{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/eie-efficient-inference-engine-on-compressed","title":"EIE: Efficient Inference Engine on Compressed Deep Neural Network","arxiv_id":"1602.01528","date":"2016-02-04","proceeding":null,"authors":["Song Han","Xingyu Liu","Huizi Mao","Jing Pu","Ardavan Pedram","Mark A. Horowitz","William J. Dally"],"abstract":"State-of-the-art deep neural networks (DNNs) have hundreds of millions of\nconnections and are both computationally and memory intensive, making them\ndifficult to deploy on embedded systems with limited hardware resources and\npower budgets. While custom hardware helps the computation, fetching weights\nfrom DRAM is two orders of magnitude more expensive than ALU operations, and\ndominates the required power.\n  Previously proposed 'Deep Compression' makes it possible to fit large DNNs\n(AlexNet and VGGNet) fully in on-chip SRAM. This compression is achieved by\npruning the redundant connections and having multiple connections share the\nsame weight. We propose an energy efficient inference engine (EIE) that\nperforms inference on this compressed network model and accelerates the\nresulting sparse matrix-vector multiplication with weight sharing. Going from\nDRAM to SRAM gives EIE 120x energy saving; Exploiting sparsity saves 10x;\nWeight sharing gives 8x; Skipping zero activations from ReLU saves another 3x.\nEvaluated on nine DNN benchmarks, EIE is 189x and 13x faster when compared to\nCPU and GPU implementations of the same DNN without compression. EIE has a\nprocessing power of 102GOPS/s working directly on a compressed network,\ncorresponding to 3TOPS/s on an uncompressed network, and processes FC layers of\nAlexNet at 1.88x10^4 frames/sec with a power dissipation of only 600mW. It is\n24,000x and 3,400x more energy efficient than a CPU and GPU respectively.\nCompared with DaDianNao, EIE has 2.9x, 19x and 3x better throughput, energy\nefficiency and area efficiency.","url_abs":"http://arxiv.org/abs/1602.01528v2","url_pdf":"http://arxiv.org/pdf/1602.01528v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"eie-efficient-inference-engine-on-compressed","repo_url":"https://github.com/RalphMao/EIE-simulator","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"eie-efficient-inference-engine-on-compressed","repo_url":"https://github.com/cucapra/fodlam","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"caffe2","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"eie-efficient-inference-engine-on-compressed","repo_url":"https://github.com/songhan/Deep-Compression-AlexNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"caffe2","reach":{"status":"ok","spdx":"BSD-2-Clause"}},{"paper_slug":"eie-efficient-inference-engine-on-compressed","repo_url":"https://github.com/songhan/SqueezeNet-Deep-Compression","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"caffe2","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":null,"task_name":"CPU"},{"task_slug":null,"task_name":"GPU"}],"methods":[{"method_slug":"relu","method_name":"ReLU"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1602.01528","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1602.01528"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/songhan/SqueezeNet-Deep-Compression","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/RalphMao/EIE-simulator","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/cucapra/fodlam","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/songhan/Deep-Compression-AlexNet","reach":{"status":"ok","spdx":"BSD-2-Clause"}}],"summary":{"unverified":3},"by_repo_kind":{"listed":{"samples":3,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"aed98e44716625ad","entry":"layer_costs","repo":"cucapra/fodlam","repo_kind":"listed","path":"fodlam.py","file_url":"https://github.com/cucapra/fodlam/blob/HEAD/fodlam.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"aed98e44716625ad"}},{"code_sha256_prefix":"a1c9b7bced69fd39","entry":"load_net","repo":"cucapra/fodlam","repo_kind":"listed","path":"fodlam.py","file_url":"https://github.com/cucapra/fodlam/blob/HEAD/fodlam.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a1c9b7bced69fd39"}},{"code_sha256_prefix":"f9d32c4968a00fd9","entry":"norm_layer_name","repo":"cucapra/fodlam","repo_kind":"listed","path":"fodlam.py","file_url":"https://github.com/cucapra/fodlam/blob/HEAD/fodlam.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f9d32c4968a00fd9"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}