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
We are an OO software house based in Brussels, Belgium, Europe.
For a semicon based project we are undertaking here in Belgium we
currently have need someone with skills we do not possess.
We currently looking for an experienced VHDL designer with experience
of Synopsis/Ambit or similar. The work is targeted at eventual IC/FPGA
implementation.
The work would be for at least 12 months.
Because of the nature of the work we can only accept EU nationals for
this project.
If this is you can you urgently email or fax us
- your CV
- Availability
- Daily Rate
by either of the following means.
Email : wil@informate.be
Fax : [32 for Belgium]-2-7122051
Many thanks,
Jean Williams,
Informate International Ltd.
Article: 32251I was asking for the same reason. I use NT4 because it is stable and reasonably secure. I'd like to be able to use USB especially. The user interface is very windows95-like and I haven't had any problems running win95 stuff under NT...except for those applications that need unsupported hardware such as the USB, and the modem extensions that link caller ID into the system. I tried win2k briefly last year when I upgraded systems but found the likes of xilinx, synplicity and modelsim didn't fare too well. Nial Stewart wrote: > Ray Andraka wrote: > > > > NT4.0 pretty much has all the 'niceties' of win95. Are you running any of the > > CAE tools under w2000? If so, which ones, and are they stable too? > > Ray, > > As far as I know NT 4.0 still can't drive USB devices which is why > I was wondering about win2K. If it's as stable as NT with the > hardware drivers of win 98/95 then it sounds a good bet. > > Nial. -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32252
Dear all, I'm searching any board in 144 pin SO-DIMM format: --> flash module --> ram module --> fpga module --> dsp module --> MCU / MPU module --> ethernet module --> ... --> any module in 144 pin SO-DIMM format. Thank you for giving me the references of the modules you know. (Please send me a email to laurent.gauch@amontec.com) Reagards Laurent www.amontec.comArticle: 32253
The 16 bit FFT is a kernel that can be used for larger (powers of 16) FFTs. If you look at the FFT, it can be decomposed into smaller FFTs using the mixed radix algorithm. The decomposition consists of a data reordering and a phase rotation of the intermediate results. We currently are one of the vendors offering a shrink-wrapped 16 point core. We have used that core for some very high speed 256 point and 4K point FFTs. For those, the block RAM speed is the limiting factor. Our 16 point core is more compact and faster than our competitor's cores because we do not use radix 4 kernels to construct the core. We do have a phase rotator and reordering address generators available too, but they are not in state for general sale yet. finish wrote: > hello, > > From my modest background, i know that for performing the FFT > transform on an input signal, i have to extend it, if required, by > zeros to 2^n. > FFT is a global transform,i.e the whole input sequence should be > available. > > In practise, most often we take 1024 or 512, but i see some commercial > hardware implementation for just 16 input data. > > How far will this limited input size transform affect the overall > performance ? > > thanks > > H.S -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32254
You can compile the smaller devices using the webpack software on-line, which is free. For students and small projects this should be sufficient and will save you the expense of buying a seat. Antonio wrote: > I need the latest version of Xilinx Foundation Software, there is some > way to have it free ?? I'm just a poor student, thank you for your > answer ... > > Antonio D'Ottavio -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32255
Hi all, I want to examine a sub-module of my design regarding the exact number of used CLBs in the Xilinx XC4000E-Series. But now some LUTs to invert signals are not used any more because the internal ports are pads in this case and so some inversions are done in the IOBs. Therefore I don't get the correct number of CLBs. Is there a possibility to force the IOBs not to use the inverted input port or any other solution to get the right number? I use Xact-M3.1 . PascalArticle: 32256
Ray Andraka wrote: > > I was asking for the same reason. I use NT4 because it is stable and reasonably > secure. I'd like to be able to use USB especially. The user interface is very > windows95-like and I haven't had any problems running win95 stuff under NT...except > for those applications that need unsupported hardware such as the USB, and the modem > extensions that link caller ID into the system. I tried win2k briefly last year > when I upgraded systems but found the likes of xilinx, synplicity and modelsim > didn't fare too well. From what I've read over the last day, that's surprising. Most people who've tried Win2000 report that it's been very reliable and would recommend it. Are you sure you didn't try WinMe which most people seem to think should be avoided like the plague? Nial.Article: 32257
"Antonio" <dottavio@ised.it> wrote in message
news:fb35ea96.0106210439.30a69a6d@posting.google.com...
> I need the latest version of Xilinx Foundation Software, there is some
> way to have it free ?? I'm just a poor student, thank you for your
> answer ...
