Custom Search
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
hello: when i am doing timing simulator(using the reference design xapp205.zip :fifoctlr_ccmw1.v/fifoctlr_ccmw2.v) for a project,the follwoing message occures: "P1/U1/U11/BU0/INTERNAL_BLOCKRAM--READ Violation.Attempt to read from cell that is also being written to." the capacity of the dual block ram i mentioned is 16x1024--64x256,that is,when write_enable is active,read_enable must inactive.But when i replace it with 16x2048---64x512,it is right,and the phenomena of the warning is not . why?Article: 26851
It's the crashing during p&r bug .... recently reintroduced by popular demand ... coming to you at a p&r run soon! Fixed by setting VCCGND_OFF=1 Gary. Ray Andraka wrote: > I'm using it. It fixes a number of bugs, but there is one that it created. > Can't remember off hand which one it was, as I currently have several cases > open. It is supposed to be fixed in the service pack due out next week. > > "S. Ramirez" wrote: > > > > Dear Newsgroupies, > > Has anyone here used Xilinx 3.2i? I am presently using 3.1i, with > > service_pack_31i_3_2_03i_pc and am fairly happy with it. Does 3.2i with SP4 > > bring anything great to the table? Does it have any bugs that anyone has > > uncovered? > > Thanks. > > -Simon Ramirez, Consultant > > Synchronous Design, Inc. > > -- > -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: 26852
In article <SeJL5.25863$68.6416528@typhoon.tampabay.rr.com>, sramirez@deleet.cfl.rr.com (S. Ramirez) wrote: > That sucks! I thought all those Unix systems were compatible "Compatible" in Unix-land has never really meant binary image compatible. Frequently, it means "you can recompile the source with only a modest amount of aggro (provided you know what you're doing)" :-) -- Steve Rencontre http://www.rsn-tech.co.uk //#include <disclaimer.h>Article: 26853
Paul Smith wrote: > Leon Heller wrote: > > > > > I've just been quoted 14-18 weeks lead-time for some XC2S50-5TQ144C by > > our distributor, Insight Memec, in the UK > > > > I was just quoted 16 weeks for XC2S30-5VQ100C > > Anyone able to buy XC2S30 or XC2S50 with reasonable leadtimes? Maybe somebody should remind Xilinx and/or the disti's of the concept of time-to-market. Around here 16 weeks is well beyond the no design-in limit. Perhaps Xilinx could take some people/fab time off the G-gate devices to actually get some useful parts out the door.Article: 26854
Hi all, I would like to ask anyone knows anything about JBits? -- Cheers, Ivan ~~~~~~~~~~~~~~~~ Ivan LeungArticle: 26855
Ivan Leung wrote: > > Hi all, > > I would like to ask anyone knows anything about JBits? Yup, What do you want to know? Phil -- --------------------------------------------------------------------- __ / /\/ Dr Phil James-Roxby Direct Dial: 303-544-5545 \ \ Staff Software Engineer Fax: Unreliable use email :-) / / Loki/DARPA Email: phil.james-roxby@xilinx.com \_\/\ Xilinx Boulder ---------------------------------------------------------------------Article: 26856
Duane <junkmail@junkmail.com> wrote: ... : ngdanno, ngd2vhdl, etc... The "Design manager" does not work under wine, : as for as I know. So I have a somewhat elaborate but very easy (I think) : to use script that runs through the process, creating directories, I didn't notice you bugging the Wine developpers on comp.os.ms-windows.emulators with good bug reports ... : ... : I actually purchased VMware, since I still need it to run the synthesis : tools. But that is all I use it for. All my other tools are either : native Linux or run under wine. For purely running Windows application on Linux, I guess one should have a look at Win4Lin. It requires much less resources than VMWare and runs faster. The new beta should also be able to access the parallel port. Bye -- Uwe Bonnes bon@elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------Article: 26857
In article <8tpjil$hkf$1@eng-ser1.erg.cuhk.edu.hk>, Ivan Leung <khleung@cse.cuhk.edu.hk> wrote: >Hi all, > > I would like to ask anyone knows anything about JBits? What is there to know? It is an abstract API, written in java, designed to directly manipulate the Virtex bitfile. -- Nicholas C. Weaver nweaver@cs.berkeley.eduArticle: 26858
