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
I am working on a product design that is tentatively planned to be based around a Spartan-II device with a small Micro-P core implemented in the FPGA as the system controller. I was wondering if anybody here could give me any advice and/or direct me to good information on purchased and/or open-source processor cores for Xilinx FPGAs, particularly the Spartan and Spartan-II devices. I haven't managed to learn much so far via the obvious links on the Xilinx web-site. Gaston BiessenerArticle: 30076
Luke, For BGA specifically, check out 3M Textool: http://www.3m.com/ehpd/sockets/ 3M Textool 6801 River Place Blvd. Austin, TX 78726-9000 (800) 328-0411 (612) 736-7167 There are more manufacturers listed in the databook under "Packages and Thermal Characteristics" http://www.xilinx.com/partinfo/pkgs_pdf/pkgs1.pdf subsection "Sockets". None of the other manufacturers besides 3M show that they supply a BG socket though. Also check out http://www.xilinx.com/support/programr/sockets.htm for a similar but more complete list. I hope this helps. Ryan Laity Xilinx Customer Apps Luke Roth wrote: > After hearing all the whiz-bang new features of the Virtex-II > series, we'd like to migrate our university project over (currently > running original Virtex). However, BGA packages are beyond our > capabilities. Does anybody have any recommendations for a > BG575-to-something-reasonable (PGA would be ideal) socket? > It's too bad Xilinx is making things more difficult for those of > us with limited resources. Nothing that we shouldn't be able to work > around, though. > Thanks, > LukeArticle: 30077
Does anyone know of any commercially available boards offered with the XCV812E on it? Looking for one for a fast prototype. -- -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: 30078
Alpha Data perhaps (the ADC-RC1000). Look at www.alphadata.co.uk. /Johan Ray Andraka wrote in message <3ABA5C75.68E00301@andraka.com>... >Does anyone know of any commercially available boards offered with the XCV812E >on it? Looking for one for a fast prototype. > > >-- >-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: 30079
Why isn't it safe? It makes sense to me. http://www.xilinx.com/xapp/xapp058.pdf Stan Ramsden wrote: > Here is the scenario: > I know this is not safe to do, but the board has been layed out and we are > stuck with what we have. I need to be able to program the XC18v04 > configuration memory thru the configured FPGA using the non dedicated JTAG > pins. > Has anyone done such a design?Article: 30080
Gaston, We have worked with a open source 6502 core. We spent some time=20 optimizing it for the Spartan family. You can find information about it at: http://www.free-ip.com/6502/ It synthesized without a problem using the Xilinx tools. If I remember=20= correctly without modifications it easily fit into a X2S50. There are some others out there. Also http://www.silicore.net/ has a low cost and very good PIC core. Let me know if you need any more information. Regards,=20 Tom Dillon Dillon Engineering, Inc. http://www.dilloneng.com >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<< On 3/22/2001, 1:20:47 PM, Gaston Biessener <gaston@yuck.net> wrote=20 regarding Looking for Processor Core info/advice: > I am working on a product design that is tentatively planned to be > based around a Spartan-II device with a small Micro-P core implemented= > in the FPGA as the system controller. > I was wondering if anybody here could give me any advice and/or direct= > me to good information on purchased and/or open-source processor cores= > for Xilinx FPGAs, particularly the Spartan and Spartan-II devices. I= > haven't managed to learn much so far via the obvious links on the Xili= nx > web-site. > Gaston BiessenerArticle: 30081
Austin Franklin wrote: > but something is broken, unless the > carry chain really does take 0ns through 48 levels... > Carry delay through 48 bits should be between 2 and 3 ns. Peter AlfkeArticle: 30082
