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 Mon, 12 Apr 1999 19:44:25 GMT, ems@riverside-machines.com.NOSPAM wrote: >The amount of logic required to detect an illegal state isn't really >the problem. > yada i've been away a few days, and I got a couple of mails in response to this pointing out that the error detector could be pipelined, since it may not matter how long the fsm takes to recover, which is a good point. evanArticle: 15826
I like to possibilities offered by reconfigurable FPGAs, but it seems most of the devices around are designed to be loaded once on power up, and maybe reconfigured to do something different if the designer is being really ingenuitive. I've heard talk of some new FPGAs from Xilinx which alow partial reconfigurability (6000 series?), but i don't see these as being particularly more useful.... ? Anyone disagree here ? What i think would be really interesting, is being able to re-configure an entire FPGA really quickly (say 1 system clock period ideally). This translates into the idea of having 'layers' of FPGA config data which can be latched into the FPGA config area. The inactive 'layers' being updated or replaced while not in use -> pretty much like a video display where an image is built up in the off screen buffer before the active video buffer is toggled. The implications of such an architecture are of course ghastly amounts of configuration data flying around and a squadron of PentiumIV's working overtime trying to place, route and load all this into our liquid cooled UltraFPGA. But if applied to a pretty small block of reconfigurable logic i think the creative designer could do some really neat things. ....any comments ? Are there any such devices out there already ? Regds, Mark K. -- Life is about wanting what you don't know you don't want. --Article: 15827
Article: 15828
Hello, I just want to first thank everyone who wrote back to my earilier enquiries. I have one more. I am using the FPGA express package from synopsys to do my synthesis. I used the intelliflow program they have to try to synthesize a design. When I bring my design into intelliflow, it checks my design to make sure that it is synthesizable, and it passes the checks. When I then run the routing and placement routines after targeting a device, they also work fine. But, the program gets stuck in the process of generating a board level symbol. This symbol I guess would show me the schematic of how the CLBs are actualyl wired toghether. The program stalls on the ORD file generation method which has to be completed before a board level symbol can be generated. ANy help is greatly appreciated. -- Eldho Kuriakose Nature Photography http://kepler.poly.edu/~ekuria01/ -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your OwnArticle: 15829
hi, you're correct ... viewlogic was spun back off (minus a piece or two) and is back as an independent company. rk Austin Franklin wrote: > Though true a little while ago, I believe that has recently changed and > Viewlogic is a separate company. > > Thanks, > > Austin > > APS <resp@associatedpro.com> wrote in article > <3712ACB8.65695A01@associatedpro.com>... > > Synopsys owns Viewlogic. They are the same. XILINX may be a rev back. > > Also the viewlogic can do multiple vendors. > > > > Austin Franklin wrote: > > > > > Anyone know of any differences between the Viewlogic offering of FPGA > > > Express and the Xilinx offering of FPGA Express? > > > > > > Thanks, > > > > > > Austin Franklin > > > austin@darkroom.com > > > > -- > > __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ > > > > Richard Schwarz, President EDA & Engineering Tools > > Associated Professional Systems (APS) http://www.associatedpro.com > > 3003 Latrobe Court richard@associatedpro.com > > Abingdon, Maryland 21009 > > Phone: 410.569.5897 Fax:410.661.2760 > > > > __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ > > > > > >Article: 15830
Actually, it is not that urgent, I just want to get the "you will be flamed to death if you ask these" questions so I know what's going on. I am new to this, and have no formal training, so an FAQ would be mighty useful :) Thanx DavidArticle: 15831
sounds like satisfiability to me. It is NP complete :) If you solve the problem generally, let me know, don't tell anyone else, and I will get the Nobel Prize :) All Hail Alan Turing! DavidArticle: 15832