>
> Antonio D'Ottavio
You can find a list of free and low cost design software at
http://www.optimagic.com/lowcost.shtml
otherwise just try the gnutella file sharing network
(http://www.limewire.com)
----------------------------------------------------------------------------
----
Vincent MACK
LSI / DE / EPFL
CH-1015 Lausanne
+41 / (0)21 693 56 88
vincent.mack@epfl.ch
Article: 32258Here are five reasons for migrating from NT4 to W2K: 1. USB If you use a scanner, or multiplie printers, or a digital camera, or a portable ZIP drive; USB is a much better connectivity implementation and architecture. 2. True Plug'n'Play -- NT4SPxx has a sorta OK PnP implementation, but with W2K it was designed in from the start. It really does make a difference. 3. Support for >8GB disks on boot drive. With NT, support for large drives (>8GB) on the boot drive is conditional... the boot files and records must be within the "first 8GB" of the drive. This makes restoring the OS (or copyig onto a new and larger drive) a problematic process. The service packs add full large drive support, but this happens only after the initial load of the OS. With W2K, no issues with large drives. I run with two 60GB drives, and never a problem. 4. Support for FAT32 partitions Do you dual-boot between W98 and NT or W2K? If so, then you know that the only way to pass or share files between the two OSs is through a 2GB (max) FAT16 partition. FAT32 partitions disappear in NT, NTFS partitions dissapear in W98. With W2K I can install 1 copy of an application, and that will be enough for both W2K and W98. 5. W2K system console This is a new facility for repairing a system impaired by a wayward application to the point that it won't boot. It's a lifesaver when under stress. I've needed it only twice, once for a very badly written device driver and once for trying (in vain) to get Adaptec/Roxio Easy CD Creator to install properly. There are some other subtle differences between NT and W2K, but these 5 are the more compelling differences (in my mind). -- Bob Elkind, fpga design/consulting www.aracnet.com/~eteam mail to: eteam@aracnet.com Ray Andraka wrote: > I was asking for the same reason. I use NT4 because it is stable and reasonably > secure. I'd like to be able to use USB especially. The user interface is very > windows95-like and I haven't had any problems running win95 stuff under NT...except > for those applications that need unsupported hardware such as the USB, and the modem > extensions that link caller ID into the system. I tried win2k briefly last year > when I upgraded systems but found the likes of xilinx, synplicity and modelsim > didn't fare too well. >Article: 32259
Try www.nmi.co.uk/uengine.htm . andrew. "Laurent Gauch" <laurent.gauch@amontec.com> wrote in message news:3B31FC92.2070409@amontec.com... > Dear all, > > I'm searching any board in 144 pin SO-DIMM format: > --> flash module > --> ram module > --> fpga module > --> dsp module > --> MCU / MPU module > --> ethernet module > --> ... > --> any module in 144 pin SO-DIMM format. > > Thank you for giving me the references of the modules you know. > (Please send me a email to laurent.gauch@amontec.com) > > > Reagards > Laurent > www.amontec.com >Article: 32260
0.102546681502 = 102546681502 / 10^12 Let's say we want 11-bit precision, so 102546681502 / 10^12 = x / 2^11 so x (rounded) = .102546681502*2048 = 210 = 0b00011010010 so .102546681502 = 0b0.0001101001 (with 10 bits after the radix point) Since it's positive, you put the sign bit (zero) on the front for a 12-bit coefficient of 000001101001 Normally, coefficients are scaled so the magnitude is less than/equal to one. That is, a 12-bit coefficient (including sign bit) would represent a number in the range of -1..(2^11-1)/(2^11). Remember that having one binary digit to the left of the radix point in a binary number means your 12-bit number would have a range of about -2...2. ScopeFIR is a good program for FIR development and manipulation and scaling of coefficients. It's $100. -Kevin "Antonio" <dottavio@ised.it> wrote in message news:fb35ea96.0106210435.738ae496@posting.google.com... > I've the coefficient 0.102546681502 that I want to translate in > 2'complement fixed point binary number having 1 bit for the sign, 1 > bit before the point and 10 bit after, what I have to do ?? Is there > any software to make this automatically also for the opposite > conversion ??? > Thanks you a lot > > Antonio D'Ottavio > > Post a follow-up to this >Article: 32261
To get 10 bits after the decimal in fixed binary, multiply the coefficient by 2^10 and use the rounded integer part. 0.102546681502 * 2^10 = 105.007801858048 so your fixed point binary coefficient will be 105. One doesn't really need special software for this kind of conversion. Enjoy! Antonio wrote: > I've the coefficient 0.102546681502 that I want to translate in > 2'complement fixed point binary number having 1 bit for the sign, 1 > bit before the point and 10 bit after, what I have to do ?? Is there > any software to make this automatically also for the opposite > conversion ??? > Thanks you a lot > > Antonio D'Ottavio > > Post a follow-up to thisArticle: 32262