hi, I have an application where I have to detect whether two clocks are within a certain frequency range with each other. IOW, I need to be able to detect whether the frequencies are within %2 of each other. The idea I am using is to run two counters, one on each clock and when one overflows, check the other to see how close it is. The decision is made based on the value of the other counter being close enough to zero or not. The other counter is then reset so that both start at zero again. I am running the decision logic on one of the clocks and the other counter is a gray counter. But of course there are a couple of issues with respect to metastability here. The value of one counter needs to cross one clock boundary to be compared against zero and the reset signal on overflow needs to cross a clock boundary in the other direction. I am having difficulty coming up with a robust design and asking for suggestions. thanks, MuzafferArticle: 30083
Hi can anyone recommend books for beginner level programming of XS40 and XS95 boards for someone with zero knowledge of FPGA VHDL programming? These are required to teach both staff and students, so it should preferably be easy to follow. Thanks very much. --Article: 30084
"Austin Lesea" <austin.lesea@xilinx.com> wrote in message news:3ABA26B6.520A1AC7@xilinx.com... > > With > 95% of all FPGA designs being done in HDL's, we think we are spending time and > silicon on what makes that flow the fastest and best possible. Well, picking an HDL: DataOut <= Reg0Val when Addr=REG0 else (others=>'z'); DataOut <= Reg1Val when Addr=REG1 else (others=>'z'); DataOut <= Reg2Val when Addr=REG2 else (others=>'z'); and the Verilog metaphor is similar. This is HDL, easy to understand, and a good mapping to TBUFs. Of course, ASICs do not usually support T/S lines. Even the Xilinx implementation is an emulation in Virtex, though the 3K and 4K had the real thing.Article: 30085
Tom Dillon <tdillon@dilloneng.com> writes: > We have worked with a open source 6502 core. We spent some time > optimizing it for the Spartan family. Can you make your Spartan optimized version available?Article: 30086
muzaffer@dspia.com wrote: > > hi, > I have an application where I have to detect whether two clocks are > within a certain frequency range with each other. IOW, I need to be > able to detect whether the frequencies are within %2 of each other. > The idea I am using is to run two counters, one on each clock and when > one overflows, check the other to see how close it is. The decision is > made based on the value of the other counter being close enough to > zero or not. The other counter is then reset so that both start at > zero again. I am running the decision logic on one of the clocks and > the other counter is a gray counter. But of course there are a couple > of issues with respect to metastability here. The value of one counter > needs to cross one clock boundary to be compared against zero and the > reset signal on overflow needs to cross a clock boundary in the other > direction. I am having difficulty coming up with a robust design and > asking for suggestions. > > thanks, > > Muzaffer Since you know the window, you do not need to pass the whole counters across the bondary. Generate a terminal count (eg 1024) and use to clear both counters to 00. then generate a window close to 512 on one ( 502..522 is +/- 2% ) ( one or 2 lines, 2 if you want dirn of error ) and use the other ctrs ==512 to sample these 2 lines. - just 3 lines cross clock domains. -jgArticle: 30087
This is probably best done like a frequency counter. Use one clock to generate a measurement interval, then count the cycles of the other clock that occur within that interval. This reduces the signals that cross the clock boundary to just one, and that can be synchronized with as many FFs as you want. Choose one clock you want on the output side,this is what the output is sync'd to, and is the clock you will count. Now, with the other clock, run a free running counter with a modulo that is convenient for the accuracy you wish to measure (this is your measurement interval). All you need out of that counter is a terminal count. Pass the terminal count across your clock domain boundary, synchronizing it to the other clock as you do. You want your synchronizer to produce a one clock wide pulse in the output domain in response to a TC from the input domain. More on that in a second. Use the one clock wide TC pulse to synchronously reset a second counter that is incremented on each cycle of the output side clock. If you want to measure the relative frequency, use the TC to clock-enable a register fed by the second counter outputs on the same cycle you apply the sync reset. In your case though, you are only interested in the frequency being