I'm using a 16v8Z gal from AMD/ VANTIS. Its a zero-power device - sleeping when no input activity is sensed(50ns timeout) and waking up when any input transition is detected. After a few days trying to debug some inexplicable behaviour shown by a straightforward bit of combinational logic in it - basically during some system states certain input transitions were not 'coming through' to the outputs - I am now beginning to think that during the zero power mode this device will not reliably respond (ie wakeup and 'turn-on' the logic) to an input with a slow rise-time (say >200ns). The slow rise was from the o/p of a voltage comparator. I was able to speed up its rise-time to c100ns and found that the mis-behaviour was noticeably less frequent. I contacted the manufacturer and after a long wait got a less than explicit affirmation of this. For example they wouldn't say what was a suitable 'wakeup' rise time. Rien ca change. Maybe this 'phenomenon' is old hat but I have never seen any mention of it in the data sheets or app. notes. Yet one would think that it is a real caveat when interfacing these devices to comparators - hardly a rare application. Anyone else seen similar behaviour? regds Mike -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your OwnArticle: 15833
Be sure that you are not doing a process or two using both clock edges (positive and negative). For example two processes, one using the positive clk edge and another using the negative clk edge and operating on the same signal. That won't work. ekuria01@kepler.poly.edu wrote: > Hello, I am trying to synthesize a pretty simple interface board using Xilinx > xc4000 series and Fpga express. I am having some trouble. > > 1. When I go into FPGA express and load my code (a simple vhdl file with 1 > entity and 1 behavioral arch), it compiles fine. I see that the icon that > represents my code has a green check on it. When i clik on this icon, i get a > yellow box with a pic of an or gate on it. Now, I click on FPGA express > "create implimentation". This gives me an error saying that my negative clock > edge triggered event will not work. > > Error: This use of clock edge specification not supported > in routine QSM1 line 129 in file 'C:/din/qsm1.vhd' (HDL-109) > > this is line 129 > 129: if (ext_clk'event and ext_clk ='0') then -- and so on.. > > I know that teh xc4000 can do neg clock edges from looking at its CLB > architechture. Any ideas on why this is so ? Thanks a lot for any help. > -- > Eldho Kuriakose > Nature Photography > http://kepler.poly.edu/~ekuria01/ > > -----------== Posted via Deja News, The Discussion Network ==---------- > http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own -- __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ Richard Schwarz, President Associated Professional Systems Inc. (APS) email: richard@associatedpro.com web site: http://www.associatedpro.com Phone: 410-569-5897 Fax: 410-661-2760 __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/ __/Article: 15834
Hi, does anybody know how to implement a bidirectional FPGA port using MAXPLUSII VHDL tools ? I tried following construct: ------------------------------------------------ entity datamux is port ( DATA : inout std_logic_vector (15 downto 0); OE_n : in std_logic; OUT_DATA : in std_logic_vector (15 downto 0); TEST_OUT : out std_logic_vector (15 downto 0) ); end datamux; architecture rtl of datamux is begin -- put OUT_DATA to FPGA pin when enabled with OE_n -- this statement is synthesized correct DATA <= OUT_DATA when (OE_n = '0') else (others => 'Z'); -- here i want to read the FPGA input -- this does not work with ALTERA MAXPLUS Synthesizer -- same works with e.g. Compass ASIC Synthesizer TEST_OUT <= DATA; end rtl; ---------------------------------------------------------- Thanx AloisArticle: 15835
Brad Taylor <blt@cmln.com> writes: > Oooops! Big mistake here. I somehow sent this message to the news group > here instead of my co-worker. (I can't explain how, but I must have been > browsing the newsgroup when I started up the message). I guess it could > have been much worse. Does anyone know how to get a message off this > newsgroup? This isn't the first time I've wanted to retract a post, but > this is the first time I've had an excuse. You need to find out how to CANCEL articles (your newsreader should support this). The milk is still spilled, as the CANCEL has to follow your message before it can erase them. The message is thus seen by fewer people and possibly doesn't propagate as far in UseNet, but you never know if it wasn't archived by someone already. So better check next time what you send out and where to. Achim Gratz. --+<[ It's the small pleasures that make life so miserable. ]>+-- WWW: http://www.inf.tu-dresden.de/~ag7/{english/} E-Mail: gratz@ite.inf.tu-dresden.de Phone: +49 351 463 - 8325Article: 15836
This one should work: library IEEE; use IEEE.std_logic_1164.all; entity bidir is PORT (in_data : IN STD_LOGIC; data_oe: IN STD_LOGIC; out_data: OUT STD_LOGIC; bi_data: inout std_ulogic); end bidir; architecture bidir_arch of bidir is begin process (in_data, data_oe, bi_data) begin if data_oe='1' then bi_data <= in_data; out_data <= bi_data; else bi_data <= 'Z'; out_data <= bi_data; end if; end process; end bidir_arch; -- ____________________________ Thomas Hellerforth, AMO GmbH mailto: hellerforth@amo.deArticle: 15837
