SPARTA：面向长上下文边缘推理的自适应注意力稀疏三值 LLM 加速器

View PDF HTML (experimental)

Abstract:LLMs on edge devices are constrained by limited memory and bandwidth. Although 1.58-bit ternary weights greatly reduce model storage, long-context inference is increasingly bottlenecked by KV-cache traffic and attention. Existing accelerators also use separate datapaths for ternary linear layers and attention, incurring hardware and execution overhead. We present SPARTA, a sparse ternary LLM accelerator for efficient long-context edge inference. SPARTA integrates: (1) a phase-aware sparse dataflow exploiting heterogeneous sparsity across compute-bound prefill and memory-bound decode; (2) a unified compute fabric mapping ternary SpMM/SpMV and bit-serial attention onto shared processing elements to exploit zero-weight and zero-bit sparsity; and (3) an adaptive two-stage attention engine combining page-level Top-k selection to reduce K-cache accesses with token-level Top-p pruning to reduce V-cache traffic. SPARTA reduces KV traffic by 4.1-4.8x over full attention and 1.3-1.5x over a state-of-the-art sparse-attention scheme, with task-dependent quality loss. Compared with NVIDIA Jetson AGX Orin, SPARTA achieves 3.9x higher generation throughput and 8.3x higher end-to-end energy efficiency on average; versus AMD Ryzen 7 9700X, the gains are 5.0x and 81.7x. It also delivers 1.6-12.0x higher throughput, 2.0-4.2x higher energy efficiency, and 1.3-51.4x higher area efficiency than FIGLUT, TeLLMe, and MCBP.

Subjects:	Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
Cite as:	arXiv:2610.34824 [cs.AR]
	(or arXiv:2610.34824v1 [cs.AR] for this version)
	https://doi.org/10.48550/arXiv.2610.34824 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Weihong Xu [view email]
[v1] Mon, 26 Oct 2026 04:13:47 UTC (1,488 KB)