Evaluation of the Hardwired Sequence Control System Generated by High-Level Synthesis,” Proc

Total Page:16

File Type:pdf, Size:1020Kb

Evaluation of the Hardwired Sequence Control System Generated by High-Level Synthesis,” Proc This is the accepted version of the following article: Naoki Fujieda, Shuichi Ichikawa, Yoshiki Ishigaki, and Tasuku Tanaka, “Evaluation of the hardwired sequence control system generated by high-level synthesis,” Proc. 2017 IEEE International Symposium on Industrial Electronics (ISIE 2017) (06/2017), which has been published in final form at http://dx.doi.org/10.1109/ISIE.2017.8001426. ⃝c 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Evaluation of the hardwired sequence control system generated by high-level synthesis Naoki Fujieda, Shuichi Ichikawa, Yoshiki Ishigaki, and Tasuku Tanaka Department of Electrical and Electronic Information Engineering, Toyohashi University of Technology, Toyohashi, Aichi, Japan [email protected] (Naoki Fujieda), [email protected] (Shuichi Ichikawa) Abstract—This study presents the application of the commer- because the logic circuit is more difficult to target for analysis cial High Level Synthesis (HLS) to a hardwired control appli- and duplication than software. cation with quantitative comparison to the traditional approach that uses logic synthesis with HDL. Though the derived circuits In early studies, Hardware Description Languages (HDL) from HLS are comparable to that of logic synthesis, the design and logic synthesis systems were used to generate a hardwired trade-offs in HLS are difficult to control. This study also presents control circuit. Then, High Level Synthesis (HLS) systems the design and evaluation of the whole system of hardwired appeared and became commercially available to generate a control with a Xilinx Zynq-7000 FPGA platform. From our logic circuit from a popular programming language (e.g. C experiments, two performance bottlenecks were identified: the RAM for memory elements that serializes the read/write accesses, language). With HLS systems, designers can specify hardware and data transfer time via the peripheral bus. According to our functionality at higher level of abstraction to reduce the time results, the sole hardwired control was 10 times faster than the and efforts required for hardware design [14]. original software, while the overall performance was 4 to 50 times worse than the original software. The use of flipflops and On the other hand, HLS tends to be more complex in nature dedicated I/O pins are necessary for high-performance systems. than logic synthesis. Many techniques and optimizations are applied to the design, which involves a large numbers of options and settings. Though the default features of HLS may I. INTRODUCTION yield good results in many cases, it is not always optimal Programmable Logic Controller (PLC) is widely adopted for for a specific application. It might be difficult for users to the sequence control of industrial machinery. Although PLC is control HLS to generate a good hardwired control circuit, and flexible and reliable, there are two problems; i.e., performance the derived circuit might be larger or slower than the circuit and security. The first problem is that the performance of PLC generated from HDL with logic synthesis. All these points does not always satisfy the requirements of large or highly should be clarified before adopting HLS for practical projects. responsive control systems. The second problem comes from The first purpose of this study is to examine the potential of the fact that PLC software is an easy target to duplicate and commercial HLS for control applications, compared to the tra- analyze. ditional approach that adopts logic synthesis and HDL. Three As an answer to these problems, there have been studies to sample programs are converted to logic circuits with HLS and implement PLC software as a hardwired control circuit on a traditional methods, and the derived circuits are quantitatively reconfigurable logic device, e.g. FPGA (Field Programmable evaluated and compared. Though various preceding studies Gate Array). The performance of hardwired logic is generally reported the hardwired control generated by synthesis, HLS, higher than PLC software, and the flexibility of control logic and specially developed tools, they did not compare different can be sustained by the reconfiguration of an FPGA device. approaches in a quantitative manner. To the best of the authors’ Hardwired control is also resistant to duplication and analysis, knowledge, this is the first work that quantitatively discusses the pros and cons of the HLS approach over the traditional approach for hardwired control circuits. ^_ '$SS ,"S,-2 TTX\\T2&#, Another purpose of this work is to examine the performance [X\\. #*1# and the resource requirements of the whole system, which [X\ \. includes the hardwired control circuit as a peripheral device. #,"'$. Though the preceding studies focused on the hardwired control circuit itself (e.g., [5]), a practical system