I have a qty of 10 Ariel ARUBA DSP cards. Here is some basic info. ISA board carries two C31 floating-point DSPs. The PC-C31 Aruba is a $2495 ISA bus PC-expansion card with two 40-MHz TMS320C31 floating-point DSPs, 2 Mbytes of zero-wait-state SRAM, and two high-speed RS-422 serial ports. The SRAM is split into asymmetric banks, one per DSP. One bank contains 1.5 Mbytes; the other contains 512 kbytes. In addition, each DSP can access a dual-ported, 8-kbyte RAM. norman@pcseng.comArticle: 15838
Thanks Thomas, The code is now accepted by the MAXPLUS synthesizer. But ... Simulation of the resulting netlist shows that the block does not work as I expected. I want to have a birectional IO Pad, whose logic state can be read inside the chip (at 'out_data'). In your example the synthesizer places at tristate buffer to 'out_data'. So 'outdata' becomes High-Z exactly at the time I want to use the input function. Any ideas ? Alois Thomas Hellerforth wrote: > > This one should work: > > library IEEE; > use IEEE.std_logic_1164.all; > > entity bidir is > PORT (in_data : IN STD_LOGIC; > data_oe: IN STD_LOGIC; > out_data: OUT STD_LOGIC; > bi_data: inout std_ulogic); > end bidir; > ......... -- __ __ +-------------------------- |_ \ / \ ------------------------+ | Alois HAHN _\ \ \ | Thomas NEUROTH GmbH | | alois.hahn@neuroth.co.at | |\ \ | | Hietzinger Hauptstr. 22 | | Tel.: (+43/1) 8775645-51 | | \ \| | | | Fax.: (+43/1) 8764920 | \ \ | | A-1130 Vienna/Austria | +-------------------------- |__/ \__| ------------------------+ http://www.neuroth.co.atArticle: 15839
Mark, I don't know of any such devices. I have used the Xilinx 6200 series parts and played around with my thoughts for an appropriate swapping architecture for hardware. The best ideas seemed to come from the theory of memory paging. I presented a paper on this at DesignCon98 that you may be interested in reading. Follow the link below and go to the "Reconfigurable Computing" section. http://www.netrino.com/Publications/ Cheers, MichaelArticle: 15840
You may find the FAQ on The Programmable Logic Jump Station of some use (see http://www.optimagic.com/faq.html). Also, the entire jump station is packed with info on devices, software, boards, books, etc.--all about programmable logic. ----------------------------------------------------------- Steven K. Knapp OptiMagic, Inc. -- "Great Designs Happen 'OptiMagic'-ally" E-mail: sknapp@optimagic.com Web: http://www.optimagic.com ----------------------------------------------------------- David wrote in message <3716BA8E.58C5@hopper.net>... >Actually, it is not that urgent, I just want to get the "you will be >flamed to death if you ask these" questions so I know what's going on. >I am new to this, and have no formal training, so an FAQ would be mighty >useful :) >Thanx >DavidArticle: 15841
Hi, I would like to ask how can I make use of the wire-AND property in the longline of Xilinx XC4000 series? Can I write VHDL code to describe it? or I can only do it by manual routing? Any advice is much appreciated. Regards, KenArticle: 15842
You can find lots of information on FPGAs and other kinds of programmable logic on The Programmable Logic Jump Station at http://www.optimagic.com. There is a Xilinx Student Edition, which includes development software. There is also a companion Student Lab Book. Both are list at http://www.optimagic.com/books.html#Xilinx. Other FPGA and CPLD companies provide free or low-cost versions of their software, some available for download over the web. See http://www.optimagic.com/lowcost.shtml. ----------------------------------------------------------- Steven K. Knapp OptiMagic, Inc. -- "Great Designs Happen 'OptiMagic'-ally" E-mail: sknapp@optimagic.com Web: http://www.optimagic.com ----------------------------------------------------------- Vic Lopez wrote in message <7f3qfd$43b$1@news.chatlink.com>... >Hi everyone, > Has been dissapointing not to get any responses to my question. So I'll >pose it again. What is the best way to learn about fpga's, specially XILINX? >I want to be able to get my hands dirty with them, since I think they are >becoming an important trend in logic design. Schools don't do a good job, so >I want to know what's available ( student editions maybe?) for people like >me that want to learn about this subject. Thanks in advanced. Vic Lopez. > >Article: 15843
Hi, I was wondering what the following slice of code is all about (it is from the std_logic_arith package in the IEEE library) : in function CONV_INTEGER (line 2045 :) ) ... assert ARG'length <= 32 report "ARG is too large in CONV_INTEGER" severity FAILURE; ... my question : should the 'smaller than or equal' sign not be a 'larger' sign? If not, could you please explain? greetings, PieterArticle: 15844
