{"title":"Designing Modular Hardware Accelerators in C with ROCCC 2.0","authors":"J. Villarreal, A. Park, W. Najjar, R. Halstead","doi":"10.1109/FCCM.2010.28","DOIUrl":"https://doi.org/10.1109/FCCM.2010.28","url":null,"abstract":"While FPGA-based hardware accelerators have repeatedly been demonstrated as a viable option, their programmability remains a major barrier to their wider acceptance by application code developers. These platforms are typically programmed in a low level hardware description language, a skill not common among application developers and a process that is often tedious and error-prone. Programming FPGAs from high level languages would provide easier integration with software systems as well as open up hardware accelerators to a wider spectrum of application developers. In this paper, we present a major revision to the Riverside Optimizing Compiler for Configurable Circuits (ROCCC) designed to create hardware accelerators from C programs. Novel additions to ROCCC include (1) intuitive modular bottom-up design of circuits from C, and (2) separation of code generation from specific FPGA platforms. The additions we make do not introduce any new syntax to the C code and maintain the high level optimizations from the ROCCC system that generate efficient code. The modular code we support functions identically as software or hardware. Additionally, we enable user control of hardware optimizations such as systolic array generation and temporal common subexpression elimination. We evaluate the quality of the ROCCC 2.0 tool by comparing it to hand-written VHDL code. We show comparable clock frequencies and a 18% higher throughput. The productivity advantages of ROCCC 2.0 is evaluated using the metrics of lines of code and programming time showing an average of 15x improvement over hand-written VHDL.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131237061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Design of a Reconfigurable Hybrid Database System","authors":"B. Scheuermann","doi":"10.1109/FCCM.2010.44","DOIUrl":"https://doi.org/10.1109/FCCM.2010.44","url":null,"abstract":"This paper proposes a design of a database system which accelerates the execution of database transactions by offloading database operators (e.g. joins, scans and sorting) in hardware algorithms executed on runtime reconfigurable computing platforms. Furthermore, a hybrid database system is described which exploits the strengths of the new reconfigurable hardware-based database system in combination with pre-existing technologies for main memory or disc resident database systems. Due to the parallel and pipelined design style of the hardware algorithms, the new database system offers a potential speedup over the traditional sequential execution on instruction stream processors. Moreover, data access times can be reduced by placing circuits on the reconfigurable fabric close to embedded memory which further allows for customizing high bandwidth memory interfaces. As a consequence of the accelerated execution speed, the new system has the potential to reliably meet constraints in real-time scenarios, to reduce the chance of lock contention and cache flushes, and to decrease the cost for concurrency control.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131255724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A Large-Scale Architecture for Restricted Boltzmann Machines","authors":"Sang Kyun Kim, P. McMahon, K. Olukotun","doi":"10.1109/FCCM.2010.38","DOIUrl":"https://doi.org/10.1109/FCCM.2010.38","url":null,"abstract":"Deep Belief Nets (DBNs) are an emerging application in the machine learning domain, which use Restricted Boltzmann Machines (RBMs) as their basic building block. Although small scale DBNs have shown great potential, the computational cost of RBM training has been a major challenge in scaling to large networks. In this paper we present a highly scalable architecture for Deep Belief Net processing on hardware systems that can handle hundreds of boards, if not more, of customized logic with near linear performance increase. We elucidate tradeoffs between flexibility in the neuron connections, and the hardware resources, such as memory and communication bandwidth, required to build a custom processor design that has optimal efficiency. We illustrate how our architecture can easily support sparse networks with dense regions of connections between neighboring sets of neurons, which is relevant to applications where there are obvious spatial correlations in the data, such as in image processing. We demonstrate the feasibility of our approach by implementing a multi-FPGA system. We show that a speedup of 46X-112X over an optimized single core CPU implementation can be achieved for a four-FPGA implementation.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120951823","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A Comparative Study on the Architecture Templates for Dynamic Nested Loops","authors":"J. Cong, Yi Zou","doi":"10.1109/FCCM.2010.45","DOIUrl":"https://doi.org/10.1109/FCCM.2010.45","url":null,"abstract":"Loops are the most typical constructs that the FCCM community tries to accelerate. Many loop constructs have dynamic loop bounds and may face load balancing issues in their parallel realizations. In this paper we discuss different architecture templates for these dynamic nested loops, and show the benefits and trade-offs between various implementation strategies. We show that a statically scheduled architecture may perform better if the overhead in banking conflicts overwrites the benefits in load balancing by dynamic scheduled architectures.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114689525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reaping the Processing Potential of FPGA on Double-Precision Floating-Point Operations: An Eigenvalue Solver Case Study","authors":"Miaoqing Huang, Özlem Kilic","doi":"10.1109/FCCM.2010.23","DOIUrl":"https://doi.org/10.1109/FCCM.2010.23","url":null,"abstract":"Many scientific applications such as electromagnatics require their operations carried out in double-precision floating-point format. The efficiency of these applications is mainly subject to the floating-point processing performance on the target processors. In this work, we use an eigenvalue solver application as a case study to demonstrate the processing potential of an FPGA device when dealing with floating-point operations. Relying on deep pipelines and large local memory directly accessible within the FPGA device, more than 20$times$ performance improvement has been achieved for this particular case whose computation is intrinsically sequential. The methodology described in this paper can be conveniently applied to other domain applications that share similar data processing characteristics, demonstrating an impact on a broad range of scientific fields, e.g., numerical linear algebra.