FFT-COMP-EP-U1

FFT Compiler IP Core User’s Guide August 2011 IPUG54_01.9 Table of Contents Chapter 1. Introduction .......................................................................................................................... 4 Quick Facts ........................................................................................................................................................... 4 Features ................................................................................................................................................................ 6 Chapter 2. Functional Description ........................................................................................................ 8 Block Diagram....................................................................................................................................................... 8 High Performance Architecture .................................................................................................................... 9 Low Resource Architecture .......................................................................................................................... 9 Configuring the FFT Compiler............................................................................................................................. 10 Number of Points ....................................................................................................................................... 10 Architecture ................................................................................................................................................ 10 Output Order .............................................................................................................................................. 10 Scaling Mode ............................................................................................................................................. 11 Precision Reduction Method ...................................................................................................................... 11 Signal Descriptions ............................................................................................................................................. 11 Interfacing with the FFT Compiler ....................................................................................................................... 12 Configuration Signals ................................................................................................................................. 12 Handshake Signals .................................................................................................................................... 13 Exponent Output ........................................................................................................................................ 13 Exceptions.................................................................................................................................................. 13 Timing Specifications .......................................................................................................................................... 13 Chapter 3. Parameter Settings ............................................................................................................ 16 Points/Mode Tab ................................................................................................................................................. 17 Number of Points ....................................................................................................................................... 17 Architecture ................................................................................................................................................ 17 FFT Mode................................................................................................................................................... 17 Output Order .............................................................................................................................................. 18 Scaling Width Tab ............................................................................................................................................... 18 Scaling Mode ............................................................................................................................................. 18 Data Width ................................................................................................................................................. 18 Precision Reduction Method ...................................................................................................................... 19 Implementation Tab ............................................................................................................................................ 19 Multiplier Type............................................................................................................................................ 19 Pipeline ...................................................................................................................................................... 19 Memory Type ............................................................................................................................................. 19 Synthesis and Simulation Tools Tab................................................................................................................... 20 Support Synplify ......................................................................................................................................... 20 Support Precision....................................................................................................................................... 20 Support ModelSim...................................................................................................................................... 20 Support ALDEC.......................................................................................................................................... 20 Chapter 4. IP Core Generation............................................................................................................. 21 Licensing the IP Core.......................................................................................................................................... 21 Getting Started .................................................................................................................................................... 21 IPexpress-Created Files and Top Level Directory Structure............................................................................... 24 Instantiating the Core .......................................................................................................................................... 25 Running Functional Simulation ........................................................................................................................... 25 Synthesizing and Implementing the Core in a Top-Level Design ....................................................................... 25 Hardware Evaluation........................................................................................................................................... 26 Enabling Hardware Evaluation in Diamond................................................................................................ 26 Enabling Hardware Evaluation in ispLEVER.............................................................................................. 26 © 2011 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice. IPUG54_01.9, August 2011 2 FFT Compiler IP Core User’s Guide Lattice Semiconductor Table of Contents Updating/Regenerating the IP Core .................................................................................................................... 27 Regenerating an IP Core in Diamond ........................................................................................................ 