the Coregenerator in > > my structural design. > > The first time I didn't have any problems i my behavioral simulation, > > but the second time the following error appeared. > > > > Internal ERROR : Access violation > > Occurred in architecture BEHAVIORAL of entity > > C_MEM_BLOCK_V1_0 (which is a single port RAM block). > > > > Where does this problem come from? > > > > christophe We're using Fusion from Viewlogic, and indeed, sometimes it behaves very strange. I simply copied the entire design in a new project directory and the problem was solved. christopheArticle: 26078
Qian Zhang wrote: > > Hi Friends > > Here I have Xilinx demo board, > but I dont know how to use it. > Where can I find some information? > > Also Who can tell me how to generate bit file > so that I can download to demo board? > Thank you very much! > > Qian Look on the Xilinx website. In the application notes you might find some info about demo boards. If you are using Alliance from Xilinx you can either generate the bit file by setting the configuration option ON in the options. There is also an icon in the GUI that will generate your bit file. greetz, christopheArticle: 26079
Rick Filipkiewicz a écrit : > XST, Xilinx's own Verilog & VHDL synth tool, is now just about useable > since the 3.1i release has a decent-ish manual. It also runs nicely > from the command line. I'm going to try & benchmark it against > Synplify fairly soon. Hi Tell me about your benchmarking so that I can do it against Leonardo -- Nicolas MATRINGE IPricot European Headquarters Conception electronique 16 rue du Moulin des Bruyeres Tel +33 1 46 67 51 11 F-92400 COURBEVOIE - FRANCE Fax +33 1 46 67 51 01 http://www.IPricot.com/Article: 26080
Hi Steve, I am quite sure the programming sequence is correct as I have tried with different output frequencies. The problem arises when I have a divider (even a divide by 2 circuit) in FPGA. Below is the divide-by two circuit that I use in VHDL:- library IEEE; use IEEE.std_logic_1164.all; entity clk_div2 is port(RESETN : in STD_LOGIC; CLK : in STD_LOGIC; ---------------------------------------------------------------------------- --- HCLK : out STD_LOGIC ); end clk_div2; architecture RTL clk_div2 is signal hclk_sig : STD_LOGIC; begin hclk <= hclk_sig; process(RESETN, clk) begin if RESETN = '0' then hclk_sig <= '0'; elsif (clk'event AND clk = '1') then hclk_sig <= NOT hclk_sig; end if; end process; end RTL; Any problem on the above?? Sherdyn Steve Wiseman <steve@steves-house.org.uk> wrote in message news:39D8D348.D6EF4BC1@steves-house.org.uk... > Sherdyn wrote: > > > > Hi, > > > > I have encountered a strange problem with Cypress PLL ICD2061A. > > I've used these successfully several times - they're great, but not > always obvious. Have you used the bitstream generation proglet, or > hand-cranked the calculations? > Things I've got wrong in the past:- > * Bitstream backwards, (the datasheet used to be rather ambigous) > * non-monotonic serial clock, <- this was the real problem, use a > fast, fast scope , or a few pF just to take the edge off anyway. > * looking at the wrong output pin. (D'oh!) > > I'd pay good money for one of these devices, in a regular 14-pin > oscmod case, with a watch LCD and up-down buttons for frequency, > sometimes, just to see how close to the edge a design is running. > Only the non-availability of watch LCDs has stopped me doing this > myself...) > > > So what could FPGA causes the PLL not to work well? Jitter or something > > else. Any clue? > > The don't seem too vulnerable to jitter, but I generally never hung > them off anything other than other oscmods, or used their internal > crystal drivers. > > Good luck - > > SteveArticle: 26081
Why not post or make available your files so those with other tools can benchmark it also? Rick Filipkiewicz wrote: > rickman wrote: > > > I also noticed that there are big differences other than just the chips targeted. The existing > > WebPack does not include FPGA Express synthesis, (not sure about XST, whatever that is) but the > > press release indicates that it now includes a full HDL capability. > > -- > > > > XST, Xilinx's own Verilog & VHDL synth tool, is now just about useable since the 3.1i release has a > decent-ish manual. It also runs nicely from the command line. I'm going to try & benchmark it against > Synplify fairly soon. > > Now, Xilinx, how about actually getting some SpartanII's out the door so we can use the new free s/w.Article: 26082
Hi, I'm thinking about implementing a JVM ( Java Virtual Machine ) processor in a FPGA. But I don't want to invent the wheel once more. Does enyone know if this is done allready? Thanks Martin ------------------------------------------------------ Martin Schoeberl, mailto:martin@good-ear.com Need some ear training? Take a look at -----> http://www.good-ear.com/ <-----Article: 26083