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115516344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Automated Precision Analysis: A Polynomial Algebraic Approach","authors":"D. Boland, G. Constantinides","doi":"10.1109/FCCM.2010.32","DOIUrl":"https://doi.org/10.1109/FCCM.2010.32","url":null,"abstract":"When migrating an algorithm onto hardware, the potential saving that can be obtained by tuning the precision used in the algorithm to meet a range or error specification is often overlooked; the major reason is that it is hard to choose a number system which can guarantee any such specification can be met. Instead, the problem is mitigated by opting to use IEEE standard single or double precision so as to be ‘no worse’ than a software implementation. However, the flexibility in the number representation is one of the key factors that can only be exploited on FPGAs, unlike GPUs and general purpose processors, and hence ignoring this potential significantly limits the performance achievable on an FPGA. To this end, this paper describes a tool which analyses algorithms with given input ranges under a finite precision to provide information that could be used to tune the hardware to the algorithm specifications. We demonstrate the proposed procedure on an iteration of the conjugate gradient algorithm, achieving a reduction in slices of over 40% when meeting the same error specification found by traditional methods. We also show it achieves comparable bounds to recent literature in a small fraction of the execution time, with greater scalability.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"79 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123143564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Performing Floating-Point Accumulation on a Modern FPGA in Single and Double Precision","authors":"T. O. Bachir, J. David","doi":"10.1109/FCCM.2010.24","DOIUrl":"https://doi.org/10.1109/FCCM.2010.24","url":null,"abstract":"In this paper, we discuss the feasibility of a floating-point accumulator (FPACC) on modern high-end FPGA devices. We explore different implementation scenarios and propose new FPACC architectures for both single and double precision floating-point addends. The proposed strategies can be easily adapted to the implement a multiply-accumulator (FPMAC), with one or two rounding stages, in both single and double precision as well. All the aforementioned designs are characterized by high operating frequencies (ranging from 130 to 300 MHz) and moderate occupation area (from 300 to 800 slices) when implemented on the VC5VSX50T FPGA, an entry level Virtex 5 from Xilinx.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"137 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131377770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Junguk Cho, Bridget Benson, Sunsern Cheamanunkul, R. Kastner
{"title":"Increased Performace of FPGA-Based Color Classification System","authors":"Junguk Cho, Bridget Benson, Sunsern Cheamanunkul, R. Kastner","doi":"10.1109/FCCM.2010.50","DOIUrl":"https://doi.org/10.1109/FCCM.2010.50","url":null,"abstract":"This paper presents a hardware architecture for increased performance of color classification. In our architecture, color classification, based on an AdaBoost algorithm, identifies a pixel as having the color of interest or not. We designed the proposed architecture using Verilog HDL and implemented the design in a Xilinx Virtex-5 FPGA. The architecture for color classification can have 598 times performance improvement over an equivalent software solution and 1.9 times performance improvement over the leading hardware color classifier.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123165737","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A TDMA Ethernet Switch for Dynamic Real-Time Communication","authors":"G. Carvajal, S. Fischmeister","doi":"10.1109/FCCM.2010.27","DOIUrl":"https://doi.org/10.1109/FCCM.2010.27","url":null,"abstract":"A real-time communication medium must provide a special coordination mechanism to guarantee bounded communication delays. Implementing this mechanism in software offers flexibility but reduces reliability and performance. On the other hand, customized hardware solutions deliver high throughput and predictability, but they increase the implementation cost and are unable to adapt to the specific needs of individual applications. In this work, we introduce a switch that implements a programmable dedicated time-triggered packet switching mechanism on top of Ethernet. The switch, called the Network Code Switch bases on the NetFPGA system and executes flexible but verifiable state-based schedules encoded in the Network Code programming language. This permits the user to tailor the communication behavior to the needs of the distributed application with verifiable performance. We discuss our experience starting at the designing to the implementation of the prototype, and describe how we exploited modularity and code reutilization to reduce the implementation costs and increase the flexibility of the architecture. We also validate our design by evaluating the overhead and throughput of the implemented prototype.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133637358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A High-Speed and Memory Efficient Pipeline Architecture for Packet Classification","authors":"Yeim-Kuan Chang, Yi-Shang Lin, Cheng-Chien Su","doi":"10.1109/FCCM.2010.40","DOIUrl":"https://doi.org/10.1109/FCCM.2010.40","url":null,"abstract":"Multi-field Packet classification is the main function in high-performance routers. The current router design goal of achieving a throughput higher than 40 Gbps and supporting large rule sets simultaneously is difficult to be fulfilled by software approaches. In this paper, a set pruning trie based pipelined architecture called Set Pruning Multi-Bit Trie (SPMT) is proposed for multi-field packet classification. However, the problem of rule duplications in SPMT that may cause a memory blowup must be solved in order to implement SPMT with large rule sets in FPGA devices consisting of limited on-chip memory. We will propose two rule grouping schemes to reduce rule duplications in SPMT. The first scheme called Partition by Wildcards (PW) divides the rules into subgroups based on the positions of their wildcard fields. The second scheme called Partition by Length (PL) rules partitions the rules into subgroups according to their prefix lengths. Based on our performance experiments on Xilinx Virtex-5 FPGA device, the proposed pipeline architecture can achieve a throughput of over 100 Gbps with dual port memory. Also, the rule sets of up to 10k rules can be fit into the on-chip memory of Xilinx Virtex-5 FPGA device.","PeriodicalId":185657,"journal":{"name":"2010 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines","volume":"48 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116306915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}