consists of many other parts such as an embedded processor, bus interfaces, Fig. 1. An example of conversion from PLC program to VHDL [6]. memory modules, etc. The performance of the whole system might be much different from that of the sole hardwired control part, because various overheads are involved in the optimization efforts might be drastically reduced by using whole system. The resource requirement of the whole system HLS, which incorporates such optimization techniques [14]. should be also examined along with the performance. This To utilize recent HLS system, PLC instructions have to be study quantitatively discusses such practical aspects of the converted into C language (Figure 2). hardwired control. Economakos [3][4] translated the STL language of Siemens The rest of this paper is organized as follows. Section II S7 PLC into C language, which was then converted to HDL outlines the background and related studies of this work. with Catapult C HLS software. Economakos discussed the Section III introduces the methods to convert PLC software effects of coding styles to the area and the performance in into hardware. Section IV describes the overall design of the FPGA implementations. The use of C-based HLS is par- hardware control system, and Section V presents the evaluation ticularly attractive when the user would like to implement the results. Section VI concludes the paper. hardware partially, leaving most of the software executed as software. Recent HLS systems support hardware/software co- II. RELATED STUDIES design, which enables users to examine the trade-offs between The hardware implementation of control logic has been software and hardware. This approach is also expected to studied since the middle of 1990s, along with the rapid be applicable to C-based PLCs, which have become popular evolution of FPGA technologies. Adamski [1] and Wegrzyn among PLC users. [15] presented systems that transfer the PLC program in Despite the pioneering works by Economakos, it is still Petri-net format into HDL. Ikeshita et al. [7] converted PLC indefinite whether the current HLS system can replace the old software in SFC into Verilog HDL, while Miyazawa et al. [13] design scheme. If the optimizations of HLS are insufficient proposed to convert the ladder diagram to VHDL. These early or not suited to hardwired control, the derived circuit might studies adopted HDL and logic synthesis to generate the logic be slower or larger than that generated by HDL and logic circuit that corresponds to the original PLC software. synthesis. Even if HLS is well established, it might be difficult Figure 1 illustrates an example of PLC software and the to derive the best result, since a commercial CAD system is a corresponding VHDL code [5]. First, a rung of ladder diagram kind of black box. It applies every possible techniques in the is compiled into the corresponding PLC instructions. The black box, and shows the sole result of trade-offs out of vast converter then reads the PLC instructions and converts them variety of possibilities. It is sometimes difficult to control the into the corresponding VHDL, rung by rung. The condition optimization process, and to analyze the factors separately. part of the rung is converted into the corresponding conditional This work discusses the pros and cons of an HLS approach statement, and the output part is converted into the correspond- over the traditional approach for hardwired control circuits, ing assignments. using Xilinx tools and FPGAs. The following studies explored the optimization schemes for hardwired control. As an example, Ichikawa et al. [5][6] III. CONVERSION OF PLC PROGRAM proposed to convert PLC software into logic circuit for higher Throughout this study, Mitsubishi MELSEC-Q series performance and higher security. They converted PLC instruc- PLC [12] is assumed as the target platform. Mitsubishi GX tions to VHDL code, which was then synthesized by using the Works [11] software with GX Converter [10] is adopted commercial logic synthesizer. The performance advantages of to generate the instruction list as a text file. The derived three design options (Sequential Design, Levelized Design, instruction list is specific to MELSEC PLCs, while it well and Flat Design) were quantitatively evaluated, where the resembles to the IEC-61131 instruction list. It is thus expected parallelism in hardwired control was utilized. Du et al. [2] that the following discussion holds good in IEC-compliant and Milik [9][8] also discussed the optimization techniques PLCs. to generate a high-performance control circuit from a PLC instruction sequence. A. PLC-to-C conversion Though the above studies discussed the techniques to utilize A custom-made converter was developed to convert a PLC the parallelism in the PLC instruction sequence, many of instruction sequence into the corresponding code of standard these techniques are common and well known in the com- C language. This converter is designated as the C-converter in puter architecture or design automation communities.
