Large language models show promise for code generation, yet their effectiveness in producing high-performance GPU kernels remains insufficiently understood. Recent agentic workflows improve over direct prompting through specialized generation roles and execution feedback, but our characterization of a representative wo...
Yang-Jie Zhou, Xing-Chen Ma, Zi-Han Liu et al.· Proceedings of the 41st IEEE...· 0 citations
A layer of diagonal quantum gates can be specified by an integer-valued function on computational basis states. This description suggests implementing the layer by evaluating that function reversibly and translating its value into a phase using a shared quantum register. We develop this viewpoint for parallel phase kic...
An-Bang Wu, Li-Qiang Lu, Pu Pang et al.· 0 citations
Deltoris is presented, an algorithm-hardware co-design framework for efficient diffusion-based VLA inference that exploits the temporal similarity of consecutive inputs and proposes a temporal-aware bit-sparsity algorithm that computes only the differences between consecutive inputs, eliminating redundant bit-level ope...
Zheng Liu, Zeyu Guo, Zi-Han Liu et al.· 1 citation
Atlas is presented, an on device city scale 3DGS rendering framework that enables scalable rendering without runtime Internet access and proposes temporal aware LoD search and stereo rasterization to avoid redundant computation in VR.
He Zhu, Zheng Liu, Xingyang Li et al.· 0 citations
Weave is presented, to the authors' knowledge the first MoE overlap system that performs fine-grained dynamic SM scheduling - deciding per layer and per GPU by routing results at runtime, and achieves a 2.89x geometric-mean MoE-layer speedup and a 1.33x geometric-mean end-to-end speedup over five state-of-the-art basel...
Ziyu Huang, Yang-Jie Zhou, Chen-Hao Zhu et al.· 0 citations
Kaleido is presented, an algorithm hardware codesign that accelerates all operations in vDiTs by exploiting channel-wise spatiotemporal correlations in latent space and skips redundant computations by reusing partial results while preserving higher generative quality than prior methods.
Wen-Xuan Miao, Haosong Liu, Weiming Hu et al.· arXiv.org· 1 citation
This work proposes a series of optimizations for these two kernels, including computation-transfer pipelining, load balancing, and memory access fusion, achieving 1.97 × to 2.16 × proof generation speedup over a state-of-the-art open source GPU acceleration library.
Xinwei Qiang, Liukun Yu, Xiyu Wang et al.· IEEE International Symposium...· 0 citations
This work proposes FIBER, a new architecture that extends the GPU SIMT (single instruction, multiple thread) model, and extends the ISA, microarchitecture, and compiler to realize shared-register addressing, conflict-free operand delivery, and fiber-based program mapping.
Zi-Han Liu, Jingwen Leng, Yang-Jie Zhou et al.· 0 citations
This work proposes GPU-Tile-Sim, a tile-centric GPU simulation framework for LLM hardware-software co-design that represents kernel execution as a warp-level tile graph whose nodes capture tile-level operations and whose edges encode data and ordering constraints.
