Guodong Yang

Kaixuan Zhang, Yunfan Cui, Shuhao Zhang et al.

The rapid expansion of Transformer-based large language models has dramatically increased the need for high-performance GPUs. As a result, there is growing demand for fast, accurate, and widely generalizable GPU performance models to support next-generation hardware selection and system-level exploration. However, current data-driven methods are limited, exhibiting poor generalization across hardware and inadequate modeling of complex production-level kernels common in modern inference stacks. To address these issues, we present PipeWeave, a unified GPU modeling framework. This approach first employs an analytical model to quantify a given kernel's demands on the GPU's heterogeneous instruction pipelines. These analytical features are then fed into a machine learning (ML) model to capture complex cross-pipeline interactions and resource dependencies, enabling high-fidelity performance prediction. Our evaluation across 11 GPU types from four generations of major architectures on two widely-used serving systems demonstrates that PipeWeave delivers high fidelity and strong generalizability. It achieves accurate predictions, with only 6.1% average error at the kernel level and 8.5% for end-to-end inference -- reducing the error of state-of-the-art methods by 6.7x and 4.4x, respectively. We also demonstrate PipeWeave's value "beyond simulation" by utilizing its performance ceiling to diagnose implementation shortcomings and guide the optimization of a production fused MoE Triton kernel, achieving up to 1.7x speedup. Code is available https://github.com/zksainx/pipeweave.

Suyi Li, Lingyun Yang, Xiaoxiao Jiang et al.

Text-to-image (T2I) generation using diffusion models has become a blockbuster service in today's AI cloud. A production T2I service typically involves a serving workflow where a base diffusion model is augmented with various "add-on" modules, notably ControlNet and LoRA, to enhance image generation control. Compared to serving the base model alone, these add-on modules introduce significant loading and computational overhead, resulting in increased latency. In this paper, we present SwiftDiffusion, a system that efficiently serves a T2I workflow through a holistic approach. SwiftDiffusion decouples ControNet from the base model and deploys it as a separate, independently scaled service on dedicated GPUs, enabling ControlNet caching, parallelization, and sharing. To mitigate the high loading overhead of LoRA serving, SwiftDiffusion employs a bounded asynchronous LoRA loading (BAL) technique, allowing LoRA loading to overlap with the initial base model execution by up to k steps without compromising image quality. Furthermore, SwiftDiffusion optimizes base model execution with a novel latent parallelism technique. Collectively, these designs enable SwiftDiffusion to outperform the state-of-the-art T2I serving systems, achieving up to 7.8x latency reduction and 1.6x throughput improvement in serving SDXL models on H800 GPUs, without sacrificing image quality.

Peijie Dong, Ruibo Fan, Yuechen Tao et al.

Training large language models using 4-bit arithmetic enhances throughput and memory efficiency. Yet, the limited dynamic range of FP4 increases sensitivity to outliers. While NVFP4 mitigates quantization error via hierarchical microscaling, a persistent loss gap remains compared to BF16. This study conducts a longitudinal analysis of outlier dynamics across architecture during NVFP4 pretraining, focusing on where they localize, why they occur, and how they evolve temporally. We find that, compared with Softmax Attention (SA), Linear Attention (LA) reduces per-tensor heavy tails but still exhibits persistent block-level spikes under block quantization. Our analysis attributes outliers to specific architectural components: Softmax in SA, gating in LA, and SwiGLU in FFN, with "post-QK" operations exhibiting higher sensitivity to quantization. Notably, outliers evolve from transient spikes early in training to a small set of persistent hot channels (i.e., channels with persistently large magnitudes) in later stages. Based on these findings, we introduce Hot-Channel Patch (HCP), an online compensation mechanism that identifies hot channels and reinjects residuals using hardware-efficient kernels. We then develop CHON, an NVFP4 training recipe integrating HCP with post-QK operation protection. On GLA-1.3B model trained for 60B tokens, CHON reduces the loss gap to BF16 from 0.94% to 0.58% while maintaining downstream accuracy.

Kaixuan Zhang, Chutong Ding, Shiyou Qian et al.

The rapid adoption of Large Language Models (LLMs) has made GPU inference efficiency an increasingly critical system concern. The runtime of LLM workloads is largely dominated by tile-based kernels, particularly General Matrix Multiplications (GEMMs). Although these kernels are highly optimized, their performance remains sensitive to a large space of runtime parameters, such as tile sizes and pipeline stages. The interaction between these parameters and hardware resources leads to a non-convex optimization landscape. Existing approaches to parameter configuration -- including search-based auto-tuning, heuristic rules, and learned cost models -- face a fundamental trade-off between performance optimality and runtime efficiency. In this paper, we present WaveTune, a wave-aware framework for runtime kernel auto-tuning. First, we introduce a unified mapping method to handle input diversity and decompose the configuration space to manage high dimensionality. Second, we develop an analytical wave-aware bilinear model that accurately predicts kernel latency. Third, we design a sparse sampling scheme based on wave structures and a lightweight dual-table retrieval mechanism to minimize runtime overhead. As a result, WaveTune enables precise and efficient runtime configuration for GPU kernels. Across three representative kernels and five GPU architectures, WaveTune consistently achieves near-optimal kernel performance, delivering up to 1.83x kernel-level speedup and up to 1.33x end-to-end TTFT reduction, while reducing runtime decision overhead by five orders of magnitude compared to exhaustive search. These results demonstrate that WaveTune effectively eliminates the traditional trade-off between configuration latency and execution optimality, providing a practical and robust solution for high-performance LLM inference.

