Progetto di circuiti su FPGA Flusso di Sviluppo • Descrizione comportamentale • Simulazione comportamentale • Sintesi • Descrizione strutturale (Netlist) • Simulazione con ritardi approssimativi • Implementazione • Simulazione dettagliata • Download • Simulazione sul campo Descrizione COMPORTAMENTALE Simulazione COMPORTAMENTALE Sintesi Descrizione STRUTTURALE (Netlist) Simulazione STRUTTURALE Implementaziome Descrizione FISICA Simulazione DETTAGLIATA Realizzazione Simulazione SUL CAMPO Descrizione Comportamentale • Descrizione del tutto ideale del funzionamento del dispositivo ad alto livello di astrazione – non prevede ritardi di alcun tipo Es: C <= (A + B)* C Sintesi • Traduzione dalla descrizione comportamentale all strutturale, ossia a livello di porte logiche (Netlist) – Si basa sulla presenza di opportune librerie Descrizione COMPORTAMENTALE Analisi • E’ suddivisa nelle seguenti fasi – Analisi (analisi sintattica del sorgente) – Compilazione • traduzione a livello RTL (Register Transfer Level) • estrazione delle macro – Ottimizzazione (miglioramento della logica, espansione delle macro) • E’ un procedimento “guidato” – abbisogna di opportuni vincoli Lib. Vincoli Compilazione Ottimizzazione Descrizione STRUTTURALE Descrizione Strutturale • Descrizione del circuito il termini di blocchi logici (porte, flip-flop, registri, memorie, …) presenti in libreria opportunamente collegati ossia in termini di NETLIST A • Le informazioni portate da B questa descrizione sono: – Ritardi della logica (ma non dei collegamenti) – Area occupata dalla logica (ma non dai collegamenti) C D Z Implementazione • Traduce la descrizione a celle logiche in una opportuna descrizione fisica – maschere per il layout (ASIC) – file di download (FPGA - CPLD) • Si divide in quattro fasi – Translate ( fusione con blocchi preconfigurati) – Mapping (mappatura della logica nei CLB) solo per FPGA – Placement (Posizionamento dei blocchi logici) – Routing Collegamento tra CLB • Il procedimento si basa su vincoli implementativi Descrizione STRUTTURALE Translate Mapping Placement Routing Descrizione FISICA Descrizione Dettagliata • La descrizione del circuito e’ molto vicina a quelle che saranno le reali prestazioni. Vi sono infatti informazioni su – – – – Ritardi della logica Ritardi nei collegamenti Area occupata dalla logica Area occupata dai canali di collegamento – Dettaglio sull’ ubicazione di ogni singola porta logica e/o dei piedini di I/O Back Annotation • Alcuni dei risultati ottenuti ai vari passi del procedimento spesso vengono utilizzati in un procedimento di “backannotation” ossia vengono riportati ai passi superiori quali vincoli o per avere descrizioni piu’ dettagliate del funzionamento del circuito anche ad alti livelli di astrazione – Vincoli – Ritardi ISE Software Flow FPGA Design Workshop Software makes a difference • Device capabilities are worthless if you can’t use them in YOUR course • Design software should support all ranges of designs from CPLD to the high-density FPGA • Works with YOUR design flow – minimize impacts to the design cycle – work with the tools you already own Foundation Series ISE • Foundation Series ISE (Integrated Software Environment) • For PC platforms: Win98, Win2000, and NT4.0 • For UNIX platforms: HP and Solaris Xilinx Design Flow Plan & Budget Create Code/ Schematic HDL RTL Simulation Implement Translate Functional Simulation Synthesize to create netlist Map Place & Route Attain Timing Closure Timing Simulation Create Bit File Advanced design management through project navigator • Unix & PC platforms • Complete file management • Automates design flow – – – – – Entry Synthesis Implementation Simulation Programming Device Support • New leading-edge device families • ISE advantages can be leveraged across all device families and design sizes Processes and Tools Step 1:Design Step 2: Synthesize to create netlist Step 3: Implement design Some tools are listed multiple times with different task names Step 4: Configuration Context Sensitive Flow Only relevant processes are displayed to the user Guides the user to the “next step” for that source HDL Module Selected HDL Test Bench Selected Process Available Includes Synthesis and P&R Only HDL Simulation process is available ISE Push Button Flow • Select a desired end result -- all necessary processe and dependencies automatically run to produce the result • Simple three-step process to get results 1 Add Files 2 3 Select Top Level Double Click Desired End Point Lab 1: ISE Flows • Introduction to the ISE flow – Step through the FPGA design flow with a simple design – Download