in/out of range. For that, decode the counter outputs to generate an in-range/out-of-range signal, and then register that using a flip-flop clock enabled by TC. You can get this to whatever accuracy you desire by extending the time interval of the TC generated by the first counter. Only 1 signal crosses the clock domain boundary. What I do to generate a 1 cycle clock in the synchronized side uses 3 FFs. The first is a toggle FF clocked in the input side domain and toggled each time you get a trigger event (in this case a TC), so that it is high for one cycle of the first counter, low for the second. This goes to a 4 state state machine clocked in the second domain which generates a pulse each time the toggle input changes. If the state machine is coded properly, no sync flip-flop between is needed because the signal only affects one flip-flop. The state machine has 2 stable states corresponding to the two toggle states: case SM is=> when "00"=> if toggle='0' then SM<="00"; else SM<="01"; end if; when "01"=> SM<="11"; when "11"=> if toggle='1' then SM<="11"; else SM<="10"; end if; when "10"=> SM<="00"; when others=> null; end case; SM(0) is the one cycle wide pulse output. Note that that flip-flop is the only one affected by the toggle input. This synchronizer works as long as the toggle input cycles slower than half the clock of the state machine. Of course if you want, you can always add additional sync FF's on the toggle input. Jim Granville wrote: > > muzaffer@dspia.com wrote: > > > > hi, > > I have an application where I have to detect whether two clocks are > > within a certain frequency range with each other. IOW, I need to be > > able to detect whether the frequencies are within %2 of each other. > > The idea I am using is to run two counters, one on each clock and when > > one overflows, check the other to see how close it is. The decision is > > made based on the value of the other counter being close enough to > > zero or not. The other counter is then reset so that both start at > > zero again. I am running the decision logic on one of the clocks and > > the other counter is a gray counter. But of course there are a couple > > of issues with respect to metastability here. The value of one counter > > needs to cross one clock boundary to be compared against zero and the > > reset signal on overflow needs to cross a clock boundary in the other > > direction. I am having difficulty coming up with a robust design and > > asking for suggestions. > > > > thanks, > > > > Muzaffer > > Since you know the window, you do not need to pass the whole counters > across the bondary. > Generate a terminal count (eg 1024) and use to clear both counters to > 00. > then generate a window close to 512 on one ( 502..522 is +/- 2% ) > ( one or 2 lines, 2 if you want dirn of error ) > and use the other ctrs ==512 to sample these 2 lines. > - just 3 lines cross clock domains. > > -jg -- -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: 30088
Thanks. ANy others? Johan Ditmar wrote: > > Alpha Data perhaps (the ADC-RC1000). Look at www.alphadata.co.uk. > > /Johan > > Ray Andraka wrote in message <3ABA5C75.68E00301@andraka.com>... > >Does anyone know of any commercially available boards offered with the > XCV812E > >on it? Looking for one for a fast prototype. > > > > > >-- > >-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 -- -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: 30089
> > but something is broken, unless the > > carry chain really does take 0ns through 48 levels... > > > > Carry delay through 48 bits should be between 2 and 3 ns. > Peter Alfke > Er, then why does FPGA Editor say 0.0 for each of the carry elements?Article: 30090
"Simon Bacon" <simonb@tile.demon.co.cut_this.uk> wrote: > >"Austin Lesea" <austin.lesea@xilinx.com> wrote in message >news:3ABA26B6.520A1AC7@xilinx.com... >> >> With > 95% of all FPGA designs being done in HDL's, we think we are spending >time and >> silicon on what makes that flow the fastest and best possible. > >Well, picking an HDL: > > DataOut <= Reg0Val when Addr=REG0 else (others=>'z'); > DataOut <= Reg1Val when Addr=REG1 else (others=>'z'); > DataOut <= Reg2Val when Addr=REG2 else (others=>'z'); > >and the Verilog metaphor is similar. This is HDL, easy to >understand, and a good mapping to TBUFs. > >Of course, ASICs do not usually support T/S lines. By ASIC I take it you mean Standard