27 Regenerating an IP Core in ispLEVER ...................................................................................................... 27 Chapter 5. Support Resources ............................................................................................................ 29 Lattice Technical Support.................................................................................................................................... 29 Online Forums............................................................................................................................................ 29 Telephone Support Hotline ........................................................................................................................ 29 E-mail Support ........................................................................................................................................... 29 Local Support ............................................................................................................................................. 29 Internet ....................................................................................................................................................... 29 References.......................................................................................................................................................... 29 LatticeECP ................................................................................................................................................. 29 LatticeECP2/M ........................................................................................................................................... 29 LatticeECP3 ............................................................................................................................................... 29 LatticeXP2.................................................................................................................................................. 30 Revision History .................................................................................................................................................. 30 Appendix A. Resource Utilization ....................................................................................................... 31 LatticeECP Devices ............................................................................................................................................ 31 Ordering Part Number................................................................................................................................ 31 LatticeECP2 Devices .......................................................................................................................................... 31 Ordering Part Number................................................................................................................................ 32 LatticeECP2M Devices ....................................................................................................................................... 32 Ordering Part Number................................................................................................................................ 32 LatticeECP3 Devices .......................................................................................................................................... 32 Ordering Part Number................................................................................................................................ 32 LatticeXP2 Devices ............................................................................................................................................. 33 Ordering Part Number................................................................................................................................ 33 IPUG54_01.9, August 2011 3 FFT Compiler IP Core User’s Guide Chapter 1: Introduction Lattice's Fast Fourier transform (FFT) Compiler intellectual property (IP) core offers forward and inverse Fast Fourier Transforms for point sizes from 64 to 16384. The FFT Compiler IP core can be configured to perform forward FFT, inverse FFT (IFFT) or port selectable forward/inverse FFT. The FFT Compiler IP core offers two choices of implementation: high performance (Streaming I/O) and low resource (Burst I/O). In the high performance implementation, the FFT Compiler IP core can perform real-time computations with continuous data streaming in and out at clock rate. There can also be arbitrary gaps between data blocks allowing discontinuous data blocks. The low resource implementation can be used when it is required to use less slice (logic unit of Lattice FPGA devices) and Embedded Block RAM (EBR) and Digital Signal Processor (DSP) resources or if the device is too small to accommodate the high performance version. To account for the data growth in fine register length implementations, the FFT Compiler IP core allows several different modes (fixed and dynamic) for scaling data after each radix-2 stage of the FFT computation. The low resource version also supports block floating point arithmetic that provides increased dynamic range for intermediate computations. The FFT Compiler IP core also allows the number of FFT points to be varied dynamically through a port. Quick Facts Table 1-1 through Table 1-5 give quick facts about the FFT Compiler IP core for LatticeECP™, LattceECP2™, LatticeECP2M™, LatticeECP3™, and LatticeXP2™ devices, respectively. Table 1-1. FFT Compiler IP Core for LatticeECP Devices Quick Facts FFT IP Configuration High performance 256 points Core Requirements LatticeECP Minimal Device Needed LFECP6E Data Path Width LUTs sysMEM EBRs Registers MULT18X18ADDSUB 16 16 16 16 2100 2700 800 1000 3 5 3 5 1700 2100 800 900 6 8 2 ® 2 ® Lattice Diamond 1.0 or ispLEVER 8.1 Synthesis Synopsys Synplify Pro for Lattice D-2009.12L-1 Aldec Active-HDL 8.2 Lattice Edition Simulation IPUG54_01.9, August 2011 Low performance 256 points LFECP33E-5F672C Lattice Implementation Design Tool Support Low performance 256 points FPGA Families Supported Targeted Device Resource Utilization High performance 1024 points Mentor Graphics ModelSim SE 6.3F 4 FFT Compiler IP Core User’s Guide Lattice Semiconductor Introduction Table 1-2. FFT Compiler IP Core for LatticeECP 2Devices Quick Facts FFT IP Configuration Core Requirements High performance 256 points High performance 1024 points LFE2-6E LFE2-12E FPGA Families Supported Minimal Device Needed Resource Utilization LUTs sysMEM EBRs Registers MULT18X18ADDSUB LFE2-6E LFE2-6E LFE2-50E-7F672C 16 16 16 16 2100 2700 800 900 3 6 3 3 1700 2100 800 800 6 8 2 2 Diamond 1.0 or ispLEVER 8.1 Lattice Implementation Design Tool Support Low performance 256 points Lattice ECP2 Targeted Device Data Path Width Low performance 256 points Synthesis Synopsys Synplify Pro for Lattice D-2009.12L-1 Aldec Active-HDL 8.2 Lattice Edition Simulation Mentor Graphics ModelSim SE 6.3F Table 1-3. FFT Compiler IP Core for LatticeECP2M Devices Quick Facts FFT IP Configuration High performance 256 points Core Requirements FPGA Families Supported Data Path Width LUTs sysMEM EBRs Registers MULT18X18ADDSUB LFE2M35E-7F672C 16 16 16 16 2100 2700 900 1000 3 6 3 3 1700 2100 800 900 6 8 2 2 Diamond 1.0 or ispLEVER 8.1 Synthesis Synopsys Synplify Pro for Lattice D-2009.12L-1 Aldec Active-HDL 8.2 Lattice Edition Simulation IPUG54_01.9, August 2011 Low performance 256 points LFE2M20E Lattice Implementation Design Tool Support Low performance 256 points Lattice ECP2M Minimal Device Needed Targeted Device Resource Utilization High performance 1024 points Mentor Graphics ModelSim SE 6.3F 5 FFT Compiler IP Core User’s Guide Lattice Semiconductor Introduction Table 1-4. FFT Compiler IP Core for LatticeECP3 Devices Quick Facts FFT IP Configuration High performance 256 points Core Requirements FPGA Families Supported Minimal Device Needed Data Path Width LUTs sysMEM EBRs Registers MULT18X18C LFE3-35EA LFE3-95E-8FN672CES 16 16 16 16 2100 2700 900 1000 3 6 3 3 1700 2100 800 