Shahzad, You identified flexibility as a key advantage of the FPGA solution. FPGAs are also competitive in terms of cost and speed. See http://www.xilinx.com/products/xaw/reed/ for some details on a Reed-Solomon solution. Similar results are attainable for Viterbi and interleavers. Another advantage with the FPGA solution is that if you use cores they are usually quickly migrated to new FPGA architectures as they appear. You can therefore take advantage of higher clock speeds, new I/O voltage standards, etc. Regards, Bill.Article: 26084
"Martin Schoeberl" <martin.schoeberl@chello.at> wrote in message news:XLkC5.64438$zJ5.3129190@news.chello.at... > Hi, > > I'm thinking about implementing a JVM ( Java Virtual Machine ) processor in > a FPGA. > But I don't want to invent the wheel once more. > Does enyone know if this is done allready? Here are a couple of Java processor core implementations that I have read have been run in an FPGA: Austin Kim (Lucent), Morris Chang (IIT) paper: Designing a Java Microprocessor Core Using FPGA Technology http://csl.cs.iit.edu/~java/publication/asic98.html Advancel Logic Corp. TinyJ http://www.advancel.com/home.htm Jedi Technologies JStar http://www.jeditech.com/jstarwp.html Derivation Systems LavaCore http://www.derivation.com/news/pressrelease_06-21-1999.html Vulcan Moon http://www.vulcanasic.com/eda.htm I suspect at least one of the Sun picoJava cores has been synthesized to an FPGA. Certainly it had been discussed. See also www.fpgacpu.org/usenet/java.html. And someday we'll have an FPGA CPU+SoC implementing (or hosting) a JVM, hosting an EDA application, with jbits, that emits version n+1 of itself. Jan Gray Gray Research LLC www.fpgacpu.orgArticle: 26085
> See also www.fpgacpu.org/usenet/java.html. Sorry, make that www.fpgacpu.org/usenet/javaproc.html. Jan Gray Gray Research LLCArticle: 26086
rickman wrote: > > This seems very odd that both FPGA Express and Synplify have the same bug that I assume worked > correctly at one time. It would be hard to imagine that this problem has been there from the start > and no one noticed it. So how could two vendors develop the same bug at the same time? Rick, I've narrowed this down, trying a whole bunch of test cases. It appears that Synplify v6.0 has a problem with the falling-edge clock for the RAM. If the RAM's clock is rising edge, the falling edge stuff gets implemented correctly. If the RAM's clock is falling-edge triggered, the tool is confused. Synplicity's tech support knows about this problem, and apparently it's been fixed in the beta of v6.1. As for FPGA Express, I *know* that versions 3.3 and earlier had no problem doing falling-edge clocks, because I did a handful of XC4KE designs that used 'em, and everything worked fine. It's v3.4 that has the problem, at least with SpartanXL, and I have a simple test case that proves it. I kept the install of v3.3, or so I thought, but it access faults every time I try to run it, so I can't test it. Kudos to Xilinx tech support: I got two e-mails from two different support engineers last night -- one was at 7:09 pm; the other at 11:06 pm (Arizona time). -- a ---------------------------- Andy Peters Sr. Electrical Engineer National Optical Astronomy Observatory 950 N Cherry Ave Tucson, AZ 85719 apeters (at) n o a o [dot] e d uArticle: 26087
rickman wrote: > Are you sure that this is not a new bug. I used both rising and falling > edge clocks in a VHDL design I did using FPGA Express and Xilinx M1.5 > about two years ago. There were no inverters in any of my clock paths. Here's the code that demonstrates the clock-inverter bug. Anyone wanna run it through FPGA Express v3.3? library ieee; use ieee.std_logic_1164.all; entity digsamp is port ( sysclk : in std_logic; -- master clock mreset_l : in std_logic; -- global reset din : in std_logic_vector(15 downto 0); -- data input den : in std_logic; -- an enable dout : out std_logic_vector (15 downto 0)); -- output data end entity digsamp; architecture foo of digsamp is -- sync input stuff with the clock: signal din_s : std_logic_vector (15 downto 0); signal den_s : std_logic; -- intermediate signals: signal din_d : std_logic_vector (15 downto 0); signal dout_fe : std_logic_vector (15 downto 0); begin syncin: process (sysclk, mreset_l) is begin if mreset_l = '0' then din_s <= (others => '0'); den_s <= '0'; elsif rising_edge(sysclk) then din_s <= din; den_s <= den; end if; end process syncin; -- latch din when enabled: latchit: process (sysclk, mreset_l) is begin if mreset_l = '0' then din_d <= (others => '0'); elsif rising_edge(sysclk) then if den_s = '1' then din_d <= din_s; end if; end if; end process latchit; -- for giggles, delay it by catching it on the falling edge of the clock: invclock: process (sysclk, mreset_l) is begin if mreset_l = '0' then dout_fe <= (others => '0'); elsif falling_edge(sysclk) then dout_fe <= din_d; end if; end process invclock; -- drive out out on the falling edge of the clock: outdrive: process (sysclk, mreset_l) is begin if mreset_l = '0' then dout <= (others => '0'); elsif rising_edge(sysclk) then dout <= dout_fe; end if; end process outdrive; end architecture foo; -- ---------------------------- Andy Peters Sr. Electrical Engineer National Optical Astronomy Observatory 950 N Cherry Ave Tucson, AZ 85719 apeters (at) n o a o [dot] e d uArticle: 26088