cell based ASIC; if that's the case, this has not been my experience. Almost all SC libraries have T/S buffers and inverters. Some even have T/S DFFs and T/S bus holders. Of course with a full custom ASIC, there is no limit on what you can do. > Even the >Xilinx implementation is an emulation in Virtex, though the >3K and 4K had the real thing. I think the problem with T/S busses are well known. Difficult to time, potential for electrical conflicts, potential for very high power consumption etc. Being more of an analog feature, one has to spend extra effort with RTL verification. You might even have to spice some paths god forbid ;-) MuzafferArticle: 30091
Hi. I have a cpld (xilinx xc9500) in my system, mainly used for controlling buffers, and I suspect I will need to add pull up or pull down resistors where appropriate to prevent damage on system because of strange levels on I/O pins during jtag programming or power-up. Is it common practice to do like this, and if so, what are the appropriate values to add to force out pins to a safe level (to prevent bustranscievers from being turned on) during high-z ? / Daniel NilssonArticle: 30092
Mesner wrote: > Hi can anyone recommend books for beginner level programming of XS40 and XS95 > boards for someone with zero knowledge of FPGA VHDL programming? These are > required to teach both staff and students, so it should preferably be easy to > follow. You have several choices, none of which completely meet your needs: 1) "The Practical Xilinx Designer Lab Book 1.5" is a complete tutorial that shows how to design logic using VHDL and/or schematics for the XS40 and XS95 Boards and Xilinx Foundation 1.5. Prentice Hall discontinued publication of this book last year. You might get a used copy from somebody. But the current Xilinx Student Edition software is based on Foundation 2.1i so the book is a bit out of date. If you get a copy, perhaps Prentice Hall will give you permission to copy the material for use in your class. 2) XESS is releasing chapters of its "Pragmatic Logic Design with Foundation 2.1i" on http://www.xess.com. The chapters try to show detailed steps of how to use the various tools in Foundation 2.1i to create logic designs for the XS40 and XS95 Boards. However, while the examples may show you how to create, synthesize, implement, and download a VHDL-based design, it does not purport to instruct you in the details of VHDL coding. Also, only three of the eight chapters are currently available and the author habitually fails to meet the deadlines for releasing new chapters. 3) "Introductory VHDL" by Sudhakar Yalamanchili (ISBN 0-13-080982-9) gives a decent introduction to VHDL coding using the Xilinx Foundation 2.1i tools. It also includes the Student Edition of the Foundation 2.1i software. It doesn't discuss how to download or test any of its example designs on the XS40 or XS95 Boards. It costs $100 but it's not available yet. (It should be soon since I have a production copy on my desk.) Perhaps some combination of items 1, 2, and 3 will meet your needs. -- || Dr. Dave Van den Bout XESS Corp. (919) 387-0076 || || devb@xess.com 2608 Sweetgum Dr. (800) 549-9377 || || http://www.xess.com Apex, NC 27502 USA FAX:(919) 387-1302 ||Article: 30093
Austin Franklin wrote: > Er, then why does FPGA Editor say 0.0 for each of the carry elements? Maybe round-off, since each individual carry ( per bit) is <0.05 ns ?? Not a good reason... PeterArticle: 30094
Michal Kvasnicka wrote: > = > OK... > From my point of view delay measurement needs two delayed signal which = are > compared in the TDE (time delay estimation) algorithm, but TOA measurem= ent > work only with one precisely sampled signal and any available additiona= l > apriory knowledge (pulse shape, noise model, etc.). > = > Radar pulse can be approximated by trapezoidal (symmetric or asymmetric= ) > pulse wit the following parameters: > pulse width =3D 0.5 - 500us (50% amplitude level) > rise time =3D 20-100ns > fall time =3D 20-200ns > sample interval =3D 1 - 10ns > Pulse repetition interval =3D 1 - 5000us > = > These pulses (pulse train) is contaminated by noise of the general form= > (colored nongaussian, spread spectrum, etc.) with low S/N ration (in ma= ny > cases). Finally is represented by 1-channel data stream sampled by 1-10= ns > and with precise time stamping. > = > Time sampling is realized by Rubidium normal (short