{"slug": "splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators", "title": "Splice-Core: Low-Order-Bit-Splicing GEMM for Energy-Efficient AI Accelerators", "summary": "A November 19, 2026 arXiv paper introduces Splice-Core, a temporal-unary/binary hybrid GEMM unit for AI inference accelerators whose two processing-element variants, Splice-BP and Splice-LUT, cut geometric-mean tile energy by 77.4% and 76.4% respectively versus a binary systolic array across 26 INT4 CNN and LLM workloads in 45nm. Splice-LUT also raises geometric-mean tile throughput by 44.4% and reduces geometric-mean energy-delay product by 6.13x across the 26 configurations, while the pair cut tile energy by 77.3-88.1% on W8A8 convolution and expert-projection workloads. Post-place-and-route INT4 results show 14% and 9% smaller area for Splice-BP and Splice-LUT than the systolic array.", "body_md": "# Computer Science > Hardware Architecture\n\n  [Submitted on 19 Nov 2026]\n\n# Title:Splice-Core: Low-Order-Bit-Splicing GEMM for Highly Energy-Efficient AI Inference Accelerators\n\n[View PDF](https://arxiv.org/pdf/2611.25617)\n\n[HTML (experimental)](https://arxiv.org/html/2611.25617v1)\n\nAbstract:Splice-Core is a temporal-unary/binary (tub) hybrid general matrix multiplication (GEMM) unit that minimizes low-order accumulation latency, while retaining value-dependent execution for high-order activation bits. We present two novel processing element (PE) variants: (i) Splice-BP utilizing bit planes, and (ii) Splice-LUT, which selects a shared weight multiple in a single cycle. Compared to a binary systolic array, Splice-BP/Splice-LUT achieve 77.4%/76.4% geometric-mean tile energy reductions, respectively, across 26 INT4 convolutional neural network (CNN) and large language model (LLM) workloads in 45nm. Splice-LUT further increases geometric-mean tile throughput by 44.4% and reduces geometric-mean energy-delay product (EDP) by 6.13x across the 26 configurations. Across the evaluated W8A8 convolution and expert-projection workloads, Splice-BP and Splice-LUT reduce tile energy by 77.3-88.1%. Post-place-and-route INT4 yields 14%/9% smaller area for Splice-BP/Splice-LUT, compared to the systolic array.\n    \n\n### Additional Features\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators", "canonical_source": "https://arxiv.org/abs/2611.25617", "published_at": "2026-11-21 05:31:30+00:00", "updated_at": "2026-11-21 05:43:17.191585+00:00", "lang": "en", "topics": ["ai-chips", "ai-infrastructure", "machine-learning", "large-language-models"], "entities": ["Splice-Core", "Splice-BP", "Splice-LUT", "arXiv"], "also_reported_by": [], "takeaway": "Splice-LUT's single-cycle shared-weight-multiple selection delivers 76.4% lower geometric-mean tile energy, 44.4% higher throughput and a 6.13x lower energy-delay product than a binary systolic array across 26 INT4 CNN and LLM workloads in 45nm.", "recommendation": {"advice": "For energy-constrained INT4 inference accelerators, splice low-order bit accumulation with value-dependent high-order execution and a shared-weight LUT PE; the paper measured 76.4% lower tile energy and 6.13x lower EDP than a binary systolic array.", "playbook": "inference-serving", "playbook_url": "https://wpnews.pro/playbooks/inference-serving"}, "alternates": {"html": "https://wpnews.pro/news/splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators", "markdown": "https://wpnews.pro/news/splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators.md", "text": "https://wpnews.pro/news/splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators.txt", "jsonld": "https://wpnews.pro/news/splice-core-low-order-bit-splicing-gemm-for-energy-efficient-ai-accelerators.jsonld"}}