Lingyun Yang, Suyi Li, Tianyu Feng et al.

Text-to-image generation executes a diffusion workflow comprising multiple models centered on a base diffusion model. Existing serving systems treat each workflow as an opaque monolith, provisioning, placing, and scaling all constituent models together, which obscures internal dataflow, prevents model sharing, and enforces coarse-grained resource management. In this paper, we make a case for micro-serving diffusion workflows with LegoDiffusion, a system that decomposes a workflow into loosely coupled model-execution nodes that can be independently managed and scheduled. By explicitly managing individual model inference, LegoDiffusion unlocks cluster-scale optimizations, including per-model scaling, model sharing, and adaptive model parallelism. Collectively, LegoDiffusion outperforms existing diffusion workflow serving systems, sustaining up to 3x higher request rates and tolerating up to 8x higher burst traffic.

Fangcheng Jin, Yuqi Wang, Peixin Ma et al.

Achieving robust locomotion on complex terrains remains a challenge due to high dimensional control and environmental uncertainties. This paper introduces a teacher prior framework based on the teacher student paradigm, integrating imitation and auxiliary task learning to improve learning efficiency and generalization. Unlike traditional paradigms that strongly rely on encoder-based state embeddings, our framework decouples the network design, simplifying the policy network and deployment. A high performance teacher policy is first trained using privileged information to acquire generalizable motion skills. The teacher's motion distribution is transferred to the student policy, which relies only on noisy proprioceptive data, via a generative adversarial mechanism to mitigate performance degradation caused by distributional shifts. Additionally, auxiliary task learning enhances the student policy's feature representation, speeding up convergence and improving adaptability to varying terrains. The framework is validated on a humanoid robot, showing a great improvement in locomotion stability on dynamic terrains and significant reductions in development costs. This work provides a practical solution for deploying robust locomotion strategies in humanoid robots.

Zhiqiang Yang, Renshuai Tao, Xiaolong Zheng et al.

Existing deepfake detection methods heavily depend on labeled training data. However, as AI-generated content becomes increasingly realistic, even \textbf{human annotators struggle to distinguish} between deepfakes and authentic images. This makes the labeling process both time-consuming and less reliable. Specifically, there is a growing demand for approaches that can effectively utilize large-scale unlabeled data from online social networks. Unlike typical unsupervised learning tasks, where categories are distinct, AI-generated faces closely mimic real image distributions and share strong similarities, causing performance drop in conventional strategies. In this paper, we introduce the Dual-Path Guidance Network (DPGNet), to tackle two key challenges: (1) bridging the domain gap between faces from different generation models, and (2) utilizing unlabeled image samples. The method features two core modules: text-guided cross-domain alignment, which uses learnable prompts to unify visual and textual embeddings into a domain-invariant feature space, and curriculum-driven pseudo label generation, which dynamically exploit more informative unlabeled samples. To prevent catastrophic forgetting, we also facilitate bridging between domains via cross-domain knowledge distillation. Extensive experiments on \textbf{11 popular datasets}, show that DPGNet outperforms SoTA approaches by \textbf{6.3\%}, highlighting its effectiveness in leveraging unlabeled data to address the annotation challenges posed by the increasing realism of deepfakes.

Xiaoxiao Jiang, Suyi Li, Lingyun Yang et al.

Generative image editing using diffusion models has become a prevalent application in today's AI cloud services. In production environments, image editing typically involves a mask that specifies the regions of an image template to be edited. The use of masks provides direct control over the editing process and introduces sparsity in the model inference. In this paper, we present InstGenIE, a system that efficiently serves image editing requests. The key insight behind InstGenIE is that image editing only modifies the masked regions of image templates while preserving the original content in the unmasked areas. Driven by this insight, InstGenIE judiciously skips redundant computations associated with the unmasked areas by reusing cached intermediate activations from previous inferences. To mitigate the high cache loading overhead, InstGenIE employs a bubble-free pipeline scheme that overlaps computation with cache loading. Additionally, to reduce queuing latency in online serving while improving the GPU utilization, InstGenIE proposes a novel continuous batching strategy for diffusion model serving, allowing newly arrived requests to join the running batch in just one step of denoising computation, without waiting for the entire batch to complete. As heterogeneous masks induce imbalanced loads, InstGenIE also develops a load balancing strategy that takes into account the loads of both computation and cache loading. Collectively, InstGenIE outperforms state-of-the-art diffusion serving systems for image editing, achieving up to 3x higher throughput and reducing average request latency by up to 14.7x while ensuring image quality.