term stability abou= t > 10^-12 ) connected with GPS time receiver for long term stability about= > 10^-13 - 10^-15. > = > Required TOA accuracy is about 1-10ns. > = > So, I need effective and robust TOA algorithm which can be realized on = DSP > or FPGA chip sufficiently fast (see PRI value ~ 1-5000us). > = > Regards, > = > Michal > = > "Kolja Sulimma" <kolja@bnl.gov> p=EDse v diskusn=EDm pr=EDspevku > news:3AB65BBE.2055F00D@bnl.gov... > > Can you describe the rwquirements in more detail? > > Is it just delay measurement with high resolution? > > What resolution do you need? And how many channels? > > > > CU, > > Kolja Sulimma > > > > Michal Kvasnicka wrote: > > > > > Does anyone know of any texts or references concerning implementing= TOA > > > (time of arrival) measurement of the radar signal on FPGA or DSP ch= ips? > > > > > > Thanks, > > > > > > Michal > > I take it you have a passive multichannel receiver monitoring returns from a vehicle illuminated with multiple radars. You can use correlation techniques : each return stream will correlate with itself well enough to give you a probable time of return. By itself, a radar would not have that much variability pulse to pulse. Is your situation one where the radars transmit random seeming patterns as part of an identification/EW/multi-target scheme? In that case you are lost unless you have independent data as to the pattern transmitted. Differential GPS is now often used to do what I think you need. Each radar transmission encodes GPS data for a satellite. If the receiver can also see the satellite, it can then estimate the path delay. Having path delays to multiple radars solves the problem for a most probable location of the target. Use DGPS for the coarse measurement, then use doppler to estimate velocities relative to the radars and get fine measurements. Estimate position, velocity, use velocity to predict next position. use predicted - measured as position errors to feed back and correct the next position measurement. = -- = Russ Lyttle "World Domination through Penguin Power" The Universal Automotive Testset Project at <http://home.earthlink.net/~lyttlec>Article: 30095
Hello,
I'm currently targeting a Virtex-E for a project where I'm using a DLL
to de-skew a board level clock and another dll to route the same clock
internally. It's the same thing as the dll_mirror_1.v file from the
xapp132 because I get the same problem with that file:
-------
module dll_mirror_1 (CLKIN, CLKFB, CLK0_ext, CLK0_int);
input CLKIN, CLKFB;
output CLK0_ext, CLK0_int;
wire CLKIN_w, CLKFB_w, CLK0_int_dll, CLK0_ext_dll;
wire logic0;
assign logic0 = 1'b0;
IBUFG clkpad (.I(CLKIN), .O(CLKIN_w));
IBUFG clkfbpad (.I(CLKFB), .O(CLKFB_w));
CLKDLL dllint (.CLKIN(CLKIN_w), .CLKFB(CLK0_int), .RST(logic0),
.CLK0(CLK0_int_dll), .CLK90(), .CLK180(), .CLK270(),
.CLK2X(), .CLKDV(), .LOCKED());
CLKDLL dllext (.CLKIN(CLKIN_w), .CLKFB(CLKFB_w), .RST(logic0),
.CLK0(CLK0_ext_dll), .CLK90(), .CLK180(), .CLK270(),
.CLK2X(), .CLKDV(), .LOCKED());
BUFG clkg (.I(CLK0_int_dll), .O(CLK0_int));
OBUF clkextpad (.I(CLK0_ext_dll), .O(CLK0_ext));
endmodule
--------
The package that I'm targeting is a XCV2000E-7-FG680 and I'm using the
GCLK0 P and N pins for the clock input and feedback:
AW19 = Clock (CLKIN in the example)
AT21 = Clock feedback (CLKFB in the example).
I'm also using the following UCF file:
NET "CLKIN" LOC = "AW19";
NET "CLKFB" LOC = "AT21";
Synthesis runs fine (synplify 6.0 and synplify_pro with amplify 2.1.2).
Ngdbuild runs fine.
The problem is with the mapper. I get a segmentation fault every time I
run it:
>map -pr b -p xcv2000e-6-fg680 test
Release v3.3.07i - Map D.26
Copyright (c) 1995-2000 Xilinx, Inc. All rights reserved.
Using target part "v2000efg680-6".
Reading NGD file "test.ngd"...
Processing FMAPs...
Removing unused or disabled logic...
Running cover...
Writing file test.ngm...
Running directed packing...
Segmentation fault (core dumped)
Anyone has an idea about what could be the problem? Our Xilinx FAEs
told us that it should be possible to do it, they have no clue about why
it does not work and the datasheet & xapp132 does not mention that it's
not possible. They are working on the problem but I need a solution
ASAP!
BTW, no, we can't change the pinout because the board is ready. It
works if the feed-back is at the P input of another GCLK so it really
seems to be a problem with the tool.
Thanks!