Recommended publications
  • Review of FPD's Languages, Compilers, Interpreters and Tools
    ISSN 2394-7314 International Journal of Novel Research in Computer Science and Software Engineering Vol. 3, Issue 1, pp: (140-158), Month: January-April 2016, Available at: www.noveltyjournals.com Review of FPD'S Languages, Compilers, Interpreters and Tools 1Amr Rashed, 2Bedir Yousif, 3Ahmed Shaban Samra 1Higher studies Deanship, Taif university, Taif, Saudi Arabia 2Communication and Electronics Department, Faculty of engineering, Kafrelsheikh University, Egypt 3Communication and Electronics Department, Faculty of engineering, Mansoura University, Egypt Abstract: FPGAs have achieved quick acceptance, spread and growth over the past years because they can be applied to a variety of applications. Some of these applications includes: random logic, bioinformatics, video and image processing, device controllers, communication encoding, modulation, and filtering, limited size systems with RAM blocks, and many more. For example, for video and image processing application it is very difficult and time consuming to use traditional HDL languages, so it’s obligatory to search for other efficient, synthesis tools to implement your design. The question is what is the best comparable language or tool to implement desired application. Also this research is very helpful for language developers to know strength points, weakness points, ease of use and efficiency of each tool or language. This research faced many challenges one of them is that there is no complete reference of all FPGA languages and tools, also available references and guides are few and almost not good. Searching for a simple example to learn some of these tools or languages would be a time consuming. This paper represents a review study or guide of almost all PLD's languages, interpreters and tools that can be used for programming, simulating and synthesizing PLD's for analog, digital & mixed signals and systems supported with simple examples.
    [Show full text]
  • Reconfigurable Computing in the New Age of Parallelism
    Reconfigurable Computing in the New Age of Parallelism Walid Najjar and Jason Villarreal Department of Computer Science and Engineering University of California Riverside Riverside, CA 92521, USA {najjar,villarre}@cs.ucr.edu Abstract. Reconfigurable computing is an emerging paradigm enabled by the growth in size and speed of FPGAs. In this paper we discuss its place in the evolution of computing as a technology as well as the role it can play in the cur- rent technology outlook. We discuss the evolution of ROCCC (Riverside Opti- mizing Compiler for Configurable Computing) in this context. Keywords: Reconfigurable computing, FPGAs. 1 Introduction Reconfigurable computing (RC) has emerged in recent years as a novel computing paradigm most often proposed as complementing traditional CPU-based computing. This paper is an attempt to situate this computing model in the historical evolution of computing in general over the past half century and in doing so define the parameters of its viability, its potentials and the challenges it faces. In this section we briefly summarize the main factors that have contributed to the current rise of RC as a computing paradigm. The RC model, its potentials and chal- lenges are described in Section 2. Section 3 describes the ROCCC 2.0 (Riverside Optimizing Compiler for Configurable Computing), a C to HDL compilation tool who objective is to raise the programming abstraction level for RC while providing the user with both a top down and a bottom-up approach to designing FPGA-based code accelerators. 1.1 The Role of the von Neumann Model Over a little more than half a century, computing has emerged from non-existence, as a technology, to being a major component in the world’s economy.
