Akihiro Goda, Yukio Miyasaka, A. M. Gharehbaghi, M. Fujita
{"title":"Synthesis and Generalization of Parallel Algorithms Considering Communication Constraints","authors":"Akihiro Goda, Yukio Miyasaka, A. M. Gharehbaghi, M. Fujita","doi":"10.1109/ISQED48828.2020.9137022","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9137022","url":null,"abstract":"Recently, the opportunities of parallel computing are expanding rapidly in various applications including neural networks and machine learning. It is, however, not at all straightforward to develop an efficient algorithm for each parallel computing environment since communications always introduce overhead in computation. In this paper, we propose a design method of optimum parallel computing under user-specified communication constraints. The basic strategy is to automatically generate optimum scheduling from small instances of the target problem and then they are semi-automatically generalized to much larger problems. Several experiments targeting matrix vector multiplication and convolutional neural networks have been conducted. Their results show the correctness and usefulness of the proposed method as well as its scalability.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124621879","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":"Secure, Scalable and Low-Power Junction Temperature Sensing for Multi-Processor Systems-on-Chip","authors":"G. A. Kumar","doi":"10.1109/ISQED48828.2020.9136996","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9136996","url":null,"abstract":"The measurement of silicon junction temperature is often required in embedded applications for monitoring and control, device protection and different types of temperature based compensation. The systems-on-chip (SoC) are growing complex these days with multiple processors and high degree of peripheral integration. Also the power consumption of the SoC requires being very low for battery operated or energy harvesting based embedded applications like Internet of Things (IoT). This paper describes the architecture, integration, programming model and working scheme of a silicon junction temperature sensing module that achieves security, scalability and low-power requirements for multi-processor SoCs.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127116441","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}
Shujuan Yin, Zheyu Liu, Guihong Li, F. Qiao, Qi Wei, Yuanfeng Wu, Lianru Gao, Xinjun Liu, Huazhong Yang
{"title":"RARA: Dataflow Based Error Compensation Methods with Runtime Accuracy-Reconfigurable Adder","authors":"Shujuan Yin, Zheyu Liu, Guihong Li, F. Qiao, Qi Wei, Yuanfeng Wu, Lianru Gao, Xinjun Liu, Huazhong Yang","doi":"10.1109/ISQED48828.2020.9136984","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9136984","url":null,"abstract":"The promulgation of Internet-of-Things technologies requires higher energy efficiency than the past. Approximate computing is a promising computation paradigm in the post-Moore era. It seeks a subtle balance between computation accuracy and many other metrics, especially power consumption. Thereinto, approximate adders are important because addition is the essential operation in most applications. In this article, we propose a runtime accuracy reconfigurable adder with the accurate mode and 16 approximate modes with different accuracy. The statistical error model of the adder is built on two typical dataflow graphs: the adder chain and adder tree. Then, we introduce two methods to compensate for the accuracy loss based on the error model: Input Gating and Dataflow Reorganization. Our proposed adder achieves higher configuration flexibility with much less area overhead. The experiment results show our methods can make average output error reduce up to 61% without energy cost.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127169402","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}
Abhijitt Dhavlle, Raj Mehta, S. Rafatirad, H. Homayoun, Sai Manoj Pudukotai Dinakarrao
{"title":"Entropy-Shield:Side-Channel Entropy Maximization for Timing-based Side-Channel Attacks","authors":"Abhijitt Dhavlle, Raj Mehta, S. Rafatirad, H. Homayoun, Sai Manoj Pudukotai Dinakarrao","doi":"10.1109/ISQED48828.2020.9137008","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9137008","url":null,"abstract":"The hardware systems have experienced a plethora of side-channel attacks (SCAs) in recent years with cache-based SCAs being one of the dominant threats. The SCAs exploit the architectural caveats, which invariably leak essential information during an application's execution. Shutting down the side-channels is not a feasible approach due to various restrictions, such as architectural changes and complexity. To overcome such concerns and protect the data integrity, we introduce Entropy-Shield in this work. The proposed Entropy-Shield aims to maximize the entropy in the leaked side-channel information rather than attempting to close the side-channels. To achieve this, the proposed Entropy-Shield introduces carefully and sensibly crafted perturbations into the victim application, thereby increasing the entropy of the information obtained by the attacker to deduce the secret key, while the information being observed looks legit yet futile. This methodology has been successfully tested on cache targeted SCAs such as Flush+Reload and Flush+Flush and the key information retrieved by the attacker is shown to be ultimately futile, indicating the success of proposed Entropy-Shield.