hey, Sourcing a GND and VCC was an issue weeks ago in this news group. Xilinx claims that the Virtex is rich in terms of routing ressources, so why not source the Pwr and Gnd wires from the IOB? I know a friend who does so in the xc4k, is it not possible in virtex ? --Erika Sent via Deja.com http://www.deja.com/ Before you buy.Article: 26089
In article <NiHB5.7086$1H2.873605@typhoon.tampabay.rr.com>, sramirez@deleet.cfl.rr.com (S. Ramirez) wrote: > I thought that Altera made "EPLDs" and CPLDs. > However, I thought that the Flex 10K/6K families are gate arrays, > thereby making them an FPGAs, technically. > Please correct me if I am wrong. Altera don't call their Flex devices FPGAs but everyone else does. Are they 'really' FPGAs or not? It's an angels-on-pinheads question. > "Netscape User" <your_email@here.net> wrote in message > news:39D6C2D8.CE6E3315@here.net... > > First of all, Altera makes CPLD's not FPGAs...minor technical > > distinction. I'm of the belief that recruiters who post technically > > accurate job listings get better technical candidates! > > > > > We are setting up a design center in new delhi, india. > > > > > > Pre-IPO company. > > > > > > Stock options in US parent company. > > > > > > www.forasic.com > > > > Have you considered listing your job postings under www.jobtrak.com? > > > > > -- Steve Rencontre http://www.rsn-tech.co.uk //#include <disclaimer.h>Article: 26090
what about power consumption ? an issue never risen why? In article <ee6e1ae.0@WebX.sUN8CHnE>, "Bill Wilkie" <bill.wilkie@xilinx.com> wrote: > Shahzad, > > You identified flexibility as a key advantage of the FPGA solution. FPGAs are also competitive in terms of cost and speed. See http://www.xilinx.com/products/xaw/reed/ for some details on > a Reed-Solomon solution. Similar results are attainable for Viterbi and interleavers. > > Another advantage with the FPGA solution is that if you use cores they are usually quickly migrated to new FPGA architectures as they appear. You can therefore take advantage of higher clock speeds, new I/O voltage standards, etc. > > Regards, > Bill. > Sent via Deja.com http://www.deja.com/ Before you buy.Article: 26091
Personally, I'm not going to spend the $$$ for it. I've already built a fairly sizable library of completely placed thingies (technical term for macros), that will run rings around what amplify will do, plus, it'll work with a customer's tool suite, and has a better than even shot of being portable across synthesis tools. As for the design effort, for first time through it will reduce the effort, but once you have the widgets made that time is invisible. If you are not reusing the pieces (one time designs), then it is fast to just place them in the floorplanner (xilinx) and be done with it. Phil Hays wrote: > > "Keith R. Williams" wrote: > > > ..Amplify is another issue altogether. I'd like some information > > from anyone with experience with it. > > I had written a longer response to this, and was cut off by Dr. Watson. Joy. > > I have used Amplify. I would recommend it if you writing HDL that pushes the > silicon for speed. Amplify will allow for clock rate improvement by improvement > of mapping and by placement information. I think it is a good tool. > > Amplify fits between a "push the button" type flow and a "I'll put every F-map > in the exact right spot" type flow. It will get better results (meaning clock > rate) than the first. It will NOT get better results than the second, however > it takes rather less design effort. > > It is a spendy tool, however schedules and buying faster parts mean money as > well. > > Questions, ask. > > -- > Phil Hays -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.com or http://www.fpga-guru.comArticle: 26092
SO you are using up a pin to get your fixed logic '1' or '0' ???? Perhaps from an unbonded pin... erika_uk@my-deja.com wrote: > > hey, > > Sourcing a GND and VCC was an issue weeks ago in this news group. > > Xilinx claims that the Virtex is rich in terms of routing ressources, > so why not source the Pwr and Gnd wires from the IOB? > > I know a friend who does so in the xc4k, is it not possible in virtex ? > > --Erika > > Sent via Deja.com http://www.deja.com/ > Before you buy. -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.com or http://www.fpga-guru.comArticle: 26093