"S. Ramirez" wrote: > > Andy, > That sucks! I thought all those Unix systems were compatible and > that's why Peter had no problem running his EDA tools under Linux. > Apparently, I have been mislead by the Linux people! Good thing I haven't > switched -- yet. Different CPUs! My Sparcstation has (obviously) a Sparc processor; HP/UX workstations have HP processors, etc etc. I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can get source code for things (and get it to compile, which never ever ever works the first time). If the source is not available, like say for Star Office, I can't use it because it's not a x86 box. I was rarely able to get Linux code to build on an x86 box. Configure scripts never work. Whatta bunch of hooey: "Just modify the source code!" Gimme a break. -- 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 u "It is better to be silent and thought a fool, than to send an e-mail to the entire company and remove all doubt."Article: 26859
Andy Peters wrote: > > "S. Ramirez" wrote: > > > > Andy, > > That sucks! I thought all those Unix systems were compatible and > > that's why Peter had no problem running his EDA tools under Linux. > > Apparently, I have been mislead by the Linux people! Good thing I haven't > > switched -- yet. Ohh, go ahead and switch! Just expect, contrary to some of the hype, to spend a fair amount of time (at least initially) messing with things to get everything to work the way you want. Some Unix sysadmin experience helps out a whole lot here, but is not essential. I started using Linux back in April, and have been extremely happy with it as a CAE platform. It is very nice to be able to use a really high powered platform, configured exactly the way I want it, at a remarkably low price. > > Different CPUs! My Sparcstation has (obviously) a Sparc processor; > HP/UX workstations have HP processors, etc etc. > > I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can > get source code for things (and get it to compile, which never ever ever > works the first time). If the source is not available, like say for > Star Office, I can't use it because it's not a x86 box. > > I was rarely able to get Linux code to build on an x86 box. Configure > scripts never work. Whatta bunch of hooey: "Just modify the source > code!" Gimme a break. Yep, that's for sure. I much preferred Imake, which had the potential if correctly configured of truly being portable. But seriously broken vendor implementations really caused problems here, and it appears its days are waning. -- My real email is akamail.com@dclark (or something like that).Article: 26860
Uwe Bonnes wrote: > > Duane <junkmail@junkmail.com> wrote: > > ... > : ngdanno, ngd2vhdl, etc... The "Design manager" does not work under wine, > : as for as I know. So I have a somewhat elaborate but very easy (I think) > : to use script that runs through the process, creating directories, > > I didn't notice you bugging the Wine developpers on > comp.os.ms-windows.emulators with good bug reports ... The design manager and flow engine sure do look nice, at least they did a couple of years when I last used them. And they were definitely helpful when I was first learning the tools. But I suspect a significant percentage of experienced designers abandon them eventually. I see that I may have given the idea that I wasn't using them because I couldn't get them to work with wine, but the real reason is because I prefer the control I get from the script method. -- My real email is akamail.com@dclark (or something like that).Article: 26861
"news" <apteuk@vsnl.net> writes: > > type mismatches. It seems like some operations (add, subtract) what the > > inputs to be integers while other operations (and, or) what them to be > > bit_vectors or std_logic_vectors. How can I do this? I have tried > u can try to include: > library ieee; > use ieee.std_logic_arith.all; > use ieee.std_logic_unsigned.all; No, don't do that. The correct conversion functions (unless your tools do not accept them) are in ieee.numeric_std. See e.g. http://tech-www.informatik.uni-hamburg.de/vhdl/doc/faq/FAQ1.html#4.10 http://tech-www.informatik.uni-hamburg.de/vhdl/doc/faq/FAQ1.html#integer_bit_vector ColinArticle: 26862
I represent a mid-sized firm in Pittsburgh, PA who design, manufacture and market proprietary electronic equipment-Seeking Senior Hardware Engineers to be responsible for managing the digital hardware team- Requires a BSEE with more than 5 years experience in analog and digital circuit design. Experience with Xilinx FPGAs and C and Assembly programming also required. Contact Autumn Feldmeier 724)387-2668 or autumn@osearch.com Sent via Deja.com http://www.deja.com/ Before you buy.Article: 26863