900 16 4 4 12 Diamond 1.0 or ispLEVER 8.1 Lattice Implementation Design Tool Support Low performance 256 points Low performance 256 points Lattice ECP3 Targeted Device Resource Utilization High performance 1024 points Synthesis Synopsys Synplify Pro for Lattice D-2009.12L-1 Aldec Active-HDL 8.2 Lattice Edition Simulation Mentor Graphics ModelSim SE 6.3F Table 1-5. FFT Compiler IP Core for LatticeXP2 Devices Quick Facts FFT IP Configuration Core Requirements High performance 256 points High performance 1024 points LFXP2-5E LFXP2-8E FPGA Families Supported Minimal Device Needed Resource Utilization LUTs sysMEM EBRs Registers MULT18X18ADDSUB LFXP2-5E LFXP2-5E LFXP2-17E-7F484CES 16 16 16 16 2100 2700 900 1000 3 6 3 3 1700 2100 800 900 6 8 2 2 Diamond 1.0 or ispLEVER 8.1 Lattice Implementation Design Tool Support Low performance 256 points Lattice XP2 Targeted Device Data Path Width Low performance 256 points Synthesis Synopsys Synplify Pro for Lattice D-2009.12L-1 Aldec Active-HDL 8.2 Lattice Edition Simulation Mentor Graphics ModelSim SE 6.3F Features • Wide range of points sizes: 64, 128, 256, 512, 1024, 2048, 4096, 8192, and 16384 • Choice of high-performance (streaming I/O) or low resource (burst I/O) versions • Run-time variable FFT point size • Forward, inverse or port-configurable forward/inverse transform modes • Choice of no scaling, fixed scaling (RS111/RS211) or dynamically variable stage-wise scaling IPUG54_01.9, August 2011 6 FFT Compiler IP Core User’s Guide Lattice Semiconductor Introduction • Data precision of 8 to 24 bits • Twiddle factor precision of 8 to 24 bits • Natural order for input and choice of bit-reversed or natural order for output • Support for arbitrary gaps between input data blocks in high-performance realization • Block floating point scaling support in low resource configurations IPUG54_01.9, August 2011 7 FFT Compiler IP Core User’s Guide Chapter 2: Functional Description This chapter provides a functional description of the Lattice FFT Compiler IP core. Block Diagram Figure 2-1 shows the interface diagram for the FFT compiler. The diagram shows all of the available ports for the IP. It should be noted that not all the I/O ports are available for a chosen configuration. Figure 2-1. FFT Compiler Interface Diagram clk rstn exponent except mode modeset sfact sfactset points pointset rfib ibend FFT Compiler obstart outvalid dore ibstart dire doim diim FFT is a fast algorithm to implement the following N point Discrete Fourier Transform (DFT) function. N-1 nk X(k) = (1) x(n)WN n=0 where WN is given by: -j WN = e 2π N (2) The inverse DFT is given by: N-1 -nk 1 X(n) = N X(k)WN (3) k=0 However the output of the FFT Compiler IP core will differ from the true output by a scale factor determined by the scaling scheme. If “right-shift by 1 at all stages” scaling mode (RS111) is used, there is a division by 2 at every stage resulting in an output that is 1/Nth of the true output of Equation (1). The output for inverse FFT will match with Equation (3) for this scaling mode. Using other scaling modes will result in outputs scaled by other appropriate scale factors. The FFT Compiler IP corer offers two different implementation modes - High Performance and Low Resource. In High Performance mode, the FFT Compiler IP core can continuously read in and give out one data sample per clock, block after block. The FFT throughput is equal to the clock rate when the data blocks are applied continuously. Low Resource mode uses less logic and memory resources, but requires 4 to 8 block periods to compute the IPUG54_01.9, August 2011 8 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description FFT for one block of input data. Both versions of FFT do not allow breaks in the data stream within a block but do allow arbitrary additional gaps between data blocks. High Performance Architecture The implementation diagram for the high performance FFT is shown in Figure 2-2. Figure 2-2. Implementation Diagram for High-Performance FFT bfly control Scaling din bfly mem Scaling mem mem Input Registers and Control PE(log2N-1) PE1 control bfly Scaling PE0 Output Processing, Bit-Reversal & Control dout control The high performance FFT implementation consists of several processing elements (PEs) connected in cascade. The number of PEs is equal to log2 N, where N is the number of FFT points. Each PE has a radix-2 decimation-infrequency (DIF) butterfly (bfly), a memory (mem), an address generation and control logic (control), and a scaling unit (scaling). Some of the butterflies include a twiddle multiplier and a twiddle factor memory. The scaling unit performs a division by 2, a division by 4, or no division, depending on the scaling mode and scale factor inputs at the port. There is an input-processing block at the beginning of the PE chain and an output-processing block at the end of the PE chain. The input processing block has registers and control logic for handshake signals and dynamic mode control. The output-processing block contains handshake, mode control and bit-reversal logic, if configured for natural ordered output. The high performance FFT implementation enables streaming I/O operation, where the data is processed at clock speed without any gaps between blocks. This implementation can also be employed for burst I/O situations by using the handshake signals. Low Resource Architecture The low resource implementation employs only one physical radix-2 butterfly and reuses the same butterfly over multiple time periods to perform all stages of the FFT computation. Hence the resource requirement (EBR and slices) is lower compared to the high-performance version. Depending on the number of points, an N-point FFT computation may require 4N to 8N cycles. The implementation diagram for the low-resource FFT is shown in Figure 2-3. IPUG54_01.9, August 2011 9 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Figure 2-3. Low-resource FFT Data Flow Diagram Twiddle ROM mem a Butterfly mem b Commutator Input Data Output Data As Figure 2-3 shows, the FFT module is built with a butterfly reading from and writing to two memories at the same time. There is a commutator after the butterfly to handle the writing sequence of the intermediate outputs. The twiddle memory contains the pre-computed twiddle factors for the FFT. When an input block is applied, the first half of the block is written into memory a and the second half into memory b in a bit-reversed order. The butterfly reads from the two memories, performs stage 0 computation and writes out the intermediate results to the same sites in each memory. Again, for stage 1 computation, the butterfly reads from the two memories, performs computation and writes back into the two memories through the commutator. A similar process of reading, computing and writing continues for each of the remaining stages. For every block of input data read, four to eight blocks of computation time is required for this scheme. Due to the twin memory architecture, when data is unloaded form FFT in bitreversed mode, the data in memory a (points 0 to N/2-1) will be unloaded first, followed by the data in memory b (N/2 to N-1), both in a bit-reversed order. Configuring the FFT Compiler Number of Points The number of points can be either fixed and specified in the GUI or specified through the points input port. If the FFT core is going to be used for fixed points, it is better to configure it that way, to avoid the additional resource utilization in dynamic points configurations. In dynamic points configurations, the hardware resources used correspond to the maximum number of points. So to minimize resource utilization, the minimum and maximum points must be set to the minimum and maximum values expected in the usage. Architecture The operation mode must be chosen to suit the throughput requirements and resource availability. If block floating point scaling is required, then the Low Resource architecture must be selected. Output Order Output data can be chosen either to be in bit-reversed order or natural order. Natural order output is preferred in many applications, where the output is directly fed