{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,21]],"date-time":"2025-11-21T18:10:27Z","timestamp":1763748627670,"version":"3.28.0"},"reference-count":68,"publisher":"IEEE","license":[{"start":{"date-parts":[[2022,5,15]],"date-time":"2022-05-15T00:00:00Z","timestamp":1652572800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2022,5,15]],"date-time":"2022-05-15T00:00:00Z","timestamp":1652572800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2022,5,15]]},"DOI":"10.1109\/fccm53951.2022.9786164","type":"proceedings-article","created":{"date-parts":[[2022,6,3]],"date-time":"2022-06-03T19:48:36Z","timestamp":1654285716000},"page":"1-10","source":"Crossref","is-referenced-by-count":5,"title":["A Generator of Numerically-Tailored and High-Throughput Accelerators for Batched GEMMs"],"prefix":"10.1109","author":[{"given":"Louis","family":"Ledoux","sequence":"first","affiliation":[{"name":"Universitat Politecnica de Catalunya (UPC),Barcelona Supercomputing Center (BSC),Barcelona,Spain"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Marc","family":"Casas","sequence":"additional","affiliation":[{"name":"Universitat Politecnica de Catalunya (UPC),Barcelona Supercomputing Center (BSC),Barcelona,Spain"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"journal-title":"Systolic arrays for (VLSI)","year":"1978","author":"kung","key":"ref39"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/MC.1982.1653825"},{"article-title":"CLARINET: A RISC-V Based Framework for Posit Arithmetic Empiricism","year":"2020","author":"jain","key":"ref33"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.23919\/DATE.2017.7927080"},{"year":"2021","key":"ref31","article-title":"Intel® Xeon® Processor E5 Family Product Specifications"},{"year":"2019","key":"ref30","article-title":"An Introduction to the Intel® QuickPath Interconnect"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1137\/1028001"},{"key":"ref36","volume":"iii","author":"khan","year":"2013","journal-title":"Informatics and Algorithms"},{"article-title":"A Study of BFLOAT16 for Deep Learning Training","year":"2019","author":"kalamkar","key":"ref35"},{"article-title":"Rethinking floating point for deep learning","year":"2018","author":"johnson","key":"ref34"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/FPL.2019.00026"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1145\/3440722.3440880"},{"article-title":"BFloat16: The secret to high performance on Cloud TPUs","year":"0","author":"wang","key":"ref61"},{"article-title":"Benchmarking TPU, GPU, and CPU Platforms for Deep Learning","year":"2019","author":"wang","key":"ref63"},{"article-title":"Reproducible and Accurate Matrix Multiplication for GPU Accelerators","year":"2015","author":"iakymchuk","key":"ref28"},{"year":"2019","key":"ref64","article-title":"Coherent Accelerator Processor Interface"},{"key":"ref27","first-page":"13","article-title":"Floating-Point Design with Vivado HLS","author":"hrica","year":"2012"},{"year":"2020","key":"ref65","article-title":"PCI Express"},{"year":"2016","key":"ref66","article-title":"Virtex UltraScale+"},{"year":"2019","key":"ref29","article-title":"CAPI SNAP Framework Hardware and Software. Contribute to open-power\/snap development by creating an account on GitHub"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.1109\/TCSII.2020.2980531"},{"key":"ref68","doi-asserted-by":"crossref","first-page":"1715","DOI":"10.1103\/PhysRevLett.77.1715","article-title":"Qed corrections of o(mc2?7ln?) to the fine structure splittings of helium and he-like ions","volume":"77","author":"zhang","year":"1996","journal-title":"Physical Review Letters"},{"article-title":"TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems","year":"2016","author":"abadi","key":"ref2"},{"key":"ref1","first-page":"1","article-title":"Ieee standard for binary floating-point arithmetic","year":"1985","journal-title":"ANSI\/IEEE Std 754-1985"},{"article-title":"Gemmini: An Agile Systolic Array Generator Enabling Systematic Evaluations of Deep-Learning Architectures","year":"2019","author":"genc","key":"ref20"},{"year":"0","key":"ref22","article-title":"System Architecture | Cloud TPU"},{"key":"ref21","doi-asserted-by":"crossref","first-page":"399","DOI":"10.1109\/TIT.1966.1053907","article-title":"Run-length encodings (Corresp.)","volume":"12","author":"golomb","year":"2006","journal-title":"IEEE Transactions on Information Theory"},{"key":"ref24","first-page":"16","article-title":"Beating Floating Point at its Own Game: Posit Arithmetic","author":"gustafson","year":"0"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1109\/FCCM.2017.25"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/335231.335253"},{"article-title":"Deep Residual Learning for Image Recognition","year":"2015","author":"he","key":"ref25"},{"article-title":"Power9 To