    [Show full text]
  • PACT HDL: AC Compiler Targeting Asics and Fpgas With
    PACT HDL: A C Compiler Targeting ASICs and FPGAs with Power and Performance Optimizations* Alex Jones Debabrata Bagchi Satrajit Pal Xiaoyong Tang Alok Choudhary Prith Banerjee Center for Parallel and Distributed Computing Department of Electrical and Computer Engineering Technological Institute, Northwestern University 2145 Sheridan Road, Evanston, IL 60208-3118 Phone: (847) 491-3641 Fax: (847) 491-4455 Email: {akjones, bagchi, satrajit, tang, choudhar, banerjee}@ece.northwestern.edu ABSTRACT 1. INTRODUCTION Chip fabrication technology continues to plunge deeper into sub- As chip fabrication processes progress deep into the sub-micron micron levels requiring hardware designers to utilize ever- level and Integrated Circuits (ICs) and Field Programmable Gate increasing amounts of logic and shorten design time. Toward that Arrays (FPGAs) can support larger and larger amounts of logic, end, high-level languages such as C/C++ are becoming popular system designers require increasingly high-level tools to keep up. for hardware description and synthesis in order to more quickly Recently, industry has targeted C/C++ and variants as potential leverage complex algorithms. Similarly, as logic density long-term replacements for Hardware Description Languages increases due to technology, power dissipation becomes a (HDLs) such as VHDL and Verilog currently employed for progressively more important metric of hardware design. PACT today’s hardware design. Also, as technologies increase in HDL, a C to HDL compiler, merges automated hardware density in both the fabricated and reconfigurable areas, power- synthesis of high-level algorithms with power and performance consumption becomes a progressively more important problem. optimizations and targets arbitrary hardware architectures, While some work has been done in targeting C/C++ as an HDL particularly in a System on a Chip (SoC) setting that incorporates and considering power-consumption in hardware synthesis, reprogrammable and application-specific hardware.
    [Show full text]
  • Embedded Processing Using Fpgas Agenda
    Embedded Processing Using FPGAs Agenda • Why FPGA Platform Based Embedded Processing • Embedded Use Models And Their FPGA Based Solutions • Architecture/Topology Choices • A Reoccurring Question: Hardware Or Software • Reconfigurable Hardware • Tool Flows For FPGA Based Embedded Systems 2 - Embedded Processing using FPGAs www.xilinx.com Agenda • Why FPGA Platform Based Embedded Processing • Embedded Use Models And Their FPGA Based Solutions • Architecture/Topology Choices • A Reoccurring Question: Hardware Or Software • Reconfigurable Hardware • Tool Flows For FPGA Based Embedded Systems 3 - Embedded Processing using FPGAs www.xilinx.com Why Use Processors In the First Place • Microcontrollers (µC) and Microprocessors (µP) Provide a Higher Level of Design Abstraction – Most µC functions can be implemented using VHDL or Verilog - downsides are parallelism & complexity – Using C/C++ abstraction & serial execution make certain functions much easier to implement in a µC • Discrete µCs are Inexpensive and Widely Used – µCs have years of momentum and software designers have vast experience using them 4 - Embedded Processing using FPGAs www.xilinx.com Why Embedded Design using FPGAs In Addition To The Universal Drive Towards Smaller Cheaper Faster With Less Power…. 1 Difficult to Find the Required Mix of Peripherals in Off the Shelf (OTS) Microcontroller Solutions •2 Selecting a Single Processor Core with Long Term Solution Viability is Difficult at Best •3 Without Direct Ownership of the Processing Solution, Obsolescence is Always a Concern
    [Show full text]
  • HDL and Programming Languages ■ 6 Languages ■ 6.1 Analogue Circuit Design ■ 6.2 Digital Circuit Design ■ 6.3 Printed Circuit Board Design ■ 7 See Also
    Hardware description language - Wikipedia, the free encyclopedia 페이지 1 / 11 Hardware description language From Wikipedia, the free encyclopedia In electronics, a hardware description language or HDL is any language from a class of computer languages, specification languages, or modeling languages for formal description and design of electronic circuits, and most-commonly, digital logic. It can describe the circuit's operation, its design and organization, and tests to verify its operation by means of simulation.[citation needed] HDLs are standard text-based expressions of the spatial and temporal structure and behaviour of electronic systems. Like concurrent programming languages, HDL syntax and semantics includes explicit notations for expressing concurrency. However, in contrast to most software programming languages, HDLs also include an explicit notion of time, which is a primary attribute of hardware. Languages whose only characteristic is to express circuit connectivity between a hierarchy of blocks are properly classified as netlist languages used on electric computer-aided design (CAD). HDLs are used to write executable specifications of some piece of hardware. A simulation program, designed to implement the underlying semantics of the language statements, coupled with simulating the progress of time, provides the hardware designer with the ability to model a piece of hardware before it is created physically. It is this executability that gives HDLs the illusion of being programming languages, when they are more-precisely classed as specification languages or modeling languages. Simulators capable of supporting discrete-event (digital) and continuous-time (analog) modeling exist, and HDLs targeted for each are available. It is certainly possible to represent hardware semantics using traditional programming languages such as C++, although to function such programs must be augmented with extensive and unwieldy class libraries.