to the following stage. But in applications where a FIFO or a buffer is used between the output of FFT and the input to the next processing block, as in multi-clock systems, bitreversal can be done in that glue memory. Selecting bit-reversed output results in reduced latency and/or lesser EBR usage. In low resource modes, the bit-reversed order is applicable separately to the lower and upper half of the output. For a N point FFT, the first N/2 points are available in bit-reversed order first, followed by the second N/2 points in a bit-reversed order. IPUG54_01.9, August 2011 10 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Scaling Mode The selection of scaling mode depends on the statistical properties of the data and the trade-off between increased dynamic range and occasional overflows. If a large dynamic range is required for the internal computations and if the low-resource implementation is used, block floating point will be the best choice. If memory and logic resource utilization (and possibly throughput reduction) is not a problem, full precision arithmetic (scaling selected as “None”) can be used. The data width grows by one bit every stage requiring bigger multipliers, wider data path logic and wider memories. The output width is equal to input width + log2N. Among the fixed scale factors, RS111 is a more uniform scaling method than RS211. If the input data magnitude is known to be higher on the average, the RS211 method may be employed. Dynamic scale mode is for more complex scenarios where the statistical properties of the input differ between data blocks and the user wants to use more than one scaling scheme. Precision Reduction Method Precision reduction method selection applies to the scaling process after every stage. When scaling is employed (a divide by 2 or 4), the scaled data is obtained either by truncating the data (discarding last one or two bits) or rounding the data to the nearest number in the scaled precision (discarding one or two bits and making correction to the output data based on discarded bits). Truncation results in less logic utilization while rounding improves the accuracy of results. Signal Descriptions The top-level interface diagram for the FFT Compiler is shown in Figure 2-1 and the details of the I/O ports are summarized in Table 2-1. Table 2-1. Top level I/O interface Port Bits I/O Description Clock and Reset clk 1 I System clock. rstn 1 I System wide asynchronous active-low reset signal. I Data in real. Real part of the input data. Data Input and Output dire 8 - 24 diim 8 - 24 I Data in imaginary. Imaginary part of the input data. dore 8 - 24 O Data out real. Real part of the output data. doim exponent 8 - 24 O Data out imaginary. Imaginary part of the output data. Log2(N)+1 O Exponent output for block floating point. This value denotes the effective scaling that was done during block floating scaling. I FFT mode signal. Programs core to perform forward or inverse FFT. Configuration Signals mode 1 0 - Forward FFT 1 - Inverse FFT The value at mode is loaded into the system whenever modeset input goes high. The changes are effective from the start of the next input data block, i.e., for an ibstart going high during or after modeset. modeset 1 IPUG54_01.9, August 2011 I Set FFT mode signal. When this input signal goes high, the value at mode port is read and the FFT mode (forward or inverse FFT) is set. 11 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Table 2-1. Top level I/O interface (Continued) Port sfact Bits I/O Description 12 - 28 I Stage-wise scaling factors. This signal is a concatenation of individual 2-bit stage scaling factors. The most significant 2 bits correspond to stage 1 scale factor, the next significant 2 bits to stage 2 scale factor and so on. When the number of point is not a power of 4, the last stage has only 1-bit scaling factor. Each scaling factor denotes the number of right shifts performed to that stage’s output data. The scale factors are loaded into the system when sfactset input goes high. The changes are effective from the start of the next input data block, i.e., for an ibstart going high during or after sfactset. sfactset points 1 I Set scale factor signal. When this signal goes high, stage-wise scaling factors are set with the values at sfact input port. 3 if maximum points = 128 I Number of FFT points. This input is used to specify the number of points in the dynamic points mode. The value at this port must be equal to the log2 of the number of points represented in unsigned binary form. The valid range of values is from 6 to 14. A value less than 6 will be read as 6 and a value greater than 14 will be read as 14. I Set number of points signal. When this input signal goes high, the value at points port is read and the number of FFT points is set accordingly. The new number of points is effective from the next block of data, i.e., for a valid ibstart applied after pointset going high. For low resource mode, pointset can be applied during or before ibstart. For high performance mode, pointset must be applied five cycles before ibstart. 4, otherwise pointset 1 Status and Handshake Signals ibstart 1 I Input block start. Asserted high by the user to identify the start of an input data block. Once this signal goes high for a cycle, the core enters an “input read cycle”, during which input data is read in N consecutive cycles (N is the number of FFT points). Any ibstart signal during an input read cycle is ignored. except 1 O Exception. Denotes that an exception (overflow) has occurred in the computation. This could be due to the use of a wrong set of scaling factors. The exception always corresponds to a problem with the data that is currently output and not with a problem with the data that is being processed. rfib 1 O Ready for input block. This signal indicates that the core is ready to receive the next block of input data. The driving system can assert ibstart one cycle after rfib goes high. After ibstart goes high, the core will pull rfib low in the next clock cycle. ibend 1 O Input block end. This signal goes high for one cycle to coincide with the last sample of the current input data block that is being read through input ports. obstart 1 O Output block start. This signal is asserted high by the core for one clock cycle, to signify the start of an output block of data. outvalid 1 O Output data valid. This signal indicates that the core is now giving out valid output data through dore and doim ports. Interfacing with the FFT Compiler Configuration Signals There are three dynamic configuration signals used in the FFT compiler: mode, sfact and points. The user independently selects each of these. The configuration signals are sampled and stored when the corresponding set signals are active. Specifically mode is set when modeset is active, sfact is set when sfactset is active, and points is set when pointset becomes active. However these values must be set during or before the start of a block for them to be effective for that block. In other words, the set signals must be active at or before ibstart going active, for them to be effective for that block. There are a couple of exceptions to this rule. In low-resource implementation and in bit-reversed output mode, the set signals are ignored when outvalid is high. Refer to IPUG54_01.9, August 2011 12 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Figure 2-6 for an illustration of configuration signal timing. In high performance implementation, the pointset must be applied five cycles before ibstart for it to be effective for the next block. Handshake Signals The input ibstart is used to specify the start of a data block and it is assumed to coincide with first data point in the input block. This signal also sets the configuration of the core based on the values set by the corresponding set signals. Once ibstart is valid, the core starts reading the input in consecutive clocks without gap until all the Npoints are read. When the last data in a block is being read from input, the output ibend is asserted high by the core. The output control signal rfib indicates that the