Binglei Lou
Publications
Beyond GEMM-Centric NPUs: Enabling Efficient Diffusion LLM Sampling
Diffusion Large Language Models (dLLMs) introduce iterative denoising to enable parallel token generation, but their sampling phase displays fundamentally different characteristics compared to GEMM-centric transformer layers. Profiling on modern GPUs reveals that sampling can account for up to 70% of total model inference latency-primarily due to substantial memory loads and writes from vocabulary-wide logits, reduction-based token selection, and iterative masked updates. These processes demand large on-chip SRAM and involve irregular memory accesses that conventional NPUs struggle to handle efficiently. To address this, we identify a set of critical instructions that an NPU architecture must specifically optimize for dLLM sampling. Our design employs lightweight non-GEMM vector primitives, in-place memory reuse strategies, and a decoupled mixed-precision memory hierarchy. Together, these optimizations deliver up to a 2.53x speedup over the NVIDIA RTX A6000 GPU under an equivalent nm technology node. We also open-source our cycle-accurate simulation and post-synthesis RTL verification code, confirming functional equivalence with current dLLM PyTorch implementations.
Enhancing LUT-based Deep Neural Networks Inference through Architecture and Connectivity Optimization
Deploying deep neural networks (DNNs) on resource-constrained edge devices such as FPGAs requires a careful balance among latency, power, and hardware resource usage, while maintaining high accuracy. Existing Lookup Table (LUT)-based DNNs -- such as LogicNets, PolyLUT, and NeuraLUT -- face two critical challenges: the exponential growth of LUT size and inefficient random sparse connectivity. This paper presents SparseLUT, a comprehensive framework that addresses these challenges through two orthogonal optimizations. First, we propose an architectural enhancement that aggregates multiple PolyLUT sub-neurons via an adder, significantly reducing LUT consumption by 2.0x-13.9x and lowering inference latency by 1.2x-1.6x, all while maintaining comparable accuracy. Building upon this foundation, we further introduce a non-greedy training algorithm that optimizes neuron connectivity by selectively pruning less significant inputs and strategically regrowing more effective ones. This training optimization, which incurs no additional area and latency overhead, delivers consistent accuracy improvements across benchmarks -- achieving up to a 2.13% gain on MNIST and 0.94% on Jet Substructure Classification compared to existing LUT-DNN approaches.