Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 최정욱 | - |
dc.date.accessioned | 2020-09-24T01:43:10Z | - |
dc.date.available | 2020-09-24T01:43:10Z | - |
dc.date.issued | 2019-09 | - |
dc.identifier.citation | IEEE MICRO, v. 39, no. 5, Page. 102-111 | en_US |
dc.identifier.issn | 0272-1732 | - |
dc.identifier.issn | 1937-4143 | - |
dc.identifier.uri | https://ieeexplore.ieee.org/document/8782645 | - |
dc.identifier.uri | https://repository.hanyang.ac.kr/handle/20.500.11754/154102 | - |
dc.description.abstract | The ubiquitous adoption of systems specialized for AI requires bridging two seemingly conflicting challenges-the need to deliver extreme processing efficiencies while employing familiar programming interfaces, making them compelling even for nonexpert users. We take a significant first step towards this goal and present an end-to-end software stack for the RAPID AI accelerator developed by IBM Research. We present a set of software extensions, called DEEPTOOLS, that leverage and work within popular deep learning frameworks. DEEPTOOLS requires no additional user input and enables aggressive, accelerator-specific performance optimization akin to a full, custom framework. DEEPTOOLS has two key components: 1) a compiler runtime called DeepRT, which automatically identifies how best to execute a given DNN graph on RAPID and constructs the requisite program binaries; and 2) an execution runtime called RAPiDLiB, which triggers and manages the execution of compute and data-transfer operations on RAPID. We integrate DEEPTOOLS with TensorFlow and map popular DNNs (AlexNet, VGG, ResNet, LSTM) to RAPID. We demonstrate substantial improvement in performance over hand-tuned mappings. | en_US |
dc.language.iso | en | en_US |
dc.publisher | IEEE COMPUTER SOC | en_US |
dc.subject | Deep Learning | en_US |
dc.subject | Machine learning accelerators | en_US |
dc.subject | Software stack for AI | en_US |
dc.title | DEEPTOOLS: Compiler and Execution Runtime Extensions for RAPiD AI Accelerator | en_US |
dc.type | Article | en_US |
dc.relation.no | 5 | - |
dc.relation.volume | 39 | - |
dc.identifier.doi | 10.1109/MM.2019.2931584 | - |
dc.relation.page | 102-111 | - |
dc.relation.journal | IEEE MICRO | - |
dc.contributor.googleauthor | Venkataramani, Swagath | - |
dc.contributor.googleauthor | Choi, Jungwook | - |
dc.contributor.googleauthor | Srinivasan, Vijayalakshmi | - |
dc.contributor.googleauthor | Wang, Wei | - |
dc.contributor.googleauthor | Zhang, Jintao | - |
dc.contributor.googleauthor | Schaal, Marcel | - |
dc.contributor.googleauthor | Serrano, Mauricio J. | - |
dc.contributor.googleauthor | Ishizaki, Kazuaki | - |
dc.contributor.googleauthor | Inoue, Hiroshi | - |
dc.contributor.googleauthor | Ogawa, Eri | - |
dc.relation.code | 2019003506 | - |
dc.sector.campus | S | - |
dc.sector.daehak | COLLEGE OF ENGINEERING[S] | - |
dc.sector.department | DEPARTMENT OF ELECTRONIC ENGINEERING | - |
dc.identifier.pid | choij | - |
dc.identifier.researcherID | AAA-2088-2020 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.