core is ready for a new input block. One cycle after ibstart, the output rfib goes low and it remains low until one cycle before the next block can be applied. The external driving system can check for rfib and start an input block in the next cycle after rfib goes high. Refer to Figure 2-6 for an illustration of configuration signal timing. Exponent Output The output port exponent gives the value of the exponent of the multiplicative factor for the output to get the true FFT output. The value of exponent is an unsigned number. The true (i)FFT output is given by: True (i)FFT output = (dore * 2^exponent) + j (doim * 2^exponent) Exceptions Exceptions occur if there is an internal overflow in the computation of an output block. An exception is notified by the except signal going high during a valid output. The except signal goes high if one or more overflows occur during the computation of a block. The severity and number of overflow exceptions in a block depends on the scaling scheme used and the property of the input data. If the user is using an appropriate scaling method for the expected input data and can tolerate occasional exceptions, the except output may be left unconnected leading to a slightly reduced resource utilization. Timing Specifications The top-level timing diagrams for several cases are given in Figure 2-4 through Figure 2-7. Figure 2-4. Timing Diagram for Streaming I/O with Continuous Data Blocks (64 Points) clk rfib ibstart dire, diim x1(0) x1(1) x1(2) ... x1(62) X1(2) ... x1(63) x2(0) x2(1) x2(2) x2(3) X1(62) X1(63) X2(0) X2(1) ibend dore, doim output latency X1(0) X1(1) obstart outvalid IPUG54_01.9, August 2011 13 FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Figure 2-5. Timing Diagram for Streaming I/O with Gaps Between Data Blocks (64 Points) clk rfib ibstart dire, diim x1(0) x1(1) ... x1(62) X1(2) ... x1(2) x1(63) x2(0) x2(1) x2(2) ibend output latency dore, doim X1(0) X1(1) X1(62) X1(63) X2(0) obstart outvalid Figure 2-6. Timing Diagram Showing Handshake Signals for Low Resource FFT clk rfib mode x mode1 x x sfact1 x x points1 x x mode2 modeset sfact x sfact2 x sfactset points x points2 x pointset ibstart dire, diim x1(0) x1(1) ... x1(63) x2(0)x (1) 2 x2(2) x2(3) ibend This block uses: mode1, sfact1 and points1 IPUG54_01.9, August 2011 14 This block uses: mode2, sfact2 and points1, as only modeset and sfactset occur at or before ibstart FFT Compiler IP Core User’s Guide Lattice Semiconductor Functional Description Figure 2-7. Timing Diagram Showing Handshake Signals for High Performance FFT clk rfib mode x mode1 x x sfact1 x mode2 x modeset sfact sfact2 x x sfactset points x points1 x x points2 pointset ibstart dire, diim x1(0) x1(1) ... x1(63) x2(0) x2(1) ibend This block uses: mode1, sfact1 and points1 This block uses: mode2, sfact2 and points1, as pointset does not arrive 5 cycles before ibstart Output Latency Table 2-2 gives the latency through the IP core as a function of FFT point size and implementation mode. Table 2-2. Latency through FFT IP Core (Clock Cycles) FFT Point Size Low Resource Mode 64 278 High Performance Mode 83 128 598 152 256 1302 282 512 2838 543 1024 6166 1057 2048 13334 2086 4096 28694 4136 8192 61462 8237 16384 131094 16431 IPUG54_01.9, August 2011 15 FFT Compiler IP Core User’s Guide Chapter 3: Parameter Settings The IPexpress™ tool is used to create IP and architectural modules in the Diamond and ispLEVER software. Refer to “IP Core Generation” on page 21 for a description on how to generate the IP. Table 3-1 provides the list of user configurable parameters for the FFT Compiler IP core. The parameter settings are specified using the FFT Compiler IP core Configuration GUI in IPexpress. Table 3-1. FFT Compiler Parameter Description Name Range Default Number of Points Points variability Fixed, Variable Fixed 64, 128, 256, 512, 1024, 2048, 4096, 8192, 16384 64 Minimum points 64, 128, 256, 512, 1024, 2048, 4096, 8192 64 Maximum points {128, 256, 512, 1024, 2048, 4096, 8192, 16384} only values > Min points are valid 128 High Performance, Low Resource Low Resource Forward, Inverse, Dynamic through Port Forward Bit-reversed, Natural Bit-reversed None, RS111, RS211, Dynamic Through Port, Block Floating Point RS111 Input Data Width 8 to 24 16 Twiddle Factor Width 8 to 24 16 Truncation, Rounding Truncation Multiplier Type DSP block based, LUT based DSP block based Adder Pipeline 0,1 0 Number of Points (N) Architecture Architecture FFT Mode FFT Mode Output Order Output order Scaling Mode Scaling Mode Data Widths Precision Reduction Method Precision Reduction Implementation Multiplier Pipeline 2,3,4 3 EBR Memory, Distributed Memory, Automatic EBR Memory Synthesis Tools Support Synplify Support Precision RTL Sythesis Both are selected Simulation Tools Support ModelSim Support Active-HDL Both are selected Memory Type Synthesis & Simulation Tools IPUG54_01.9, August 2011 16 FFT Compiler IP Core User’s Guide Lattice Semiconductor Parameter Settings Points/Mode Tab Figure 3-1 shows the contents of the Points/Mode tab. Figure 3-1. Points/Mode Tab Number of Points This parameter allows the user to specify fixed or variable number of points. Number of Points This parameter specifies the number of FFT points if points variability is “FIxed.” Maximum Points This parameter denotes the maximum for the points range if points variability is “Variable.” Minimum Points This parameter denotes the minimum for the points range if points variability is “Variable.” Architecture This option selects either High-Performance (Streaming I/O) or Low Resource (Burst I/O) architecture. The high performance implementation offers high throughput and allows streaming input and output data, with optional gaps between blocks. The low resource implementation results in low memory and logic resource utilization, but takes multiple block periods (typically between 4 to 8) of computation time to process each data block. FFT Mode This parameter configures operating mode of the core to forward FFT, inverse FFT or dynamically variable forward/inverse FFT. In the “Dynamic Through Port” mode, the FFT mode can be set to forward or inverse FFT using the input ports mode and modeset. IPUG54_01.9, August 2011 17 FFT Compiler IP Core User’s Guide Lattice Semiconductor Parameter Settings Output Order This parameter specifies whether the output data is in bit-reversed or natural order. Scaling Width Tab Figure 3-1 shows the contents of the Scaling Width tab. Figure 3-2. Scaling Width Tab Scaling Mode This parameter defines whether the data is scaled or not after each radix-2 butterfly and if so, what kind of scaling is used. The value can be None, RS111, RS211, Dynamic Through Port and Block Floating Point. • If the value is “None” there is no scaling at the output of butterflies. • The option RS111 results in a fixed scaling of “right shift by 1” in all FFT stages. • The option RS211 results in a fixed scaling of “right shift by 2” in the first stage and “right shift by 1” in the subsequent stages. • If Scale Mode is set to “Dynamic Through Port”, the scale factors for the FFT stages are read dynamically from the input port sfact for every data block. In the dynamic scaling mode, for high performance architecture, there is an option to set the last stage (or the last two stages, if FFT Mode is dynamic) scaling to fixed scaling to improve the performance. • In Block Floating Point scaling, the dynamic range for the intermediate computation is increased by extracting a common exponent for all the data points in each stage and using the full arithmetic width for processing only the mantissa. An additional output port exponent is added to the FFT compiler core when this option is selected. This option is not available for the Streaming I/O mode. Data Width Input Data Width This parameter specifies input data width (width of either of the components: real or imaginary). IPUG54_01.9, August 2011 18 FFT Compiler IP Core User’s Guide Lattice Semiconductor Parameter Settings Twiddle Factor Width This parameter specifies twiddle factor width (width of either of the components: real or imaginary). Precision Reduction Method