Martin Guibert
Article: 30096Gaston Biessener wrote: > I am working on a product design that is tentatively planned to be > based around a Spartan-II device with a small Micro-P core implemented > in the FPGA as the system controller. Might checkout: http://www.fpgacpu.org/ http://tech-www.informatik.uni-hamburg.de/vhdl/vhdl.html -- Phil HaysArticle: 30097
As a note, the PAR report said there were 11 levels of logic MAX on this
clock...yet this delay report shows 24 levels...so how can I believe that in
the PAR report it actually DID take into account the carry chain?
Here is what FPGA Editor delay gives me (not all the 0s):
Path (0) "CLB_R25C39.S0.YQ" to "CLB_R14C41.S1.G2": contains 24 levels of
logic:
Path starting from Comp: CLB_R25C39.S0.YQ
To Delay type Delay(ns) Physical Resource
Logical Resource(s)
------------------------------------------------- --------
CLB_R24C29.S1.G4 net (fanout=2) 0.000R source_gain/acc_reg[1]
CLB_R24C29.S1.COUT Topcyg 0.000R
ce_gain/comp_acc_reg_cry_1/O
source_gain/acc_reg_i[1]
urce_gain/comp_acc_reg_cry_1
CLB_R23C29.S1.CIN net (fanout=1) 0.000R
ce_gain/comp_acc_reg_cry_1/O
CLB_R23C29.S1.COUT Tbyp 0.000R
ce_gain/comp_acc_reg_cry_3/O
urce_gain/comp_acc_reg_cry_2
urce_gain/comp_acc_reg_cry_3
CLB_R22C29.S1.CIN net (fanout=1) 0.000R
ce_gain/comp_acc_reg_cry_3/O
CLB_R22C29.S1.COUT Tbyp 0.000R
ce_gain/comp_acc_reg_cry_5/O
urce_gain/comp_acc_reg_cry_4
urce_gain/comp_acc_reg_cry_5
CLB_R21C29.S1.CIN net (fanout=1) 0.000R
ce_gain/comp_acc_reg_cry_5/O
CLB_R21C29.S1.COUT Tbyp 0.000R
ce_gain/comp_acc_reg_cry_7/O
urce_gain/comp_acc_reg_cry_6
urce_gain/comp_acc_reg_cry_7
CLB_R20C29.S1.CIN net (fanout=1) 0.000R
ce_gain/comp_acc_reg_cry_7/O
CLB_R20C29.S1.COUT Tbyp 0.000R
ce_gain/comp_acc_reg_cry_9/O
urce_gain/comp_acc_reg_cry_8
urce_gain/comp_acc_reg_cry_9
CLB_R19C29.S1.CIN net (fanout=1) 0.000R
ce_gain/comp_acc_reg_cry_9/O
CLB_R19C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_11/O
rce_gain/comp_acc_reg_cry_10
rce_gain/comp_acc_reg_cry_11
CLB_R18C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_11/O
CLB_R18C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_13/O
rce_gain/comp_acc_reg_cry_12
rce_gain/comp_acc_reg_cry_13
CLB_R17C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_13/O
CLB_R17C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_15/O
rce_gain/comp_acc_reg_cry_14
rce_gain/comp_acc_reg_cry_15
CLB_R16C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_15/O
CLB_R16C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_17/O
rce_gain/comp_acc_reg_cry_16
rce_gain/comp_acc_reg_cry_17
CLB_R15C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_17/O
CLB_R15C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_19/O
rce_gain/comp_acc_reg_cry_18
rce_gain/comp_acc_reg_cry_19
CLB_R14C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_19/O
CLB_R14C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_21/O
rce_gain/comp_acc_reg_cry_20
rce_gain/comp_acc_reg_cry_21
CLB_R13C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_21/O
CLB_R13C29.S1.COUT Tbyp 0.000R
e_gain/comp_acc_reg_cry_23/O
rce_gain/comp_acc_reg_cry_22
rce_gain/comp_acc_reg_cry_23
CLB_R12C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_23/O
CLB_R12C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[24]
rce_gain/comp_acc_reg_cry_24
rce_gain/comp_acc_reg_cry_25
CLB_R11C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_25/O
CLB_R11C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[26]
rce_gain/comp_acc_reg_cry_26
rce_gain/comp_acc_reg_cry_27
CLB_R10C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_27/O
CLB_R10C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[28]
rce_gain/comp_acc_reg_cry_28
rce_gain/comp_acc_reg_cry_29
CLB_R9C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_29/O
