{"id":16833308,"url":"https://github.com/rejunity/tt05-spiking-neural-net","last_synced_at":"2026-01-03T15:50:19.925Z","repository":{"id":205127914,"uuid":"712425086","full_name":"rejunity/tt05-spiking-neural-net","owner":"rejunity","description":null,"archived":false,"fork":false,"pushed_at":"2024-08-28T21:46:13.000Z","size":19264,"stargazers_count":4,"open_issues_count":0,"forks_count":2,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-03-15T15:54:35.447Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Verilog","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/rejunity.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-10-31T12:55:49.000Z","updated_at":"2024-08-29T18:51:30.000Z","dependencies_parsed_at":"2023-11-07T03:23:19.056Z","dependency_job_id":"1677c7f3-302c-4b0a-a802-b9024e8e8bcb","html_url":"https://github.com/rejunity/tt05-spiking-neural-net","commit_stats":null,"previous_names":["rejunity/tt05-spiking-neural-net"],"tags_count":0,"template":false,"template_full_name":"TinyTapeout/tt05-verilog-demo","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rejunity%2Ftt05-spiking-neural-net","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rejunity%2Ftt05-spiking-neural-net/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rejunity%2Ftt05-spiking-neural-net/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rejunity%2Ftt05-spiking-neural-net/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/rejunity","download_url":"https://codeload.github.com/rejunity/tt05-spiking-neural-net/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244135905,"owners_count":20403797,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-10-13T11:53:03.616Z","updated_at":"2026-01-03T15:50:19.880Z","avatar_url":"https://github.com/rejunity.png","language":"Verilog","funding_links":[],"categories":[],"sub_categories":[],"readme":"![](../../workflows/gds/badge.svg) ![](../../workflows/docs/badge.svg) ![](../../workflows/test/badge.svg)\n\n# Binarized Neural Network On-Chip at Telluride Neuromorphic Workshop'23\n\n**\"The Huge\"** Binarized Neural Network On-Chip was developed during the [Telluride Neuromorphic Workshop 2023](https://sites.google.com/view/telluride-2023/home) as a project for the [OSN23: Open-Source Neuromorphic Hardware, Software and Wetware](https://sites.google.com/view/telluride-2023/topic-areas/osn23-open-source-neuromorphic-hardware-software-and-wetware) topic.\n\n**NOTE: To our knowledge this is the first Telluride project that resulted in production of a physical silicon chip!**\n\n## Design goals\nThe design goals behind this project:\n1) place whole neural network on-chip in a brain inspired manner\n2) take advantage of connection sparsity\n3) eliminate communication with external memory (low power)\n4) use [fully-open source](https://openroad.readthedocs.io/en/latest/) design process\n5) **fabricate the actual silicon chip** for the final testing.\n\n## Results\nSample neural network consisting of 40 neurons and 320 synapses fits in less than 1 square millimiter on [Sky130 nm](https://skywater-pdk.readthedocs.io/en/main/) process and uses less than 10K logic gates.\n\n## ASIC tapeout\n\n**\"The Huge\"** Binarized Neural Network On-Chip was tapedout on [FOSS 130nm Production process](https://skywater-pdk.readthedocs.io/en/main/) via [Tiny Tapeout](https://tinytapeout.com/runs/tt05/) initiative.\n\nThe project [#582](https://tinytapeout.com/runs/tt05/582/).\n\n\u003cp align=\"center\" width=\"100%\"\u003e\n  \u003cimg width=\"30%\" src=\"./tt05_full_gds.png\"\u003e\n  \u003cimg width=\"30%\" src=\"./tt05_logic_density.png\"\u003e\n\u003c/p\u003e\n\n## \nThe standalone test for a Binarized Leaky Integrate and Fire (BLIF) neuron can be found in [https://github.com/rejunity/tt04-LIF-neuron-telluride2023]([https://github.com/rejunity/tt04-LIF-neuron-telluride2023]) and it was tapedout with [Tiny Tapeout 4 / CI2309](https://tinytapeout.com/runs/tt05/) shuttle.\n\n# The team\n  - Dr. Paola Vitolo\n  - Dr. Andrew Wabnitz\n  - ReJ aka Renaldas Zioma\n\n# Topic Leader and Tapeout Sponsor\n  - [Prof. Jason Eshraghian](https://ncg.ucsc.edu/jason-eshraghian-bio/)\n\n\n# History of optimisation from older to newer commit\n#### 16 tiles\n```\n  * 25.88% 243085um  7918 cells 544 dff, 23.26 min gds, 13.55 viewer    \u003c- 384 synapses (16) x 16 x 8\n  * 40.98% 393467um 12384 cells 800 dff, 19.59 min gds,                 \u003c- 640 synapses (16) x 16 x 16 x 8\n  * 43.67% 432050um 12932 cells 928 dff, 29.0  min gds, 47.26 viewer    \u003c- 640 synapses (16) x 16 x 16 x 8 fixed the weights\n  * 45.23% 427921um 13795 cells 928 dff, 26.5  min gds                  \u003c- 640 synapses (16) x 16 x 16 x 8 **BN added**\n  * 17.09% 100777um  5116 cells 808 dff, 15.19 min gds                  \u003c- 320 synapses (16) x 16 x 16 x 8 **50% sparsity!**\n  * 24.72% 183035um  7957 cells 968 dff, 13.22 min gds                  \u003c- 320 synapses (16) x 16 x 16 x 8 **BN scale per neuron**, 50% sparsity!\n```\n#### 8 tiles\n```\n  * 49.81%, 185466um 7977 cells 968 dff, 15.45 min gds                  \u003c- 320 synapses (16) x 16 x 16 x 8 BN scale per neuron, 50% sparsity!\n  * 59.47%, 233359um 9486 cells 1128dff, 16.30 min gds                  \u003c- 320 synapses (16) x 16 x 16 x 8 **BN scale+add** per neuron, 50% sparsity!\n  * 62.84%, 213588um 9624 cells 1142dff, 15.42 min gds                  \u003c- 320 synapses (16) x 16 x 16 x 8 **threshollds** per layer, BN per neuron, 50% sparsity!\n```\n\n# What is Tiny Tapeout?\n\nTinyTapeout is an educational project that aims to make it easier and cheaper than ever to get your digital designs manufactured on a real chip.\n\nTo learn more and get started, visit https://tinytapeout.com.\n\n## Resources\n\n- [FAQ](https://tinytapeout.com/faq/)\n- [Digital design lessons](https://tinytapeout.com/digital_design/)\n- [Learn how semiconductors work](https://tinytapeout.com/siliwiz/)\n- [Join the community](https://discord.gg/rPK2nSjxy8)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frejunity%2Ftt05-spiking-neural-net","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frejunity%2Ftt05-spiking-neural-net","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frejunity%2Ftt05-spiking-neural-net/lists"}