Hi Guys, My name is Dimitrios and I am a Cybrarian at HR Advisors in Laguna Beach Ca. I have used usenet for years for my own personal interests but I have just started to tap it's usefulness as a tool to find smart technical guys like you. As you saw in the header we are looking to fill a benchmarking position at ENCAD in San Diego. See the job description below. Please send your resume directly to me with job code ENCAD-85 in the subject header to dkhradvise@hotmail.com. Thanks, Dimitrios ;-) . Job Description: Encad, Inc. is a fast-growing leading innovator of digital imaging and inkjet digital imaging, located in San Diego Design and perform comprehensive technical testing of in-house and competitive printer/software solutions as well as a wide variety of applications testing. Required: Bachelor's degree in related field or equivalent and a minimum of 3 years of relevant experience in a high tech, fast-paced marketing or sales environment. Must possess a solid understanding of scientific methods, as well the ability to write and produce comprehensive reports. A demonstrated success working with cross-functional teams and third-party relationships is critical. Computer knowledge of Microsoft Office including Word, Excel, PowerPoint. Strong communication skills and experience giving professional presentations. Will require occasional domestic and international travel. Preferred: In-depth knowledge of digital imaging technology and marketplace. Product development & management experience with a high tech product. Knowledge of color inkjet printers and software RIPs.Article: 15845
I am trying to develop a system that uses an Altera 10K and ~ 32M x 16 of Flash memory. All the flash products I have found either are too small (density) or require an enormous amount of control and configuration, which is fine for a microprocessor or a microcontroller. But when you try and incorporate all that into an FPGA, you loose most of your gates to the initial setup. Are there any Flash product that have a TAP (JTAG) port on the chip? This way I could program it externally and just read using the Altera. Has anybody seen a canned JTAG interface for the 10K series? ie, can you embed a TAP programmer and use that to program the Flash from the Altera. I know you can program the 10K itself with its built in JTAG port, but can you use that interface through the 10K to program something else. Should I be looking at some other method of non volatile memory to interface to the 10K? Thanks. PaulArticle: 15846
E-Lab Digital Engineering, Inc. announces the commencement of The Embedded Design Bulletin, a quarterly newsletter overviewing current embedded electronics design topics. Geared toward microcontroller circuit designers and other embedded design engineers, it contains recent press releases, frequently-asked questions & how-to's on current design topics, as well as cost-saving design techniques every designer can benefit from. We encourage you to offer your comments, tips, suggestions, and circuit design know-how to share with our readers. If you'd like to be on our mailing list (free of charge) just visit http://www.elabinc.com/newsletr.htm and enter your mailing information (you can select postal or e-mail delivery). In addition, you can also download a copy of the current Embedded Design Bulletin (Acrobat PDF format) from that page. We hope you find the Embedded Design Bulletin of interest now and in the years ahead. Thank you, Todd Peterson E-Lab Digital Engineering, Inc. (816) 257-9954 FAX: (816) 257-9945 http://www.elabinc.comArticle: 15847
Mentor Graphics is pleased to offer Top Down Hands On workshops in several cities in the Southeast United States. These workshops allow the attendee to spend some hands on time with Renoir, ModelSim, and LeonardoSpectrum, Mentor Graphics integrated FPGA design flow. Workshops are being offered in Tampa, Florida on April 27th and in the Ft. Lauderdale are on May 12th. A workshop in Research Triangle Park, North Carolina is tentatively scheduled for May 18th as well. If you would like more information, check out the detailed information on Mentor's web site at: http://www.mentor.com/region/SE/Workshops/index.html -- ...Mike +--------------------------------------------------------------+ | Michael P. Walsh mike@rtp-nc.mentorg.com | | mike_walsh@mentorg.com | | Applications Engineering Manager | | Mentor Graphics Corporation V:(919) 484-2505 | | 2525 Meridian Parkway, Suite 260 F:(919) 544-0701 | | Research Triangle Park, North Carolina 27713 | +--------------------------------------------------------------+Article: 15848
