Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Boxun, Yin, Yuxuan, Iyer, Vikram, Li, Peng
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2505.12281
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918023908032512
author Xu, Boxun
Yin, Yuxuan
Iyer, Vikram
Li, Peng
author_facet Xu, Boxun
Yin, Yuxuan
Iyer, Vikram
Li, Peng
contents We present Bishop, the first dedicated hardware accelerator architecture and HW/SW co-design framework for spiking transformers that optimally represents, manages, and processes spike-based workloads while exploring spatiotemporal sparsity and data reuse. Specifically, we introduce the concept of Token-Time Bundle (TTB), a container that bundles spiking data of a set of tokens over multiple time points. Our heterogeneous accelerator architecture Bishop concurrently processes workload packed in TTBs and explores intra- and inter-bundle multiple-bit weight reuse to significantly reduce memory access. Bishop utilizes a stratifier, a dense core array, and a sparse core array to process MLP blocks and projection layers. The stratifier routes high-density spiking activation workload to the dense core and low-density counterpart to the sparse core, ensuring optimized processing tailored to the given spatiotemporal sparsity level. To further reduce data access and computation, we introduce a novel Bundle Sparsity-Aware (BSA) training pipeline that enhances not only the overall but also structured TTB-level firing sparsity. Moreover, the processing efficiency of self-attention layers is boosted by the proposed Error-Constrained TTB Pruning (ECP), which trims activities in spiking queries, keys, and values both before and after the computation of spiking attention maps with a well-defined error bound. Finally, we design a reconfigurable TTB spiking attention core to efficiently compute spiking attention maps by executing highly simplified "AND" and "Accumulate" operations. On average, Bishop achieves a 5.91x speedup and 6.11x improvement in energy efficiency over previous SNN accelerators, while delivering higher accuracy across multiple datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2505_12281
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Bishop: Sparsified Bundling Spiking Transformers on Heterogeneous Cores with Error-Constrained Pruning
Xu, Boxun
Yin, Yuxuan
Iyer, Vikram
Li, Peng
Neural and Evolutionary Computing
We present Bishop, the first dedicated hardware accelerator architecture and HW/SW co-design framework for spiking transformers that optimally represents, manages, and processes spike-based workloads while exploring spatiotemporal sparsity and data reuse. Specifically, we introduce the concept of Token-Time Bundle (TTB), a container that bundles spiking data of a set of tokens over multiple time points. Our heterogeneous accelerator architecture Bishop concurrently processes workload packed in TTBs and explores intra- and inter-bundle multiple-bit weight reuse to significantly reduce memory access. Bishop utilizes a stratifier, a dense core array, and a sparse core array to process MLP blocks and projection layers. The stratifier routes high-density spiking activation workload to the dense core and low-density counterpart to the sparse core, ensuring optimized processing tailored to the given spatiotemporal sparsity level. To further reduce data access and computation, we introduce a novel Bundle Sparsity-Aware (BSA) training pipeline that enhances not only the overall but also structured TTB-level firing sparsity. Moreover, the processing efficiency of self-attention layers is boosted by the proposed Error-Constrained TTB Pruning (ECP), which trims activities in spiking queries, keys, and values both before and after the computation of spiking attention maps with a well-defined error bound. Finally, we design a reconfigurable TTB spiking attention core to efficiently compute spiking attention maps by executing highly simplified "AND" and "Accumulate" operations. On average, Bishop achieves a 5.91x speedup and 6.11x improvement in energy efficiency over previous SNN accelerators, while delivering higher accuracy across multiple datasets.
title Bishop: Sparsified Bundling Spiking Transformers on Heterogeneous Cores with Error-Constrained Pruning
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2505.12281