Hi All, I am looking for a tool to generate fault simulated test vectors (stuck-at faults) for FPGA devices from Xilinx (Virtex) and Altera (Apex). Ideally the tool would give a brief analysis of testability issues with the design when fault coverage is low. This would be to show our customer that low fault coverage was due to the design. Our previous tool vendor for CPLD and smaller FPGA devices has been Acugen Software (www.acugen.com). These tools accept Jedec, EDIF and various other back end files. Acugen may eventually be able to provide this capability for larger FPGA devices. We have evaluated TDX from Fluence (www.fluence.com) , but this tool was designed for use in ASIC design flows. It accepts the Verilog output files from the Xilinx and Altera tools, but does not currently model the FPGA primitives and UDPs correctly. Fluence may eventually provide capability for FPGA devices. Since FPGAs are non-scan for the purpose of ATPG, this further limits the list of potential tools. These are our customer's designs, and we are not allowed to alter the designs for testability purposes. For future designs, I may be able to convince their designers to use DFT tools. Does anyone know of any tools that are suited to this problem? Thanks for your time, PaulArticle: 32263
Synplicity 6.2 is "optimizing" an instantiated design. In particular, I have a design with instantiated Xilinx primitives including a carry chain. The synthesis is apparently flattening the xilinx primitives and doing its own optimization on them, which results in a multiplexer placed between the xorcy and the flip-flop. When I instantiate primitives, I don't expect to see them remapped. I didn't see this happening in 5.3.1. The instantiated primitives have a syn_black_box attribute on them. This new 'feature' makes Synplicity useless for the designs I am doing, and actually does damage for the large base I have already fielded. Has anyone else seen this? Any fixes? -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32264
Hi, is it possible to force the Device Fitter to use a specific device architecture type like P20V8C? I'm using the schematic design entry in combination with ABEL modules. Background: I want to use the complex OLMC Mode insted of the autoselected simple Mode because of the different feedback paths. But no output use the tristate feature so the fitter automaticaly selects simple mode. In the ABEL modul for my output latches I used the following statement: "device_id DEVICE 'P20V8C' ;" but the fitter seems to ignore it. When I try to force the fitter to use the tristate mode by applying the following eqation "Q.OE = 1;" I just the following error from the fitter: "Note 4072: Signal ROM_A18 cannot be assigned (to pin 22) because the enable equation on pin 22 can only be VCC or GND." Any comments how to tell the fitter to use the device architecture type P20V8C ? Thanks in advance LutzArticle: 32265
I realize the benefits of win2k, but it doesn't do me any good if the design tools are not stable running under 2k. Last year when I tried running win2k, I had numerous problems with xilinx alliance, synplicity, modelsim and others. If I can't run those reliably, then I can't migrate to the new OS. So my question still remains. Are these applications now stable under 2K? bob elkind wrote: > Here are five reasons for migrating from NT4 to W2K: > > 1. USB > If you use a scanner, or multiplie printers, or a digital camera, or a portable ZIP drive; > USB is a much better connectivity implementation and architecture. > > 2. True Plug'n'Play -- NT4SPxx has a sorta OK PnP implementation, but with W2K > it was designed in from the start. It really does make a difference. > > 3. Support for >8GB disks on boot drive. > With NT, support for large drives (>8GB) on the boot drive is conditional... the boot > files and records must be within the "first 8GB" of the drive. This makes restoring the > OS (or copyig onto a new and larger drive) a problematic process. The service packs > add full large drive support, but this happens only after the initial load of the OS. > > With W2K, no issues with large drives. I run with two 60GB drives, and never a > problem. > > 4. Support for FAT32 partitions > Do you dual-boot between W98 and NT or W2K? If so, then you know that the only > way to pass or share files between the two OSs is through a 2GB (max) FAT16 partition. > FAT32 partitions disappear in NT, NTFS partitions dissapear in W98. > > With W2K I can install 1 copy of an application, and that will be enough for both > W2K and W98. > > 5. W2K system console > This is a new facility for repairing a system impaired by a wayward application to > the point that it won't boot. It's a lifesaver when under stress. I've needed it only > twice, once for a very badly written device driver and once for trying (in vain) to > get Adaptec/Roxio Easy CD Creator to install properly. > > There are some other subtle differences between NT and W2K, but these 5 are the > more compelling differences (in my mind). > > -- Bob Elkind, fpga design/consulting > www.aracnet.com/~eteam > mail to: eteam@aracnet.com > > Ray Andraka wrote: > > > I was asking for the same reason. I use NT4 because it is stable and reasonably > > secure. I'd like to be able to use USB especially. The user interface is very > > windows95-like and I haven't had any problems running win95 stuff under NT...except > > for those applications that need unsupported hardware such as the USB, and the modem > > extensions that link caller ID into the system. I tried win2k briefly last year > > when I upgraded systems but found the likes of xilinx, synplicity and modelsim > > didn't fare too well. > > -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32266
