Anda: Unlocking Efficient LLM Inference with a Variable-Length Grouped Activation Data Format

Abstract

The widely-used, weight-only quantized large language models (LLMs), which leverage low-bit integer (INT) weights and retain floating-point (FP) activations, reduce storage requirements while maintaining accuracy. However, this shifts the energy and latency bottlenecks towards the FP activations that are associated with costly memory accesses and computations. Existing LLM accelerators focus primarily on computation optimizations, overlooking the potential of jointly optimizing FP computations and data movement, particularly for the dominant FP-INT GeMM operations in LLM inference. To address these challenges, we investigate the sensitivity of activation precision across various LLM modules and its impact on overall model accuracy. Based on our findings, we first propose the Anda data type: an adaptive data format with group-shared exponent bits and dynamic mantissa bit allocation. Secondly, we develop an iterative post-training adaptive precision search algorithm that optimizes the bit-width for different LLM modules to balance model accuracy, energy efficiency, and inference speed. Lastly, a suite of hardware optimization techniques is proposed to maximally exploit the benefits of the Anda format. These include a bit-plane-based data organization scheme, Anda-enhanced processing units with bit-serial computation, and a runtime bit-plane Anda compressor to simultaneously optimize storage, computation, and memory footprints. Our evaluations on FPINT GeMM operations show that Anda achieves a 2.4x speedup, 4.0x area efficiency, and 3.1x energy efficiency improvement on average for popular LLMs including OPT, LLaMA, and LLaMA-2 series over the GPU-like FP-FP baseline. Anda demonstrates strong adaptability across various application scenarios, accuracy requirements, and system performance, enabling efficient LLM inference across a wide range of deployment scenarios.

To appear in 2025 IEEE International Symposium on High-Performance Computer Architecture (HPCA 2025)

References in corpus (9)

• Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks
• OliVe: Accelerating Large Language Models via Hardware-friendly Outlier-Victim Pair Quantization
• Explaining Neural Scaling Laws
• Dissecting Tensor Cores via Microbenchmarks: Latency, Throughput and Numeric Behaviors
• BitWave: Exploiting Column-Based Bit-Level Sparsity for Deep Learning Acceleration
• Mokey: Enabling Narrow Fixed-Point Inference for Out-of-the-Box Floating-Point Transformer Models
• BEBERT: Efficient and Robust Binary Ensemble BERT
• A Precision-Scalable RISC-V DNN Processor with On-Device Learning Capability at the Extreme Edge
• DaCapo: Accelerating Continuous Learning in Autonomous Systems for Video Analytics

Cited by in corpus (4)

• Efficient Precision-Scalable Hardware for Microscaling (MX) Processing in Robotics Learning
• A Scheduling Framework for Efficient MoE Inference on Edge GPU-NDP Systems
• Precision-Scalable Microscaling Datapaths with Optimized Reduction Tree for Efficient NPU Integration
• CD-PIM: A High-Bandwidth and Compute-Efficient LPDDR5-Based PIM for Low-Batch LLM Acceleration on Edge-Device