This is a multi-part message in MIME format. --------------0052CFBE47E837A6B079C3D2 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit There were a couple of papers presented at FPGA'98 conference: Steve Trimberger- "Scheduling Designs into a Time-Multiplexed FPGA" Chang, Marek-Sadowska - "Partitioning Sequential Circuits on Dynamically Reconfigurable FPGAs" FPGA conference is sponsored by ACM, you may be able to get copies through them. Steve's paper concentrated on an FPGA with 8 configuration layers, the other talked more generally on partitioning issues. I don't know if any of these type devices have actually been fabricated yet, I think efficiency and tools issues are still being investigated. - John Mark wrote: > > I like to possibilities offered by reconfigurable FPGAs, but it seems > most of the devices around are designed to be loaded once on power up, > and maybe reconfigured to do something different if the designer is > being really ingenuitive. > > I've heard talk of some new FPGAs from Xilinx which alow partial > reconfigurability (6000 series?), but i don't see these as being > particularly more useful.... ? Anyone disagree here ? > > What i think would be really interesting, is being able to re-configure > an entire FPGA really quickly (say 1 system clock period ideally). This > translates into the idea of having 'layers' of FPGA config data which > can be latched into the FPGA config area. The inactive 'layers' being > updated or replaced while not in use -> pretty much like a video display > where an image is built up in the off screen buffer before the active > video buffer is toggled. > > The implications of such an architecture are of course ghastly amounts > of configuration data flying around and a squadron of PentiumIV's > working overtime trying to place, route and load all this into our > liquid cooled UltraFPGA. But if applied to a pretty small block of > reconfigurable logic i think the creative designer could do some really > neat things. > > ....any comments ? > > Are there any such devices out there already ? > Regds, > Mark K. > > -- Life is about wanting what you don't know you don't want. -- --------------0052CFBE47E837A6B079C3D2 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for John L. Smith Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: John L. Smith n: Smith;John L. org: Visicom Imaging Products adr: 1 Burlington Woods;;;Burlington;MA;01803;USA email;internet: jsmith@visicom.com title: Principal Engineer tel;work: 781-221-6700 tel;fax: 781-221-6777 note: http://www.visicom.com/products/Vigra/index.html x-mozilla-cpt: ;0 x-mozilla-html: TRUE version: 2.1 end: vcard --------------0052CFBE47E837A6B079C3D2--Article: 15849
Mark wrote: > I like to possibilities offered by reconfigurable FPGAs, but it seems > most of the devices around are designed to be loaded once on power up, > and maybe reconfigured to do something different if the designer is > being really ingenuitive. > > I've heard talk of some new FPGAs from Xilinx which alow partial > reconfigurability (6000 series?), but i don't see these as being > particularly more useful.... ? Anyone disagree here ? Yes, Lots of work was done with the 6200. In my opinonbeing able to access the configuration store through a ram style interface (address,data pairs) was extreamly powerful and allowed you to conveniently change very small portions of the design very quickly. The map and mask registers were really innovative structures that allowed all sorts of interesting behavior. > > > What i think would be really interesting, is being able to re-configure > an entire FPGA really quickly (say 1 system clock period ideally). This > translates into the idea of having 'layers' of FPGA config data which > can be latched into the FPGA config area. The inactive 'layers' being > updated or replaced while not in use -> pretty much like a video display > where an image is built up in the off screen buffer before the active > video buffer is toggled. You have to look at Andre DeHons DPGA (someone help with net pointerif one exists) and Xilinx has a patent on such a device. Lockheed-Sanders go a DRAPA contract to do such a device but I'm not sure how that worked out. There has been some amount of work in the area. > > > The implications of such an architecture are of course ghastly amounts > of configuration data flying around and a squadron of PentiumIV's > working overtime trying to place, route and load all this into our > liquid cooled UltraFPGA. But if applied to a pretty small block of > reconfigurable logic i think the creative designer could do some really > neat things. > > If your UltraFPGA is so cool why not just have it do theplace and route??? Remember "The Software IS the Computer" -- Steve Casselman, President Virtual Computer Corporation http://www.vcc.com
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