No, it was Window2000 professional. This was June 2000 when I tried it. Perhaps the tools vendors have made changes to support it since then? Nial Stewart wrote: > Ray Andraka wrote: > > > > I was asking for the same reason. I use NT4 because it is stable and reasonably > > secure. I'd like to be able to use USB especially. The user interface is very > > windows95-like and I haven't had any problems running win95 stuff under NT...except > > for those applications that need unsupported hardware such as the USB, and the modem > > extensions that link caller ID into the system. I tried win2k briefly last year > > when I upgraded systems but found the likes of xilinx, synplicity and modelsim > > didn't fare too well. > > From what I've read over the last day, that's surprising. > Most people who've tried Win2000 report that it's been > very reliable and would recommend it. > > Are you sure you didn't try WinMe which most people seem > to think should be avoided like the plague? > > Nial. -- -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.comArticle: 32267
Your friends and colleagues need you! Time for an ad-hoc poll of the audience: Please reply if you have *first-hand* experience (go or no-go) running design tool XXXX [simulater | synthesiser | fpga-back-end-design | schematic capture | version control system | etc. ] on a windows 2000 professional sytsem. Please reply (to the newsgroup) with : tool name, GO - NO-GO result, tool SW version. Thanks for your help. Bob Elkind, eteam@aracnet.com fpga design/consultingArticle: 32268
--------------DBF5A4F521D1EDFE3EF1B49F Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Tool GO-NOGO Version ==================================================== Orcad Capture GO 9.2 Altera MAX+2 GO 10.0 Altera Quartus2 GO 1.0 Synplicify GO - -- Bob Elkind eteam@aracnet.com bob elkind wrote: > Your friends and colleagues need you! > > Time for an ad-hoc poll of the audience: > > Please reply if you have *first-hand* experience (go or no-go) running design tool XXXX > [simulater | synthesiser | fpga-back-end-design | schematic capture | version control system | etc. ] > on a windows 2000 professional system. > > Please reply (to the newsgroup) with : tool name, GO - NO-GO result, tool SW version. > > Thanks for your help. > > Bob Elkind, eteam@aracnet.com > fpga design/consulting --------------DBF5A4F521D1EDFE3EF1B49F Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <!doctype html public "-//w3c//dtd html 4.0 transitional//en"> <html> <tt>Tool GO-NOGO Version</tt> <br><tt>====================================================</tt> <br><tt>Orcad Capture GO 9.2</tt> <br><tt>Altera MAX+2 GO 10.0</tt> <br><tt>Altera Quartus2 GO 1.0</tt> <br><tt>Synplicify GO -</tt><tt></tt> <p><tt>-- Bob Elkind eteam@aracnet.com</tt> <br> <p>bob elkind wrote: <blockquote TYPE=CITE>Your friends and colleagues need you! <p>Time for an ad-hoc poll of the audience: <p>Please reply if you have *first-hand* experience (go or no-go) running design tool XXXX <br>[simulater | synthesiser | fpga-back-end-design | schematic capture | version control system | etc. ] <br>on a windows 2000 professional system. <p>Please reply (to the newsgroup) with : tool name, GO - NO-GO result, tool SW version. <p>Thanks for your help. <p>Bob Elkind, eteam@aracnet.com <br>fpga design/consulting</blockquote> </html> --------------DBF5A4F521D1EDFE3EF1B49F--Article: 32269
I've seen logic tossed in between my XORCY and the flop, but my primitives never got synthesized out. When I manually enter the carry chain, I still have my MUXCY and XORCY elements but sometimes other stuff gums up the works. In my case, a synchronous reset ended up as LUTs between the XORCY and the flop but - the really weird thing - I couldn't reproduce the problem for the synplify apps folks. The problem "went away." Maybe I missed a subtle coding change. Check to make sure the ins/outs are what you expect with your primitives and that certain primitives with valid outputs really do disappear; in all the hand-coding I've done to force the efficiencies, I haven't come across an "illegitimate" optimization of my primitives. Isn't coding in primitives such fun ?! - John Ray Andraka wrote: > Synplicity 6.2 is "optimizing" an instantiated design. In particular, I > > have a design with instantiated Xilinx primitives including a carry > chain. The synthesis is apparently flattening the xilinx primitives and > > doing its own optimization on them, which results in a multiplexer > placed between the xorcy and the flip-flop. When I instantiate > primitives, I don't expect to see them remapped. I didn't see this > happening in 5.3.1. The instantiated primitives have a syn_black_box > attribute on them. This new 'feature' makes Synplicity useless for the > > designs I am doing, and actually does damage for the large base I have > already fielded. > > Has anyone else seen this? Any fixes? > > -- > -Ray Andraka, P.E. > President, the Andraka Consulting Group, Inc. > 401/884-7930 Fax 401/884-7950 > email ray@andraka.com > http://www.andraka.comArticle: 32270