This parameter specifies whether the data is truncated or rounded nearest during scaling. In rounding mode, for high performance architecture, there is an additional option to set the last stage (or the last two, if FFT Mode is dynamic) to truncation. Setting the last stage to truncation results in better throughput. Implementation Tab Figure 3-3 shows the contents of the Implementation tab. Figure 3-3. Implementation Tab Multiplier Type This option specifies whether DSP blocks or LUTs are used for implementing multipliers and multiply-add components. Pipeline Multiplier Pipeline This option is used to specify the pipeline of LUT based multipliers in low resource mode. Higher values for pipeline leads to better performance at the cost of slightly increased utilization and latency. Adder Pipeline This option is used to specify an additional pipeline after the adders. Additional pipeline leads to better performance at the cost of slightly increased utilization and latency. This option is available only when architecture is low resource and the scale mode is not block floating point. Memory Type This parameter specifies the balance between using EBR and distributed memories. If EBR memory is selected, EBRs are used for memory depths 32 and higher. If the Distributed Memory option is selected, EBR memories are used only for depths 512 or more and the rest uses distributed memories. In the automatic option, the IP generator uses a pre-defined setting to select the EBR and distributed memories based on the FFT parameters. IPUG54_01.9, August 2011 19 FFT Compiler IP Core User’s Guide Lattice Semiconductor Parameter Settings Synthesis and Simulation Tools Tab Figure 3-3 shows the contents of the Synthesis and Simulation Tools tab. Figure 3-4. Synthesis and Simulation Tools Tab Support Synplify If selected, IPexpress generates evaluation scripts and other associated files required to synthesize the top-level design using the Synplify synthesis tool. Support Precision If selected, IPexpress generates evaluation script and other associated files required to synthesize the top-level design using the Precision synthesis tool. Support ModelSim If selected, IPexpress generates evaluation script and other associated files required to run simulation using the ModelSim simulator. Support ALDEC If selected, IPexpress generates evaluation script and other associated files required to run simulation using the ALDEC simulator. IPUG54_01.9, August 2011 20 FFT Compiler IP Core User’s Guide Chapter 4: IP Core Generation This chapter provides information on licensing the FFT Compiler IP core, generating the core using the Diamond or ispLEVER software IPexpress tool, running functional simulation, and including the core in a top-level design. The Lattice FFT Compiler IP core can be used in LatticeECP3, LatticeECP2/M, LatticeECP, and LatticeXP2 device families. Licensing the IP Core An IP license is required to enable full, unrestricted use of the FFT Compiler IP core in a complete, top-level design. An IP license that specifies the IP core and device family is required to enable full use of the core in Lattice devices. Instructions on how to obtain licenses for Lattice IP cores are given at: http://www.latticesemi.com/products/intellectualproperty/aboutip/isplevercoreonlinepurchas.cfm Users may download and generate the IP core and fully evaluate the core through functional simulation and implementation (synthesis, map, place and route) without an IP license. The FFT Compiler IP core also supports Lattice’s IP hardware evaluation capability, which makes it possible to create versions of the IP core that operate in hardware for a limited time (several hours) without requiring an IP license (see “Instantiating the Core” on page 25 for further details). However, a license is required to enable timing simulation, to open the design in the Diamond or ispLEVER EPIC tool, and to generate bitstreams that do not include the hardware evaluation timeout limitation. Getting Started The FFT Compiler IP core is available for download from the Lattice IP Server using the IPexpress tool. The IP files are automatically installed using InstallShield technology in any customer-specified directory. After the IP core has been installed, the IP core will be available in the IPexpress GUI dialog box shown in Figure 4-1. The IPexpress tool GUI dialog box for the FFT Compiler IP core is shown in Figure 4-1. To generate a specific IP core configuration the user specifies: • Project Path – Path to the directory where the generated IP files will be loaded. • File Name – “username” designation given to the generated IP core and corresponding folders and files. • (Diamond) Module Output – Verilog or VHDL. • (ispLEVER) Design Entry Type – Verilog HDL or VHDL. • Device Family – Device family to which IP is to be targeted (e.g. LatticeSCM, Lattice ECP2M, LatticeECP3, etc.). Only families that support the particular IP core are listed. • Part Name – Specific targeted part within the selected device family. IPUG54_01.9, August 2011 21 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation Figure 4-1. IPexpress Dialog Box (Diamond Version) Note that if the IPexpress tool is called from within an existing project, Project Path, Module Output (Design Entry in ispLEVER), Device Family and Part Name default to the specified project parameters. Refer to the IPexpress tool online help for further information. To create a custom configuration, the user clicks the Customize button in the IPexpress tool dialog box to display the FFT Compiler IP core Configuration GUI, as shown in Figure 4-2. From this dialog box, the user can select the IP parameter options specific to their application. Refer to “Parameter Settings” on page 16 for more information on the parameter settings. IPUG54_01.9, August 2011 22 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation Figure 4-2. Configuration Dialog Box (Diamond Version) IPUG54_01.9, August 2011 23 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation IPexpress-Created Files and Top Level Directory Structure When the user clicks the Generate button in the IP Configuration dialog box, the IP core and supporting files are generated in the specified “Project Path” directory. The directory structure of the generated files is shown in Figure 4-3. Figure 4-3. Lattice FFT Compiler IP Core Directory Structure Table 4-1 provides a list of key files created by the IPexpress tool and how they are used. The IPexpress tool creates several files that are used throughout the design cycle. The names of most of the created files are customized to the user’s module name specified in the IPexpress tool. Table 4-1. File List File Description _inst.v This file provides an instance template for the IP. .v This file provides the FFT core for simulation. _beh.v This file provides a behavioral simulation model for the FFT core. _bb.v This file provides the synthesis black box for the user’s synthesis. .ngo The ngo files provide the synthesized IP core. .lpc This file contains the IPexpress tool options used to recreate or modify the core in the IPexpress tool. .ipx The IPX file holds references to all of the elements of an IP or Module after it is generated from the IPexpress tool (Diamond version only). The file is used to bring in the appropriate files during the design implementation and analysis. It is also used to re-load parameter settings into the IP/Module generation GUI when an IP/Module is being re-generated. _top.