Operator Fusion for LLM Inference on the Tensix Architecture
Mirrored from arXiv — Machine Learning for archival readability. Support the source by reading on the original site.
Computer Science > Machine Learning
Title:Operator Fusion for LLM Inference on the Tensix Architecture
Abstract:This study addresses on-device inference bottlenecks of Transformer models on Tenstorrent's Tensix architecture and proposes an operator fusion strategy that enhances data locality. RMSNorm is fused with matrix multiplication in self-attention and in the FFN, enabling back-to-back execution of memory-bound and compute-bound operators in on-chip SRAM to significantly reduce DRAM reads/writes of intermediate results and scheduling overhead. To support multi-core parallelism, a NoC-based multicast mechanism is leveraged in which row/column master nodes efficiently distribute inputs and weights across the core mesh, alleviating DRAM bandwidth contention. Experiments on the Wormhole platform with Qwen2.5-0.5B, Qwen3-0.6B, and Qwen3-4B show up to 37.44% latency reduction for attention and 15.89% for MLP, with up to 7.91% reduction per decoder layer, while Pearson Correlation Coefficient (PCC) remains above 98.75%, confirming significant end-to-end efficiency gains under numerical consistency.
| Comments: | 11 pages, 5 figures |
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2606.09879 [cs.LG] |
| (or arXiv:2606.09879v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2606.09879
arXiv-issued DOI via DataCite
|
Access Paper:
- View PDF
- HTML (experimental)
- TeX Source
References & Citations
Bookmark
Bibliographic and Citation Tools
Code, Data and Media Associated with this Article
Demos
Recommenders and Search Tools
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both 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.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
More from arXiv — Machine Learning
-
Restless bandits with imperfect binary feedback: PCL-indexability analysis and computation
Jun 11
-
Few-Shot Resampling for Scalable Statistically-Sound Data Mining
Jun 11
-
Physics-informed generative AI for semiconductor manufacturing: Enforcing hard physical constraints in generative models by construction
Jun 11
-
Mechanical Field Networks: Structured Neural Dynamics for Multivariate Systems
Jun 11
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.