Have a look at: http://www.xilinx.com/products/software/jbits/index.htm for more information on JBits. -- Steve -- 11/1/00 "Ivan Leung" <khleung@cse.cuhk.edu.hk> wrote in message news:8tpjil$hkf$1@eng-ser1.erg.cuhk.edu.hk... > > Hi all, > > I would like to ask anyone knows anything about JBits? > > -- > Cheers, > Ivan > > ~~~~~~~~~~~~~~~~ > Ivan LeungArticle: 26864
On Sun, 29 Oct 2000 21:30:45 GMT, JoeG <jNOgalleSPAM@pacbell.net> wrote: >I already have: > >1) Ashenden's "Designers guide to VHDL" >2) Stefan Sjoholm & Lindh's "VHDL for designers" >3) Douglas Smith's "HDL Chip Design" >4) Kurup & Abbasi's "Logic synthesis using Synopsys" try Lee VHDL coding and logic syntesis with synopsis ISBN 0-12-440651-3Article: 26865
This is a multi-part message in MIME format. --------------95A2AFBED30F21EE0EE4BFDD Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I apologise for the size of this paper. Until now no-one has complained. The paper is about to be removed from the websitre so download time will no longer be an issue (otherwise I would replace it with a more compact version). Kind regards, Satnam Singh --------------95A2AFBED30F21EE0EE4BFDD Content-Type: text/x-vcard; charset=us-ascii; name="Satnam.Singh.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Satnam Singh Content-Disposition: attachment; filename="Satnam.Singh.vcf" begin:vcard n:Singh;Satnam tel;pager:4087182588@messaging.cellone-sf.com tel;cell:(408) 718 2588 tel;fax:(208) 293-9440 tel;home:(408) 377 9982 tel;work:(408) 879 4693 x-mozilla-html:TRUE url:http://www.xilinx.com/labs/satnam org:<br><img src="http://www.xilinx.com/images/xlogoc.gif" alt="Xilinx">;Xilinx Labs adr:;;2100 Logic Drive;San Jose;CA;95124-3450;USA version:2.1 email;internet:Satnam.Singh@xilinx.com title:Researcher note:UK Cellular +44-7979-648412 x-mozilla-cpt:;-1856 fn:Satnam Singh end:vcard --------------95A2AFBED30F21EE0EE4BFDD--Article: 26866
This is a multi-part message in MIME format. --------------1BED79354097BF48FAFCFF9D Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I don't know if you've read the paper, but it is about why we NEED RLOCs. That is why there is a question mark in the title. There is still a huge advantage to be gained by floor-planning your design if you happen to have a good enough intuition of the underlying fabric. I give two examples where my flooor-plan beat the automatic tools and where where I failed -- but that was because my floor-plan was not good enough -- had I put in more effort I am sure I could come up with a floor-plan that beat the automatic tools (in this case for butterfly networks). I speculate that in the future we may not need them if the software gets clever enough, but that is certainly not the case for Xilinx FPGAs at the moment or the immediate future. Kind regards, Satnam Singh husby@my-deja.com wrote: > Of course an easier experiment is to remove the rlocs > from your existing designs and find that all of > them perform much worse or can't be routed at all. > > The same goes for mapping directives. > > Muzaffer Kal <muzaffer@dspia.com> wrote: > > I think this paper would be helpful to some people here and > > interesting for most. Interesting read. Available at : > > http://www.xilinx.com/labs/satnam/death_of_the_rloc.pdf > > Sent via Deja.com http://www.deja.com/ > Before you buy. --------------1BED79354097BF48FAFCFF9D Content-Type: text/x-vcard; charset=us-ascii; name="Satnam.Singh.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Satnam Singh Content-Disposition: attachment; filename="Satnam.Singh.vcf" begin:vcard n:Singh;Satnam tel;pager:4087182588@messaging.cellone-sf.com tel;cell:(408) 718 2588 tel;fax:(208) 293-9440 tel;home:(408) 377 9982 tel;work:(408) 879 4693 x-mozilla-html:TRUE url:http://www.xilinx.com/labs/satnam org:<br><img src="http://www.xilinx.com/images/xlogoc.gif" alt="Xilinx">;Xilinx Labs adr:;;2100 Logic Drive;San Jose;CA;95124-3450;USA version:2.1 email;internet:Satnam.Singh@xilinx.com title:Researcher note:UK Cellular +44-7979-648412 x-mozilla-cpt:;-1856 fn:Satnam Singh end:vcard --------------1BED79354097BF48FAFCFF9D--Article: 26867
Satnam Singh <Satnam.Singh@xilinx.com> wrote: >I apologise for the size of this paper. Until now no-one has complained. >The paper is about to be removed from the websitre so download time will >no longer be an issue (otherwise I would replace it with a more compact >version). just curious, why are you removing it ? Muzaffer http://www.dspia.comArticle: 26868
Steve,
Correct you are! What I really meant to say is "Why doesn't Xilinx (or
other FPGA vendor) make a binary for PCs running Linux?" I thought this was
the next big tsunami.