[v,vhd] This file provides a module which instantiates the FFT core. This file can be easily modified for the user's instance of the FFT core. This file is located in the fft_comp_eval//src/rtl/top directory. IPUG54_01.9, August 2011 24 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation Table 4-1. File List (Continued) File Description twidx.mem Twiddle factor to initialize ROM. The “x” in file name can be 0,1,2, etc. _generate.tcl This file is created when GUI “Generate” button is pushed and generation is invoked. This file may be run from command line. _generate.log This is the IPexpress scripts log file. _gen.log This is the IPexpress IP generation log file. Instantiating the Core The \ and subtending directories provide files supporting FFT IP core evaluation. The \ directory shown in Figure 4-3 contains files and folders with content that is constant for all configurations of the FFT. The \ subfolder (\fft_comp_core0 in this example) contains files and folders with content specific to the username configuration. The \fft_comp_eval directory is created by IPexpress the first time the core is generated and updated each time the core is regenerated. A \ directory is created by IPexpress each time the core is generated and regenerated each time the core with the same file name is regenerated. A separate \ directory is generated for cores with different names, e.g. \fft_compiler_eval0, \fft_compiler_eval1, etc. Running Functional Simulation Simulation support for the FFT IP core is provided for Aldec Active-HDL (Verilog and VHDL) simulator and for Mentor Graphics ModelSim simulator. The functional simulation includes a configuration-specific behavioral model of the FFT IP core. The test bench sources stimulus to the core, and monitors output from the core. The generated IP core package includes the configuration-specific behavior model (_beh.v) for functional simulation in the “Project Path” root directory. The simulation scripts supporting ModelSim evaluation simulation is provided in \\fft_comp_eval\\sim\modelsim\scripts. The simulation script supporting Aldec evaluation simulation is provided in  \\fft_comp_eval\\sim\aldec\scripts. Both ModelSim and Aldec simulation are supported via test bench files provided in \\fft_comp_eval\testbench. Models required for simulation are provided in the corresponding \models folder. Users may run the Aldec evaluation simulation by doing the following: 1. Open Active-HDL. 2. Under the Tools tab, select Execute Macro. 3. Browse to folder \\fft_comp_eval\\sim\aldec\scripts and execute one of the “do” scripts shown. Users may run the ModelSim evaluation simulation by doing the following: 1. Open ModelSim. 2. Under the File tab, select Change Directory and choose the folder \fft_comp_eval\\sim\modelsim\scripts. 3. Under the Tools tab, select Execute Macro and execute the ModelSim “do” script shown. Note: When the simulation completes, a pop-up window will appear asking “Are you sure you want to finish?” Answer No to analyze the results. Answering Yes closes ModelSim. Synthesizing and Implementing the Core in a Top-Level Design Synthesis support for the FFT IP core is provided for Mentor Graphics Precision or Synopsys Synplify. The FFT IP core itself is synthesized and is provided in NGO format when the core is generated in IPexpress. Users may syn- IPUG54_01.9, August 2011 25 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation thesize the core in their own top-level design by instantiating the core in their top-level as described previously and then synthesizing the entire design with either Synplify or Precision RTL Synthesis. The following text describes the evaluation implementation flow for Windows platforms. The flow for Linux and UNIX platforms is described in the Readme file included with the IP core. The top-level files _top.v are provided in \\fft_comp_eval\\src\rtl\top. Push-button implementation of the reference design is supported via Diamond or ispLEVER project files, .syn, located in the following directory: \\fft_comp_eval\\impl\(synplify or precision). To use this project file in Diamond: 1. Choose File > Open > Project. 2. Browse to  \\fft_comp_eval\\impl\synplify (or precision) in the Open Project dialog box. 3. Select and open .ldf. At this point, all of the files needed to support top-level synthesis and implementation will be imported to the project. 4. Select the Process tab in the left-hand GUI window. 5. Implement the complete design via the standard Diamond GUI flow. To use this project file in ispLEVER: 1. Choose File > Open Project. 2. Browse to  \\fft_comp_eval\\impl\synplify (or precision) in the Open Project dialog box. 3. Select and open .syn. At this point, all of the files needed to support top-level synthesis and implementation will be imported to the project. 4. Select the device top-level entry in the left-hand GUI window. 5. Implement the complete design via the standard ispLEVER GUI flow. Hardware Evaluation The FFT Compiler IP core supports supports Lattice’s IP hardware evaluation capability, which makes it possible to create versions of the IP core that operate in hardware for a limited period of time (several hours) without requiring the purchase of an IP license. It may also be used to evaluate the core in hardware in user-defined designs. Enabling Hardware Evaluation in Diamond Choose Project > Active Strategy > Translate Design Settings. The hardware evaluation capability may be enabled/disabled in the Strategy dialog box. It is enabled by default. Enabling Hardware Evaluation in ispLEVER In the Processes for Current Source pane, right-click the Build Database process and choose Properties from the dropdown menu. The hardware evaluation capability may be enabled/disabled in the Properties dialog box. It is enabled by default. IPUG54_01.9, August 2011 26 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation Updating/Regenerating the IP Core By regenerating an IP core with the IPexpress tool, you can modify any of its settings including device type, design entry method, and any of the options specific to the IP core. Regenerating can be done to modify an existing IP core or to create a new but similar one. Regenerating an IP Core in Diamond To regenerate an IP core in Diamond: 1. In IPexpress, click the Regenerate button. 2. In the Regenerate view of IPexpress, choose the IPX source file of the module or IP you wish to regenerate. 3. IPexpress shows the current settings for the module or IP in the Source box. Make your new settings in the Target box. 4. If you want to generate a new set of files in a new location, set the new location in the IPX Target File box. The base of the file name will be the base of all the new file names. The IPX Target File must end with an .ipx extension. 5. Click Regenerate. The module’s dialog box opens showing the current option settings. 6. In the dialog box, choose the desired options. To get information about the options, click Help. Also, check the About tab in IPexpress for links to technical notes and user guides. IP may come with additional information. As the options change, the schematic diagram of the module changes to show the I/O and the device resources the module will need. 7. To import the module into your project, if it’s not already there, select Import IPX to Diamond Project (not available in stand-alone mode). 8. Click Generate. 9. Check the Generate Log tab to check for warnings and error messages. 10.Click Close. The IPexpress package file (.ipx) supported by Diamond holds references to all of the elements of the generated IP core required to support simulation, synthesis and implementation. The IP core may be included in a user's design by importing the .ipx file to the associated Diamond project. To change the option settings of a module or IP that is already in a design project, double-click the module’s .ipx file in the File List view. This opens IPexpress and the module’s dialog box showing the current option settings. Then go to step 6 above. Regenerating an IP Core in ispLEVER To regenerate an IP core in ispLEVER: 1. In the IPexpress tool, choose Tools > Regenerate IP/Module. 2. In the Select a Parameter File dialog box, choose the Lattice Parameter Configuration (.lpc) file of the IP core you wish to regenerate, and click Open. 3. The Select Target Core Version, Design Entry, and Device dialog box shows the current settings for the IP core in the Source Value box. Make your new settings in the Target Value box. 4. If you want to generate a new set of files in a new location, set the location in the LPC Target File box. The base of the .lpc file name will be the base of all the new file names. The LPC Target File must end with an .lpc extension. 