    [Show full text]
  • An Automated Flow to Generate Hardware Computing Nodes from C for an FPGA-Based MPI Computing Network
    An Automated Flow to Generate Hardware Computing Nodes from C for an FPGA-Based MPI Computing Network by D.Y. Wang A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF BACHELOR OF APPLIED SCIENCE DIVISION OF ENGINEERING SCIENCE FACULTY OF APPLIED SCIENCE AND ENGINEERING UNIVERSITY OF TORONTO Supervisor: Paul Chow April 2008 Abstract Recently there have been initiatives from both the industry and academia to explore the use of FPGA-based application-specific hardware acceleration in high-performance computing platforms as traditional supercomputers based on clusters of generic CPUs fail to scale to meet the growing demand of computation-intensive applications due to limitations in power consumption and costs. Research has shown that a heteroge- neous system built on FPGAs exclusively that uses a combination of different types of computing nodes including embedded processors and application-specific hardware accelerators is a scalable way to use FPGAs for high-performance computing. An ex- ample of such a system is the TMD [11], which also uses a message-passing network to connect the computing nodes. However, the difficulty in designing high-speed hardware modules efficiently from software descriptions is preventing FPGA-based systems from being widely adopted by software developers. In this project, an auto- mated tool flow is proposed to fill this gap. The AUTO flow is developed to auto- matically generate a hardware computing node from a C program that can be used directly in the TMD system. As an example application, a Jacobi heat-equation solver is implemented in a TMD system where a soft processor is replaced by a hardware computing node generated using the AUTO flow.
    [Show full text]
  • Advances in Architectures and Tools for Fpgas and Their Impact on the Design of Complex Systems for Particle Physics
    Advances in Architectures and Tools for FPGAs and their Impact on the Design of Complex Systems for Particle Physics Anthony Gregersona, Amin Farmahini-Farahania, William Plishkerb, Zaipeng Xiea, Katherine Comptona, Shuvra Bhattacharyyab, Michael Schultea a University of Wisconsin - Madison b University of Maryland - College Park fagregerson, farmahinifar, [email protected] fplishker, [email protected] fcompton, [email protected] Abstract these trends has been a rapid adoption of FPGAs in HEP elec- tronics. A large proportion of the electronics in the Compact The continual improvement of semiconductor technology has Muon Solenoid Level-1 Trigger, for example, are based on FP- provided rapid advancements in device frequency and density. GAs, and many of the remaining ASICs are scheduled to be Designers of electronics systems for high-energy physics (HEP) replaced with FPGAs in proposed upgrades [1]. have benefited from these advancements, transitioning many de- signs from fixed-function ASICs to more flexible FPGA-based Improvements in FPGA technology are not likely to end platforms. Today’s FPGA devices provide a significantly higher soon. Today’s high-density FPGAs are based on a 40-nm sil- amount of resources than those available during the initial Large icon process and already contain an order of magnitude more Hadron Collider design phase. To take advantage of the ca- logic than the FPGAs available at planning stage of the Large pabilities of future FPGAs in the next generation of HEP ex- Hadron Collider’s electronics. 32 and 22 nm silicon process periments, designers must not only anticipate further improve- technologies have already been demonstrated to be feasible; ments in FPGA hardware, but must also adopt design tools and as FPGAs migrate to these improved technologies their logic methodologies that can scale along with that hardware.