-Simon Ramirez, Consultant
Synchronous Design, Inc.
> > That sucks! I thought all those Unix systems were compatible
>
> "Compatible" in Unix-land has never really meant binary image compatible.
> Frequently, it means "you can recompile the source with only a modest
> amount of aggro (provided you know what you're doing)" :-)
>
> --
> Steve Rencontre http://www.rsn-tech.co.uk
> //#include <disclaimer.h>
>
>
Article: 26869
Now this is what I call a good, honest ad.
-Simon Ramirez, Consultant
Synchronous Design, Inc.
<autumn@osearch.com> wrote in message news:8tq49a$mnu$1@nnrp1.deja.com...
> I represent a mid-sized firm in Pittsburgh, PA who design, manufacture
> and market proprietary electronic equipment-Seeking Senior Hardware
> Engineers to be responsible for managing the digital hardware team-
> Requires a BSEE with more than 5 years experience in analog and digital
> circuit design. Experience with Xilinx FPGAs and C and Assembly
> programming also required.
> Contact Autumn Feldmeier 724)387-2668
> or autumn@osearch.com
>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
Article: 26870Andy Peters <"apeters <"@> n o a o [.] e d u> writes: > works the first time). If the source is not available, like say for > Star Office, I can't use it because it's not a x86 box. You can find this on Sun's website: Sun is making the source code for its StarOffice[tm] software suite freely available under the GNU General Public License (GPL). OpenOffice.org and the OpenOffice.org Foundation will host the CVS tree (source code) and maintain the reference implementation of the software, APIs and file formats. Development will take place in the open, with all contributors working from the same source. According to the principles of free software and the GPL, all improvements to the software are contributed back to the project and are subsequently available to all. For further details take a look at http://www.sun.com/developers/openoffice/ You can download OpenOffice, the implied future tense in the announcement is actually past tense by now. > I was rarely able to get Linux code to build on an x86 box. Configure > scripts never work. Umm, I regularly build all sorts of stuff from source on various x86 boxes, running different versions of Linux from different distros and most of the time configure scripts work and everything compiles just as it should. I have to admit, though, that I read the README's and INSTALL's. > Whatta bunch of hooey: "Just modify the source code!" > Gimme a break. Well, that's the whole idea of free software, you *can* modify the source if you want to use it for something that it was not written for. You can, for example, port StarOffice to a PPC machine because you can have the source. You don't have to, however. If you don't like the idea, then just pretend you don't have the source and if it isn't available for or does not run on the platform of your choice, then tough, just like with closed source stuff. Zoltan -- +------------------------------------------------------------------+ | ** To reach me write to zoltan in the domain of bendor com au ** | +--------------------------------+---------------------------------+ | Zoltan Kocsi | I don't believe in miracles | | Bendor Research Pty. Ltd. | but I rely on them. | +--------------------------------+---------------------------------+Article: 26871
"S. Ramirez" <sramirez@deleet.cfl.rr.com> writes: > Andy, > That sucks! I thought all those Unix systems were compatible and > that's why Peter had no problem running his EDA tools under Linux. It is a little bit more complicated. You see Windows runs on x86 only, (although NT used to run on Alpha and PPC, but they were dropped) and thus you tend to forget that different processors have different architectures and instruction sets. In the unix world, however, we have to face the problem that any given bitstream might encode completely different instructions for different processors (if it means anything at all for a given processor). Regardless of the operating system, which from a processor's point of view is just code as anything else, you can't exchange binaries between processors. You can emulate a processor on an other but that is a rather slow process. Unix compatibility means that the system calls (i.e. the system/application program boundary) are more or less the same between different versions of unix, at least on the source level. > Apparently, I have been mislead by the Linux people! Good thing I haven't > switched -- yet. They did not necessary mislead you. Linux is indeed capable of running x86 Solaris binaries through the iBCS library. However, when a vendor sells you a Solaris tool it is almost surely for Sparc processor and would not run even on a native Solaris environment if it was on x86 iron. It is like if Windows was available for the PowerMac and you tried to run your x86 Windows application there, it would not run for the PPC can't understand x86 instructions, regardless whether the