Okay, now I'm really confused with Xilinx's Licensing. A few weeks ago, we had the distributor over, pushing the new ISE. If we pay for the time based license for 3.1i Base, we get the old GUI, support for up to Virtex '50. Or we can pay more, get the 'ISE' version (Which is explained to me as a new GUI, same tools.) Alternatively, We can get the 'Express', which is full device support, 3.1i synthesis tools, old GUI. Pay an *additional fee* for the ISE, and still get full device support, new GUI. We're using Spartan II / Virtex '150, so we can't use the Base, since it only supports up to Virtex '50. All this was last week. And now, Virtex is going to release all of this for free with the WebPack? (Up to the XCV-300, anyways.) Am I missing something, or did they not inform my distributor? Anything that can sort this mess ou is appreciated. -KentArticle: 26094
At $55, who's complaining? I only wish XSE 2.1i worked out of the box under Win2000. So far I can't seem to get the stupid flexlm to see my ethernet card. (Everything works fine under Win98SE.) Interestingly enough, the 'demo license' that comes with XSE 2.1i is identical to the commercial version of Foundation Express 2.1i (as far as I can tell.) The demo-license *appears* to allow the implementation tools to target any part up to the XCV-1000/1000E! The included FPGA-express seems to behave likewise. The demo-license doesn't support synthesis contraints, so of course you would not be able to generate an optimal design placement/routing. This is all very preliminary, of course, as I haven't actually worked through a very large design to test my theory. Not that it makes much difference to me, as my workplace has the Xilinx Foundation Express software (3.1i) It's nice to know that I ever wanted do some very large project as a hobby, the Xilinx 2.1i student-edition will allow it. Andy Peters wrote: > > Seems to me that once the "next generation" of Xilinx' tools are > released, the previous generation becomes the Student Edition. > > What an interesting way to keep selling the old stuff! > > -- a > ---------------------------- > Andy Peters > Sr. Electrical Engineer > National Optical Astronomy Observatory > 950 N Cherry Ave > Tucson, AZ 85719 > apeters (at) n o a o [dot] e d uArticle: 26095
> > That AMP connector is a board-to-board 0.8mm pitch job, so you will not > > find a cable that fits it. > > > > This sounds like a good time to vent some steam in the direction of > > whoever designed a prototypng board with a 0.8mm pitch connector on it. > > You might try looking in the Samtec catalogue. They have some 0.8 mm stuff that > might be suitable for a bodge. And you can generally get their stuff within a > few days or max 5-6 if its a custom special. Contrast AMP and or their distis: > MOQ = 10-20 GUnits, lead time after your design has qualified for museum > status. > > Company connector rules after much bitter experience: > > (1) If its not a PC world connector and its not in the Samtec book it doesn't > get used. > > (2) If a client insists on some AMP special they source it. ***SIGH*** Ok, looks like I'll work something out with the PCI-edge connector on the AVNET Xilinx board. The development kit (ADS Xilinx XCV300 Development System) seemed like a great deal, $2495 for Foundation Express 2.1i software and the FPGA board. I guess nothing is perfect. I've had good experience with an Xess XS40-4010XL board, so I may try one of their Virtex boards. At least Xess's boards have standard I/O pin headers on their boards. Unfortunately, they use the Virtex in PQ240 package, so "general-purpose I/Os" are very scarce (if you use all the nice board features like SRAM, RAMDAC, audio codec, video digitizer, etc.)Article: 26096
On Tue, 3 Oct 2000 05:59:31, Phil Hays <spampostmaster@sprynet.com> wrote: > "Keith R. Williams" wrote: > > > ..Amplify is another issue altogether. I'd like some information > > from anyone with experience with it. > > I had written a longer response to this, and was cut off by Dr. Watson. Joy. > > I have used Amplify. I would recommend it if you writing HDL that pushes the > silicon for speed. Amplify will allow for clock rate improvement by improvement > of mapping and by placement information. I think it is a good tool. Ok, this is goodness. > Amplify fits between a "push the button" type flow and a "I'll put every F-map > in the exact right spot" type flow. It will get better results (meaning clock > rate) than the first. It will NOT get better results than the second, however > it takes rather less design effort. Yes, I have some serious speed problems (200MHz in a XCV600E-7). The good news is that it's mostly dataflow. The only "arithmetic" is address decoding. Latency is a problem, but not a serious one. I can reduce stages later. As I have indicated, my problems more of a lack of experience