the generated bitstream to the XESS - XSA50 demo board Design Entry • Two design entry methods: HDL or schematic – Architecture Wizard and Core Generator available to assist design entry • Whichever method you use, you will need a tool to generate an EDIF netlist to program a Xilinx FPGA – Popular synthesis tools: Synplify, Leonardo Spectrum, FPGA Compiler II, and XST • Simulate design so that it works as expected! Plan & Budget ... Create Code/ Schematic Functional Simulation HDL RTL Simulation Synthesize to create netlist Schematic Source File • Create a new schematic source: Project New Source Schematic • Components from Xilinx Unified Libraries • HDL keywords cannot be used on schematics • Unified components require all input pins to be connected – Tie unused pins, both inputs and outputs, to GND or VCC Options and Symbols The Options tab selections change, depending on which function is selected For example, if you are adding a net name, the net name options would be shown • • • • Components are divided into categories Exact symbols are located in the Symbol box Symbol Name Filter for easier search Orientation – Rotate 0, 90,180, 270 – Mirror and rotate 0, 90, 180, 270 HDL Source File • Types of HDL source files – VHDL logic description (.vhd extension) – Verilog logic description (.v extension) – ABEL-HDL logic description (.abl extension) • Selecting these source types will open a text editor for you to enter the design code Xilinx CORE Generator System GUI Cores can be organized by function, vendor, or device family Core type, version, device support, and vendor Core Customize Window Core Overview tab provides version information and a brief functional description Parameters tab allows you to customize the core Contact tab provides information about the vendor Data sheet access What is Implementation? • More than just “Place & Route” • Implementation includes many phases – Translate: Merge multiple design files into a single netlist – Map: Group logical symbols from the netlist (gates) into physical components (CLBs and IOBs) – Place & Route: Place components onto the chip, connect them, and extract timing data into reports • Each phase generates files that allow you to use other Xilinx tools (such as Floorplanner, FPGA Editor, XPower, Multi-Pass Place & Route) Implement • Each implementation stage can be expanded to view the available sub-tools and sub-processes – Translate • Create post-translate simulation model – Map • Floorplan • Manual route with FPGA Editor – Place & Route • • • • Static timing Floorplanner, view placed design FPGA Editor, view routed design Analyze power Download • Once a design is implemented, you must create a file that the FPGA can understand – This file is called a bit stream: a BIT file (.bit extension) • The BIT file can be downloaded directly to the FPGA, or can be converted into a PROM file which stores the programming information Program the FPGA • There are two ways to program an FPGA – Through a PROM device • You will need to generate a file that the PROM programmer will understand – Directly from the computer • Use the iMPACT configuration tool RTL Viewer for XST Helps debug design connectivity, design speed Architecture Wizards • Simplifies design of complex components – Generates HDL files & .ucf • Supports: – DCM – RocketIOTM transceivers • Including Channel Bonding State Diagram Source • Files with .dia extension • Selecting this source type will invoke StateCAD HDL Bencher • Multiple Clock and Asynchronous Signal Support 2) Choose Clocks 1) Select Multiple Clocks and/or Asynch Signal Support 3) Associate Signals with Clocks or Assign as Asynchronous HDL Bencher • Multiple Clock and Asynchronous Signal Support 4) Specify Timing for Each Clock 5) Define Waveforms Incremental Design Make small changes quickly! – Re-implements only the changed modules – Keeps placement and routing – Easy set-up through floorplanning along HDL hierarchy boundaries – Works with HDL designs • don’t optimize across hierarchy – More turns per day – More repeatable results PACE simplifies pin and area assignments – PACE (Pinout and Area Constraints Editor) – Create groups for busses and standard outputs – Color-coded banks – Drag-and-drop pin assignments – Interactive DRC – Automatic differential I/O pairing – Logic size to area checking Constraints Improvement Wizard • Gives suggestions on how to constrain unconstrained paths Summary • The Xilinx design process contains only four steps: design, synthesize, implement, configure • The Xilinx design process can all be done through the ISE Project Navigator