Hello, In Quartus II, you cannot back annotate the routing, only logic cell placement. The "Local Routing" option does not constrain the fitter to place the cells within a local route (in the same or adjacent LAB). If the placement is such that a local route can take place, the "Local Route" option will occur. If you have your placement where you want it and all local routes are used, you can then place that option on the cells to ensure that local routing is used on subsequent placement and compiles. In the latest release of Quartus II 1.1, you can constrain placement of a module and then replicate that exact placement on all instances of that module through LogicLock. With this new feature (contact your local Altera FAE for more details) and the local route option, you can constrain that path in every instance of the module. I hope this helps. If you have any questions, let me know. BrianArticle: 32272
We have a parametric FFT core available that can be configured for any=20= (power of 2) length. FFT.=20 The VHDL or Verilog code to be synthesized is auto-generated from our=20= Parametric System Builder tools so that the FFT can run as fast as you=20= need, only limited by available logic in the target device. How fast does your 16 point FFT need to operate? Regards,=20 Tom Dillon Dillon Engineering, Inc. http://www.dilloneng.com >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<< On 6/21/2001, 7:05:04 AM, finishf@yahoo.com (finish) wrote regarding FFT= =20 limited size input: > hello, > From my modest background, i know that for performing the FFT > transform on an input signal, i have to extend it, if required, by > zeros to 2^n. > FFT is a global transform,i.e the whole input sequence should be > available. > In practise, most often we take 1024 or 512, but i see some commercial= > hardware implementation for just 16 input data. > How far will this limited input size transform affect the overall > performance ? > thanks > H.SArticle: 32273
Hi, I'm just going to give my opinion here. I prefer VHDL for a couple reasons. First, Verilog does not allow variable indexing. If I have 8 instances of a RAM, I can easily create a 'generate' function and place them all with the variable for the generate as the index on the bit-slices for the port map. This is not allowed in Verilog (at least I don't know how to do it... if someone knows, please tell me). Second, Verilog has just (within the past year) allowed for double indexing on arrays. So the RAM arrays can now be ram[255:0][7:0]. VHDL has allowed this for years (actually you could pretty much always do it with user defined types). In the past, that same RAM in Verilog would have to be ram[2047:0] and you would have to keep the slices straight in your head. Combine this with no variable indexing and there could be difficulites creating functions parameterizable for future use. Now with that said, I agree with some of the other posts that there are many types in VHDL that are not in Verilog. These types make designing a little more lengthy, but debugging the code is easier because you will be flagged if you leave something out while Verilog will assume that you meant it to be tied to ground. Now with all that being said, I usually only write synthesizable code, not code for simulation, therefore Verilog may be better for simulation. In summary, Verilog is easier to learn, but VHDL is more powerful for the in depth user. Hope this helps. If you have any questions, let me know. BrianArticle: 32274
This is a multi-part message in MIME format. ------=_NextPart_000_0005_01C0FAAC.72FBD700 Content-Type: text/plain; charset="iso-8859-7" Content-Transfer-Encoding: quoted-printable Thank you all for your answers. I didnt know about the archives = (Google/dejanews). Now I have much material to work with. Nikiforakis Manos ------=_NextPart_000_0005_01C0FAAC.72FBD700 Content-Type: text/html; charset="iso-8859-7" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; = charset=3Diso-8859-7"> <META content=3D"MSHTML 5.50.4134.100" name=3DGENERATOR> <STYLE></STYLE> </HEAD> <BODY bgColor=3D#ffffff> <DIV><FONT face=3DArial size=3D2>Thank you all for your answers. I didnt = know about=20 the archives (Google/dejanews). Now I have much material to work=20 with.</FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2>Nikiforakis = Manos</FONT></DIV></BODY></HTML> ------=_NextPart_000_0005_01C0FAAC.72FBD700-- -- Posted from orion.ccf.auth.gr [155.207.199.31] via Mailgate.ORG Server - http://www.Mailgate.ORG
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