The People","year":"2017","author":"morgan","key":"ref50"},{"key":"ref51","doi-asserted-by":"crossref","first-page":"1578","DOI":"10.1109\/T-C.1971.223174","article-title":"tapered floating point: a new floating-point representation","volume":"c 20","author":"morris","year":"1971","journal-title":"IEEE Transactions on Computers"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1145\/3446210"},{"article-title":"AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference","year":"2020","author":"tambe","key":"ref58"},{"article-title":"Going Deeper with Convolutions","year":"2014","author":"szegedy","key":"ref57"},{"article-title":"BFloat16 extensions for Armv8-A","year":"0","author":"stephens","key":"ref56"},{"key":"ref55","first-page":"15","article-title":"A new matrix multiplication systolic array","author":"quinton","year":"1986"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1109\/TELECOM50385.2020.9299565"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1145\/17356.17388"},{"key":"ref52","first-page":"107","article-title":"A Customizable Matrix Multiplication Framework for the Intel HARPv2 Xeon+FPGA Platform: A Deep Learning Case Study","author":"moss","year":"0"},{"article-title":"Full-Speed Deterministic Bit-Accurate Parallel Floating-Point Summation on Multi-and Many-Core Architectures","year":"2014","author":"collange","key":"ref10"},{"key":"ref11","first-page":"24","article-title":"Stratix V Device Overview","author":"corporation","year":"2020"},{"key":"ref40","doi-asserted-by":"crossref","first-page":"603","DOI":"10.1109\/TC.1987.1676945","article-title":"Optimal Systolic Design for the Transitive Closure and the Shortest Path Problems","author":"kung","year":"1987","journal-title":"IEEE Transactions on Computers"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1145\/3316279.3316285"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/MDT.2011.44"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/FPT.2008.4762363"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3373087.3375296"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1088\/1126-6708\/2009\/01\/012"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.3390\/a12080149"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/1236463.1236468"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/HCS49909.2020.9220557"},{"article-title":"PERC: Posit Enhanced Rocket Chip","year":"2020","author":"arunkumarm","key":"ref4"},{"first-page":"10","article-title":"ADM-PCIE-9V3 Support & Development Kit Release: 1.1.0 V1.1","year":"2018","key":"ref3"},{"article-title":"A Parallel Algorithm for Calculation of Large Determinants with High Accuracy for GPUs and MPI clusters","year":"2013","author":"beliakov","key":"ref6"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1002\/scj.4690190602"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/ICCD.2018.00057"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS47924.2020.00069"},{"article-title":"Opening Up The Server Bus For Coherent Acceleration","year":"2016","author":"morgan","key":"ref49"},{"key":"ref9","first-page":"140","article-title":"Architectural Enhancements in Intel Agilex FPGAs","author":"chromczak","year":"0"},{"article-title":"mightymercado\/PySigmoid","year":"2020","author":"mercado","key":"ref46"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1145\/3190339.3190344"},{"article-title":"Opening Up The Server Bus For Coherent Acceleration","year":"2016","author":"morgan","key":"ref48"},{"article-title":"Big Blue Aims For The Sky With Power9","year":"2016","author":"morgan","key":"ref47"},{"article-title":"From Sir Isaac to the Sloan Survey Calculating the Structure and Chaos Owing to Gravity in the Universe","year":"1996","author":"lake","key":"ref42"},{"key":"ref41","first-page":"15","article-title":"A Systolic 2-1) Convolution Chip","author":"kung\u2019and","year":"0"},{"key":"ref44","first-page":"16","article-title":"Understanding Performance of PCI Express Systems","author":"lawley","year":"2014"},{"article-title":"Accelerated Mathematical Engine Tesla","year":"0","author":"lamert","key":"ref43"}],"event":{"name":"2022 IEEE 30th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM)","start":{"date-parts":[[2022,5,15]]},"location":"New York City, NY, USA","end":{"date-parts":[[2022,5,18]]}},"container-title":["2022 IEEE 30th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/9786061\/9786064\/09786164.pdf?arnumber=9786164","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,6,27]],"date-time":"2022-06-27T21:02:14Z","timestamp":1656363734000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9786164\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,5,15]]},"references-count":68,"URL":"https:\/\/doi.org\/10.1109\/fccm53951.2022.9786164","relation":{},"subject":[],"published":{"date-parts":[[2022,5,15]]}}}