    [Show full text]
  • Download The
    Techniques for Enabling In-System Observation-based Debug of High-Level Synthesis Circuits on FPGAs by Jeffrey Goeders BASc Computer Engineering, University Toronto, 2010 MASc Computer Engineering, The University of British Columbia, 2012 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Electrical and Computer Engineering) The University of British Columbia (Vancouver) September 2016 c Jeffrey Goeders, 2016 Abstract High-level synthesis (HLS) is a rapidly growing design methodology that allows designers to create digital circuits using a software-like specification language. HLS promises to increase the productivity of hardware designers in the face of steadily increasing circuit sizes, and broaden the availability of hardware acceleration, allowing software designers to reap the benefits of hardware implementation. One roadblock to HLS adoption is the lack of an in-system debugging infrastructure. Existing debug technologies are limited to software emulation and cannot be used to find bugs that only occur in the final operating environment. This dissertation investigates techniques for observing HLS circuits, allowing designers to debug the circuit in the context of the original source code, while it executes at-speed in the normal operating environment. This dissertation is comprised of four major contributions toward this goal. First, we develop a debugging framework that provides users with a basic software-like debug experience, including single- stepping, breakpoints and variable inspection. This is accomplished by automatically inserting special- ized debug instrumentation into the user’s circuit, allowing an external debugger to observe the circuit. Debugging at-speed is made possible by recording circuit execution in on-chip memories and retrieving the data for offline analysis.
    [Show full text]
  • The GENCOD Project: Automated Generation of Hardware Code For
    The GENCOD project : Automated generation of Hardware code for safety critical applications on FPGA targets Pascal Pampagnin, Ludovic Letellier To cite this version: Pascal Pampagnin, Ludovic Letellier. The GENCOD project : Automated generation of Hardware code for safety critical applications on FPGA targets. ERTS2 2010, Embedded Real Time Software & Systems, May 2010, Toulouse, France. hal-02264384 HAL Id: hal-02264384 https://hal.archives-ouvertes.fr/hal-02264384 Submitted on 6 Aug 2019 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. The GENCOD project : Automated generation of Hardware code for safety critical applications on FPGA targets Pascal Pampagnin1, Ludovic Letellier2, 1: Airbus Operations SAS, 316 route de Bayonne, 31060 Toulouse, France 2: Alyotech Innovations, 22, bd Déodat de Séverac, 31770 Colomiers, France Abstract: GENCOD is a research project for year ‘80 ’90 ‘00 ‘05 ‘10 solutions to automated generation of safe code for µP 286 Intel Intel ® Intel ® Intel ® Field Programmable Gate Arrays (FPGA) targets. 486 Pentium Itanium Itanium The paper will describe typical ASIC/FPGA workflow, ™ ® 4 ® 2 ® i7 and current implementation for airborne electronic Transi 104 106 42.106 41.107 2.109 hardware design.
    [Show full text]
  • Implementation of Fpga-Based Object Tracking Algorithm
    IMPLEMENTATION OF FPGA-BASED OBJECT TRACKING ALGORITHM A PROJECT REPORT Submitted by G. SHRIKANTH (21904106079) KAUSHIK SUBRAMANIAN (21904106043) in partial fulfillment for the award of the degree of BACHELOR OF ENGINEERING In ELECTRONICS AND COMMUNICATION ENGINEERING SRI VENKATESWARA COLLEGE OF ENGINEERING, SRIPERUMBUDUR ANNA UNIVERSITY: CHENNAI 600 025 APRIL 2008 1 ANNA UNIVERSITY: CHENNAI 600 025 BONAFIDE CERTIFICATE Certified that this project report “IMPLEMENTATION OF FPGA-BASED OBJECT TRACKING ALGORITHM” is the bonafide work of “KAUSHIK SUBRAMANIAN (21904106043) AND G. SHRIKANTH (21904106079)” who carried out the project work under my supervision . SIGNATURE SIGNATURE Prof. R. Narayanan Mr. N. Venkateswaran Head of the Department SUPERVISOR Department of Electronics and Assistant Professor Communication Engineering Department of Electronics and Sri Venkateswara College of Communication Engineering Engineering, Pennalur, Sri Venkateswara College of Sriperumbudur - 602105 Engineering, Pennalur, Sriperumbudur - 602105 EXTERNAL INTERNAL EXAMINAR EXAMINAR 2 ACKNOWLEDGEMENT We are personally indebted to a number of people who gave us their useful insights to aid in our overall progress for this project. A complete acknowledgement would therefore be encyclopedic. First of all, we would like to give our deepest gratitude to our parents for permitting us to take up this course. Our sincere thanks and heartfelt sense of gratitude goes to our respected Principal, Dr. R. Ramachandran for all his efforts and administration in educating us in his premiere institution. We take this opportunity to also thank our Head of the Department, Prof. R. Narayanan for his encouragement throughout the project. We would like to express our gratitude to our Internal Coordinator, Prof. Ganesh Vaidyanathan for his commendable support and encouragement for the completion of our project with perfection.
