TranSAGE: Automating Spatial Accelerator Generation for Large-Scale Transformer Models
Open Access
- 10 April 2027
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 970-988
Abstract
No abstract availableFunding Information
- National Natural Science Foundation of China (62622214)
- National Natural Science Foundation of China (62472273)
This publication has 38 references indexed in Scilit:
- Stream-HLS: Towards Automatic Dataflow Acceleration2025
- Stream: Design Space Exploration of Layer-Fused DNNs on Heterogeneous Dataflow AcceleratorsIEEE Transactions on Computers, 2024
- PipeFuser: Building Flexible Pipeline Architecture for DNN Accelerators via Layer Fusion2024
- Automatic Generation of Spatial Accelerator for Tensor AlgebraIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2022
- High-Resolution Image Synthesis with Latent Diffusion Models2022
- DNNFusion: accelerating deep neural networks execution with advanced operator fusion2021
- Understanding Reuse, Performance, and Hardware Cost of DNN Dataflow2019
- Spatial: a language and compiler for application accelerators2018
- Eyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural NetworksIEEE Journal of Solid-State Circuits, 2016
- ShiDianNaoACM SIGARCH Computer Architecture News, 2015