5. Click Next. The IP core’s dialog box opens showing the current option settings. IPUG54_01.9, August 2011 27 FFT Compiler IP Core User’s Guide Lattice Semiconductor IP Core Generation 6. In the dialog box, choose desired options. To get information about the options, click Help. Also, check the About tab in the IPexpress tool for links to technical notes and user guides. The IP core might come with additional information. As the options change, the schematic diagram of the IP core changes to show the I/O and the device resources the IP core will need. 7. Click Generate. 8. Click the Generate Log tab to check for warnings and error messages. IPUG54_01.9, August 2011 28 FFT Compiler IP Core User’s Guide Chapter 5: Support Resources This chapter contains information about Lattice Technical Support, additional references, and document revision history. Lattice Technical Support There are a number of ways to receive technical support. Online Forums The first place to look is Lattice Forums (http://www.latticesemi.com/support/forums.cfm). Lattice Forums contain a wealth of knowledge and are actively monitored by Lattice Applications Engineers. Telephone Support Hotline Receive direct technical support for all Lattice products by calling Lattice Applications from 5:30 a.m. to 6 p.m. Pacific Time. • For USA & Canada: 1-800-LATTICE (528-8423) • For other locations: +1 503 268 8001 In Asia, call Lattice Applications from 8:30 a.m. to 5:30 p.m. Beijing Time (CST), +0800 UTC. Chinese and English language only. • For Asia: +86 21 52989090 E-mail Support • techsupport@latticesemi.com • techsupport-asia@latticesemi.com Local Support Contact your nearest Lattice Sales Office. Internet www.latticesemi.com References [1] Lawrence R. Rabiner and Bernard Gold, “Theory and Application of Digital Signal Processing”, Prentice-Hall, Inc., Englewood Cliffs, NJ, 1975. LatticeECP • HB1000, LatticeECP/EC Family Handbook LatticeECP2/M • HB1003, LatticeECP2/M Family Handbook LatticeECP3 • HB1009, LatticeECP3 Family Handbook LatticeXP2 • DS1009, Lattice XP2 Datasheet IPUG54_01.9, August 2011 29 FFT Compiler IP Core User’s Guide Lattice Semiconductor Support Resources Revision History Date Document Version IP Version Change Summary — — — Previous Lattice releases. September 2006 01.2 2.0 Added LatticeECP2 device support. February 2007 01.3 2.1 Added LatticeECP2M device support. Updated for ispLEVER 6.1 SP1. May 2007 01.4 2.2 Added support for LatticeXP2 FPGA family. February 2008 01.5 2.3 Updated appendices. May 2009 01.6 2.4 Added support for LatticeECP3, updated figures and tables. June 2010 01.7 2.4 Divided document into chapters. Added table of contents. Added Quick Facts tables in Chapter 1, “Introduction.” Added new content in Chapter 4, “IP Core Generation.” August 2010 01.8 2.5 Added support for Diamond software throughout. August 2011 01.9 2.5 Added Output Latency section. IPUG54_01.9, August 2011 30 FFT Compiler IP Core User’s Guide Appendix A: Resource Utilization This appendix gives resource utilization information for Lattice FPGAs using the FFT Compiler IP core. The IP configurations shown in this chapter were generated using the IPexpress software tool. IPexpress is the Lattice IP configuration utility, and is included as a standard feature of the Diamond and ispLEVER design tools. Details regarding the usage of IPexpress can be found in the IPexpress and Diamond or ispLEVER help system. For more information on the Diamond or ispLEVER design tools, visit the Lattice web site at: www.latticesemi.com/software. LatticeECP Devices Table A-1. Performance and Resource Utilization1 # Points Operating Mode SLICEs LUTs Registers sysMEM™ EBRs sysDSP™ Blocks fMAX (MHz) 64 Low resource 488 588 705 3 1 213 64 High performance 998 1444 1174 1 2 208 256 Low resource 575 732 772 3 1 212 256 High performance 1424 2060 1636 3 3 200 1024 Low resource 702 957 861 5 1 192 1024 High performance 1781 2589 2056 7 4 182 8192 Low resource 769 1055 913 36 1 189 8192 High performance 2579 3950 2762 38 6 179 1. Performance and utilization data are generated targeting a LFECP33E-5F672C device using Lattice Diamond 1.0 and Synplify Pro D2009.12L-1 software. Performance may vary when using a different software version or targeting a different device density or speed grade within the LatticeECP family. Ordering Part Number The Ordering Part Number (OPN) for the FFT Compiler for LatticeECP/EC devices is FFT-COMP-E2-U2. LatticeECP2 Devices Table A-2. Performance and Resource Utilization1 # Points Operating Mode SLICEs LUTs Registers sysMEM EBRs sysDSP Blocks fMAX (MHz) 64 Low resource 499 608 695 3 1 283 64 High performance 723 1435 1174 1 2 283 256 Low resource 578 745 755 3 1 250 256 High performance 1049 2078 1636 3 3 250 1024 Low resource 666 894 804 3 1 258 1024 High performance 1324 2626 2056 6 4 258 8192 Low resource 787 1123 897 18 1 240 8192 High performance 1952 3880 2762 21 6 240 1. Performance and utilization data are generated targeting a LFE2-50-7F672C device using Lattice Diamond 1.0 and Synplify Pro D2009.12L-1 software. Performance may vary when using a different software version or targeting a different device density or speed grade within the LatticeECP2 family. Ordering Part Number The Ordering Part Number (OPN) for the FFT Compiler for LatticeECP2 devices is FFT-COMP-P2-U2. IPUG54_01.9, August 2011 31 FFT Compiler IP Core User’s Guide Lattice Semiconductor Resource Utilization LatticeECP2M Devices Table A-3. Performance and Resource Utilization1 # Points 64 SLICEs LUTs Registers sysMEM EBRs sysDSP Blocks fMAX (MHz) Low resource 534 681 730 3 1 286 Operating Mode 64 High performance 989 1437 1174 1 2 243 256 Low resource 616 830 789 3 1 292 256 High performance 1419 2070 1636 3 3 266 1024 Low resource 718 991 839 3 1 273 1024 High performance 1782 2618 2056 6 4 261 8192 Low resource 787 1123 897 18 1 234 8192 High performance 2518 3867 2762 21 6 229 1. Performance and utilization data are generated targeting a LFE2M-35E-7F672C device using Lattice Diamond 1.0 and Synplify Pro D2009.12L-1 software. Performance may vary when using a different software version or targeting a different device density or speed grade within the LatticeECP2M family. Ordering Part Number The Ordering Part Number (OPN) for the FFT Compiler for LatticeECP2M devices is FFT-COMP-PM-U2. LatticeECP3 Devices Table A-4. Performance and Resource Utilization1 # Points Operating Mode SLICEs LUTs Registers sysMEM EBRs sysDSP Blocks fMAX (MHz) 64 Low resource 493 668 727 3 2 285 64 High performance 733 1433 1174 1 2 243 256 Low resource 575 821 783 3 2 255 256 High performance 1065 2074 1636 3 3 266 1024 Low resource 653 964 820 3 2 254 1024 High performance 1329 2602 2056 6 4 261 8192 Low resource 722 1089 870 18 2 258 8192 High performance 1957 3843 2762 21 6 229 1. Performance and utilization data are generated targeting a LFE3-95E-8FN672CES device using Lattice Diamond 1.0 and Synplify Pro D2009.12L-1 software. Performance may vary when using a different software version or targeting a different device density or speed grade within the LatticeECP3 family. Ordering Part Number The Ordering Part Number (OPN) for the FFT Compiler for LatticeECP3 devices is FFT-COMP-E3-U2. IPUG54_01.9, August 2011 32 FFT Compiler IP Core User’s Guide Lattice Semiconductor Resource Utilization LatticeXP2 Devices Table A-5. Performance and Resource Utilization1 # Points 64 SLICEs LUTs Registers sysMEM EBRs sysDSP Blocks fMAX (MHz) Low resource 534 681 730 3 1 261 Operating Mode 64 High performance 989 1437 1174 1 2 243 256 Low resource 616 830 789 3 1 226 256 High performance 1419 2070 1636 3 3 266 1024 Low resource 718 991 839 3 1 237 1024 High performance 1782 2618 2056 6 4 261 1. Performance and utilization data are generated targeting a LFXP2-17E-7F484C device using Lattice Diamond 1.0 and Synplify Pro D2009.12L-1 software. Performance may vary when using a different software version or targeting a different device density or speed grade within the LatticeXP2 family. Ordering Part Number The Ordering Part Number (OPN) for the FFT Compiler for LatticeXP2 devices is FFT-COMP-X2-U2. IPUG54_01.9, August 2011 33 FFT Compiler IP Core User’s Guide