CIMple delivers a 32 kb digital SRAM-based compute-in-memory accelerator for transformer self-attention that reaches 26.1 TOPS/W at 0.85 V in 28 nm with INT8 precision using dual-banked architecture and LUT-based split softmax.
Efficient Softmax approximation for deep neural networks with attention mechanism
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
verdicts
UNVERDICTED 3representative citing papers
Presents quantization, checkpointing, softmax approximation, and logits masking to achieve substantial peak memory reductions in LoRA fine-tuning of 3B LLMs.
4-bit and 6-bit integer-only quantized Transformers implemented on Spartan-7 FPGA for AIoT time-series forecasting achieve 0.63% higher test loss than 8-bit baselines but up to 132x speedup and 48x lower energy.
citing papers explorer
-
CIMple: Standard-cell SRAM-based CIM with LUT-based split softmax for attention acceleration
CIMple delivers a 32 kb digital SRAM-based compute-in-memory accelerator for transformer self-attention that reaches 26.1 TOPS/W at 0.85 V in 28 nm with INT8 precision using dual-banked architecture and LUT-based split softmax.
-
Techniques for Peak Memory Reduction for LoRA Fine-tuning of LLMs on Edge Devices
Presents quantization, checkpointing, softmax approximation, and logits masking to achieve substantial peak memory reductions in LoRA fine-tuning of 3B LLMs.
-
Integer-only Quantized Transformers for Embedded FPGA-based Time-series Forecasting in AIoT
4-bit and 6-bit integer-only quantized Transformers implemented on Spartan-7 FPGA for AIoT time-series forecasting achieve 0.63% higher test loss than 8-bit baselines but up to 132x speedup and 48x lower energy.