and even a "guru", beieve it or not. I'm learning fast, but it would be a whole lot faster if I had someone who would admit that they knew VHDL, not to mention FPGA design. > It is a spendy tool, however schedules and buying faster parts mean money as > well. Exactly. The cost of sitting me down in a chair is far more than Amplify. If I get the damned thing working I'm a hero (or at least not waste of manpower), no matter what the cost inbetween. > Questions, ask. I've gone through the Amplify PowerPoint-Ware. ...well, mostly. I have many questions, which I really should talk to their FAEs about. ...though since you've volunteered! ;-) - I've looked through the flow and it seems that my Synplify license is useless, given that I wanted to stay with Virtex (I'm still fighting with a SpartanXL, so this isn't a biggie). - Given the above, and Synplicities wish to move me to -Pro, why? Sure, I see some use for -Pro for higher speed designs, but doesn't Amplify trump -Pro here. I don't see the cross. - How hard is Amplify to learn. Can I pick it up (they did send me the manual) or should I take a *valuable* week off to go to their new training? By the time I get to the left-coast I might just as well take their -Pro classes as well. The Xilinx floor-planner has caused me nothing but problems. Yes, self-inflicted I'm quite sure. There are many more questions, but I'm about to go dream in VHDL, again. Today was not good. ---- KeithArticle: 26097
Ray Andraka wrote: > Personally, I'm not going to spend the $$$ for it. I've already built a fairly > sizable library of completely placed thingies (technical term for macros), that > will run rings around what amplify will do, plus, it'll work with a customer's > tool suite, and has a better than even shot of being portable across synthesis > tools. I'm aware of your design style and am not surprised that you don't want to use Amplify. > If you are > not reusing the pieces (one time designs), then it is fast to just place them in > the floorplanner (xilinx) and be done with it. Amplify is easier and faster to use than the Xilinx floorplanner and gets better results. It is faster to use as you have fewer things to place. The placement is stable even if the source changes. Amplify does a better job of mapping to logic by using the hints from the block level floorplan. -- Phil HaysArticle: 26098
Good luck. I have not understood Xilinx licensing since day one. I guess I understand it when a chip company says they "have" to charge for the software that you *must* use in order to make use of their chips. I understand why they give away (nearly) the same software for use by "students". Maybe I can even understand a motovation for licensing software for a limited time. But I start having problems understanding all the various options and permutations of their software for different feature sets, range of chips supported and all the other nonsense involved in trying to figure out how to get a design done in the least amount of trouble. The pricing structure changes so frequently that you could likely make big bucks not from being able to use their tools, but rather just understanding the marketing well enough to offer advice on what and how to buy them! Kinda like the weather in the northeast. If you don't like it, just wait a little while... it will change! At least the new WebPack ISE is a step in the right direction. Does anyone understand why they are limiting the chips supported? How does that benefit them as a company? korthner@inf.furukawa.co.jp wrote: > > Okay, now I'm really confused with Xilinx's Licensing. > > A few weeks ago, we had the distributor over, pushing the new ISE. > > If we pay for the time based license for 3.1i Base, we get the old > GUI, support for up to Virtex '50. Or we can pay more, get the 'ISE' > version (Which is explained to me as a new GUI, same tools.) > Alternatively, We can get the 'Express', which is full device support, > 3.1i synthesis tools, old GUI. Pay an *additional fee* for the ISE, > and still get full device support, new GUI. > > We're using Spartan II / Virtex '150, so we can't use the Base, since > it only supports up to Virtex '50. > > All this was last week. > > And now, Virtex is going to release all of this for free with the > WebPack? (Up to the XCV-300, anyways.) Am I missing something, or > did they not inform my distributor? > > Anything that can sort this mess ou is appreciated. > > -Kent -- Rick "rickman" Collins rick.collins@XYarius.com Ignore the reply address. To email me use the above address with the XY removed. Arius - A Signal Processing Solutions Company Specializing in DSP and FPGA design Arius 4 King Ave Frederick, MD 21701-3110 301-682-7772 Voice 301-682-7666 FAX Internet URL http://www.arius.comArticle: 26099
test
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