OS is MacOS or Windows. You should not, however, brush the compatibility claim of free software away before you try to appreciate what's behind it. Linux supports (runs on) more processors than any other OS except *BSD. Windows runs on 1, namely the x86. You can face severe compatibility issues between Windows versions. Interestingly enough a lot of free code runs on all sorts of CPUs each of them running all sorts of unix variants *and* the free code often runs on Windows as well. In case of devtools, like gcc, it can be even more mind boggling. When you create a crosscompiler you can compile gcc on some (processor-A,opsys-A) platform to run on (processor-B,opsys-B) and generate code for (processor-C,opsys-C). This is not for the faint in heart, but you can do it. As per if it is good or bad that you have not switched, it is only up to you. If Windows allows you to do your job, you feel comfortable with it and you are not in the mood of learning a new system (and paradigm), then don't switch. Using Linux is not compulsory. In fact, because of the lack of support from tool vendors, AFAIK it is impossible to design an FPGA from scratch to bitstream using Linux only (sans running a virtual Windows machine within an x86 Linux). There is increasing demand for Linux tools, which the vendors seem to respond (albeit very sluggishly) but currently Windows is the way to go. If Windows can satisfy your needs, then you should not switch just because Linux is a buzzword and a cool thing to use. Zoltan -- +------------------------------------------------------------------+ | ** To reach me write to zoltan in the domain of bendor com au ** | +--------------------------------+---------------------------------+ | Zoltan Kocsi | I don't believe in miracles | | Bendor Research Pty. Ltd. | but I rely on them. | +--------------------------------+---------------------------------+Article: 26872
Hi all I try to program a Virtex with a 18V02 ISP Prom using SelectMap mode. But when I made the PCB, I missed something obvious and i connected the pin D0 of the PROM to the pin D0 of the FPGA and the pin D7 to the pin D7. Xilinx says that the good connexion is D0 => D7 and D7 => D0. ;-))) Now, how can i swap these bits ? The first solution is to swap them on the board, but it would be more simple to swap them in the bitstream. There is a "swap pins" option in the Prom File Formatter, but only for Hex format, not for Mcs or Exor, and only the 2 last formats are accepted by the Jtag Programmer. Does anyone have a solution ? Thanks. -- J-P GOGLIO GETRIS S.A. 13 Chemin des Prés 38240 Meylan Tel : (+33) 4 76 18 52 10 E-mail : goglio@getris.com Fax : (+33) 4 76 18 52 01Article: 26873
On Wed, 01 Nov 2000 17:22:13 +0000, Rick Filipkiewicz <rick@algor.co.uk> wrote: > > >Paul Smith wrote: > >> Leon Heller wrote: >> >> > >> > I've just been quoted 14-18 weeks lead-time for some XC2S50-5TQ144C by >> > our distributor, Insight Memec, in the UK >> > >> >> I was just quoted 16 weeks for XC2S30-5VQ100C >> >> Anyone able to buy XC2S30 or XC2S50 with reasonable leadtimes? > >Maybe somebody should remind Xilinx and/or the disti's of the concept of >time-to-market. >Around here 16 weeks is well beyond the no design-in limit. If you've got a real design, get the FAE in and ask about availability. If you're nice to them, they may offer to supply an equivalent Virtex (make sure there is one first) at the S-II price. Worked for me. EvanArticle: 26874
On Wed, 01 Nov 2000 11:31:47 -0700, Andy Peters <"apeters <"@> n o a o [.] e d u> wrote: >I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can >get source code for things (and get it to compile, which never ever ever >works the first time). If the source is not available, like say for >Star Office, I can't use it because it's not a x86 box. > >I was rarely able to get Linux code to build on an x86 box. Configure >scripts never work. Whatta bunch of hooey: "Just modify the source >code!" Gimme a break. Exactly. And how many of us actually care about whether or not we have source code? As far as I'm concerned, it's just wasting valuable disk space. Linux developers seem to have a religious objection to distributing binaries, so we have to put up with second-rate installation programs and binaries which aren't necessarily the same as anyone elses. And what about actually installing Linux itself in the first place? I spent a day trying to get the latest Debian onto my new Dell portable recently. I went through the whole process at least six times, and couldn't get any further than a simple console interface; Debian doesn't know about AGP. I eventually found a sysadmin who managed to install X, but she couldn't get networking going. I remember installing Windows 3.1 from a box of floppies, and it was easier than this. The other bad news is that commerical Linux EDA software appears to be at Unix prices, not Windows prices, which is going to make it a non-starter. Evan
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