    [Show full text]
  • Hardware Acceleration
    CprE 488 – Embedded Systems Design Lecture 8 – Hardware Acceleration Joseph Zambreno Electrical and Computer Engineering Iowa State University www.ece.iastate.edu/~zambreno rcl.ece.iastate.edu First, solve the problem. Then, write the code. – John Johnson Motivation: Moore’s Law • Every two years: – Double the number of transistors – Build higher performance general-purpose processors • Make the transistors available to the masses • Increase performance (1.8×↑) • Lower the cost of computing (1.8×↓) • Sounds great, what’s the catch? Gordon Moore Zambreno, Spring 2017 © ISU CprE 488 (Hardware Acceleration) Lect-08.2 Motivation: Moore’s Law (cont.) • The “catch” – powering the transistors without melting the chip! 10,000,000,000 2,200,000,000 1,000,000,000 Chip Transistor 100,000,000 Count 10,000,000 1,000,000 100,000 10,000 2300 1,000 130W 100 10 0.5W 1 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Zambreno, Spring 2017 © ISU CprE 488 (Hardware Acceleration) Lect-08.3 Motivation: Dennard Scaling • As transistors get smaller their power density stays constant Transistor: 2D Voltage-Controlled Switch Dimensions Voltage ×0.7 Doping Robert Dennard Concentrations Area 0.5×↓ Capacitance 0.7×↓ Frequency 1.4×↑ Power = Capacitance × Frequency × Voltage2 Power 0.5×↓ Zambreno, Spring 2017 © ISU CprE 488 (Hardware Acceleration) Lect-08.4 Motivation Dennard Scaling (cont.) • In mid 2000s, Dennard scaling “broke” Transistor: 2D Voltage-Controlled Switch Dimensions Voltage ×0.7 Doping Concentrations Area 0.5×↓ Capacitance 0.7×↓ Frequency 1.4×↑ Power
    [Show full text]
  • Making Domain-Specific Hardware Synthesis Tools Cost-Efficient
    Making Domain-Specific Hardware Synthesis Tools Cost-Efficient Nithin George∗ David Novo∗ Tiark Rompf†‡ Martin Odersky† Paolo Ienne∗ ∗Processor Architecture Laboratory, †Programming Methods Laboratory, ‡Oracle Labs, École Polytechnique Fédérale de Lausanne, École Polytechnique Fédérale de Lausanne, Email: fi[email protected] Email: first.last@epfl.ch Email: first.last@epfl.ch Abstract—Tools to design hardware at a high level of ab- Application domain experts, who often have little knowledge straction promise software-like productivity for hardware de- of designing hardware, can use these domain-specific tools signs. Among them, tools like Spiral, HDL Coder, Optimus and effectively to create designs targeting FPGAs. These tools can, MMAlpha target specific application domains and produce highly hence, make the benefits of reconfigurable technology more efficient implementations from high-level input specifications in accessible to users within different domains. a Domain Specific Language (DSL). But, developing similar domain-specific High-Level Synthesis (HLS) tools need enormous However, developing a new domain-specific HLS tool in- effort, which might offset their many advantages. In this paper, curs enormous effort since one would need to design a new we propose a novel, cost-effective approach to develop domain- DSL, a custom compiler and, sometimes, even a development specific HLS tools. We develop the HLS tool by embedding environment; developing just the compiler itself would involve its input DSL in Scala and using Lightweight Modular Staging writing a parser, multiple analysis and optimization steps and the (LMS), a compiler framework written in Scala, to perform output code generators. This high design effort makes such tools optimizations at different abstraction levels.
    [Show full text]