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127638028","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":"Saving Time and Energy Using Partial Flash Memory Operations in Low-Power Microcontrollers","authors":"Prawar Poudel, A. Milenković","doi":"10.1109/ISQED48828.2020.9137034","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9137034","url":null,"abstract":"This paper introduces a technique that reduces time and energy consumed by critical flash memory operations in ultra-low-power microcontrollers. The proposed technique utilizes partial or aborted flash memory erase and program operations that proved to have no negative impacts on accuracy and longevity of information stored in the flash memory. Our experimental evaluation performed on a family of microcontrollers shows that the proposed technique can save 98% of the energy consumed for flash erase operations and up to 75% for flash program operations.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126303294","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":"Signal Selection Heuristics for Post-Silicon Validation","authors":"Suprajaa Tummala, Xiaobang Liu, R. Vemuri","doi":"10.1109/ISQED48828.2020.9137037","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9137037","url":null,"abstract":"Limited observability is the key challenge in post-silicon validation and can be alleviated by using an on-chip trace buffer which monitors and captures the response of certain selected signals during run-time. Use of on-chip trace buffers for debug introduces area overhead along with increased power consumption. This imposes a constraint on the number of signals selected to be traced. In this work, we review some trace signal selection algorithms proposed in the literature and propose several new heuristics. Given the constraints on the width (number of signals traced) and depth (number of cycles over which they are traced), these algorithms attempt to select the best set of signals to maximize the state restoration ratio. We evaluate the quality and performance of these heuristic signal selection algorithms while using two different techniques for state restoration: the forward propagation and backward justification (FB) method and the satisfiability (SAT) based method.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"767 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116137173","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":"Spintronics-based Reconfigurable Ising Model Architecture","authors":"Ankit Mondal, Ankur Srivastava","doi":"10.1109/ISQED48828.2020.9137043","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9137043","url":null,"abstract":"The Ising model has been explored as a framework for modeling NP-hard problems, with several diverse systems proposed to solve it. The Magnetic Tunnel Junction (MTJ)-based Magnetic RAM is capable of replacing CMOS in memory chips. In this paper, we propose the use of MTJs for representing the units of an Ising model and leveraging its intrinsic physics for finding the ground state of the system through annealing. We design the structure of a basic MTJ-based Ising cell capable of performing the functions essential to an Ising solver. A technique to use the basic Ising cell for scaling to large problems is described. We then go on to propose Ising-FPGA, a parallel and reconfigurable architecture that can be used to map a large class of NP-hard problems, and show how a standard Place and Route tool can be utilized to program the Ising-Fpga. The effects of this hardware platform on our proposed design are characterized and methods to overcome these effects are prescribed. We discuss how two representative NP-hard problems can be mapped to the Ising model. Simulation results show the effectiveness of MTJs as Ising units by producing solutions close/comparable to the optimum, and demonstrate that our design methodology holds the capability to account for the effects of the hardware.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134336095","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":"Layout Capacitance Extraction Using Automatic Pre-Characterization and Machine Learning","authors":"Zhixing Li, Weiping Shi","doi":"10.1109/ISQED48828.2020.9136970","DOIUrl":"https://doi.org/10.1109/ISQED48828.2020.9136970","url":null,"abstract":"The most widely adopted approach for layout capacitance extraction is pattern matching, where a layout is partitioned into small sections and matched against a pre-characterized pattern library. The capacitance formula associated with the matched pattern is used to compute the capacitance. The pattern matching approach can meet the speed and capacity requirements for large ASICs, but the pattern library and associated formula must be manually created by experienced engineers through a tedious process. For each new process node or any change to the technology profile, this process has to be repeated, causing a heavy burden on the EDA vendors and delays in PDK releases by the foundries. In this paper, we propose an automatic approach that significantly reduces the effort and time compared with the manual approach for the pattern matching method for capacitance extraction. Our approach consists of the following steps: 1) generate sample geometries that cover the problem space, 2) for each sample geometry, call a field solver to compute the capacitance value, 3) use unsupervised learning with the field solver values and foundry values as guidance to cluster geometries into patterns and derive the capacitance formula, and 4) use supervised learning to train a neural network that will be used to classify any layout geometry into a pattern in the library, which will be used at the extraction time. Experiment results show the proposed approach achieves satisfactory accuracy well above the foundry requirements and runs as fast as traditional pattern matching software. Although our approach is only demonstrated for 2D, it is the first framework using the automatic procedure and machine learning in capacitance extraction. As the interconnect structures and foundry requirements become more complex, our approach may be attractive to EDA vendors and foundries.","PeriodicalId":225828,"journal":{"name":"2020 21st International Symposium on Quality Electronic Design (ISQED)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114168039","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}