CLB_R9C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[30]
rce_gain/comp_acc_reg_cry_30
rce_gain/comp_acc_reg_cry_31
CLB_R8C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_31/O
CLB_R8C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[32]
rce_gain/comp_acc_reg_cry_32
rce_gain/comp_acc_reg_cry_33
CLB_R7C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_33/O
CLB_R7C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[34]
rce_gain/comp_acc_reg_cry_34
rce_gain/comp_acc_reg_cry_35
CLB_R6C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_35/O
CLB_R6C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[36]
rce_gain/comp_acc_reg_cry_36
rce_gain/comp_acc_reg_cry_37
CLB_R5C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_37/O
CLB_R5C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[38]
rce_gain/comp_acc_reg_cry_38
rce_gain/comp_acc_reg_cry_39
CLB_R4C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_39/O
CLB_R4C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[40]
rce_gain/comp_acc_reg_cry_40
rce_gain/comp_acc_reg_cry_41
CLB_R3C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_41/O
CLB_R3C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[42]
rce_gain/comp_acc_reg_cry_42
rce_gain/comp_acc_reg_cry_43
CLB_R2C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_43/O
CLB_R2C29.S1.COUT Tbyp 0.000R
source_gain/comp_acc_reg[44]
rce_gain/comp_acc_reg_cry_44
rce_gain/comp_acc_reg_cry_45
CLB_R1C29.S1.CIN net (fanout=1) 0.000R
e_gain/comp_acc_reg_cry_45/O
CLB_R1C29.S1.Y Tciny 0.000R
source_gain/comp_acc_reg[46]
rce_gain/comp_acc_reg_cry_46
ource_gain/comp_acc_reg_s_47
CLB_R14C41.S1.G2 net (fanout=1) 0.000R
source_gain/comp_acc_reg[47]
-------------------------------------------------
Total (0.000ns logic, 0.000ns route) 0.000ns (to I_clk_50m_c)
(0.0% logic, 0.0% route)
I will say that FPGA editor is so much more difficult to use than the old
XACT editor for things like path delay to name one. I know it's got a lot
of improvements, but it is a giant slide backwards in certain functionality
that I find very useful.
"Peter Alfke" <palfke@earthlink.net> wrote in message
news:3ABABB88.9E594630@earthlink.net...
>
>
> Austin Franklin wrote:
>
> > Er, then why does FPGA Editor say 0.0 for each of the carry elements?
>
> Maybe round-off, since each individual carry ( per bit) is <0.05 ns ??
> Not a good reason...
>
> Peter
>
>
Article: 30098Does anyone know off-hand the licensing for these cores? Can we use a free 6502 core in a commercial product? Ditto for PIC- compatabile cores. I know that the licensing on Jan gray's web site (www.fpgacpu.org) specifies no commercial uses. Some of these other cores might be free but the instuction sets copyrighted? (I'd like to put together a instructoin-compatable PIC core, but I can't figure out if I'd be allowed to use it. ) Anyone know? -Kent > Gaston Biessener wrote: > > I am working on a product design that is tentatively planned to be > > based around a Spartan-II device with a small Micro-P core implemented > > in the FPGA as the system controller. Phil Hays <spampostmaster@home.com> writes: > Might checkout: > > http://www.fpgacpu.org/ > > http://tech-www.informatik.uni-hamburg.de/vhdl/vhdl.html > > > -- > Phil HaysArticle: 30099
Also as opposed to "Dead Meat" or "Red Meat" from Cali. -Lee - Former Jersey Punk myself. "Andy Peters noao [.] edu>" <"apeters <"@> wrote in message news:99bdm9$2gi6$1@noao.edu... > Lee Iovino wrote: > > > > "Aggressive Meat" sounds like the name of a punk band from New Jersey. > > No, they were from across the river in Brooklyn. You're thinking of the > Meatmen, who were actually from Chicago. > > -andy > former NJ punk rocker.
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