arXiv subjects

Xiuhong Li

Publications and source records attributed to Xiuhong Li.

18 recordsLinked to original sources

arXiv2026

Tropical: Enhancing SLO Attainment in Disaggregated LLM Serving via SLO-Aware Multiplexing

To guarantee service quality in transformer based large language model (LLM) serving, it is essential to meet the latency constraints of both the prefill phase (measured by Time-to-First-Token, TTFT) and the decode phase (measured by Time-per-Output-Token, TPOT). Non-disaggregated serving places prefill and decode on the same worker, while disaggregated serving places the prefill and decode on isolated workers. However, no single architecture excels in both TTFT and TPOT metrics. After conducting a root cause analysis, we concluded that indisaggregated LLM serving, prefill execution has minimal interference with decode execution but result in high queuing times. In contrast,non-disaggregated LLM serving effectively reduces queuing times but introduces significant interference between prefills and decodes. In order to leverage the best aspects of both non-disaggregated anddisaggregated LLM serving, we have designed and implemented Tropical.Tropical introduces an sevice-level objectives (SLO)-aware multiplexing strategy that balances the queuing time and the interference, enabling the LLM serving to achieve high TTFT and TPOT SLOs simultaneously. Our evaluation of real-world datasets reveals that Tropical outperforms both state-of-the-art non-disaggregated and disaggregated LLM serving systems, achieving up to 2.09 more requests within a 90% SLO attainment. Specially, compared to the disaggregated LLM serving system, Tropicalimproves P90 TTFT performance by 9 with only an 15% reduction in P90 TPOT. Against the non-disaggregated LLM serving systems, Tropicaldelivers a 2.8 performance improvement in P90 TPOT while maintaining the same P90 TTFT.

cs.DC↗

arXiv2026

NanoCP: Request-Level Dynamic Context Parallelism for Data-Expert Parallel Decoding

Modern serving systems for Mixture-of-Experts (MoE) models adopt hybrid data-expert parallelism: expert parallelism (EP) shards experts across GPUs to scale capacity, while data parallelism (DP) replicates attention layers across instances to process independent requests. Existing systems bind each request's attention, MoE communication, and KV cache to a single instance. Because attention latency scales with KV cache size while MoE communication latency scales with batch size, this binding cannot balance both simultaneously, producing EP stragglers; it also fragments KV memory across instances, inflating tail latency under long contexts. While existing context parallelism (CP) mitigates these constraints, its uniform parallelism degree incurs prohibitive communication and attention-side overheads. We present \work, which decouples MoE communication from KV cache placement and achieves dual balance through dynamic context parallelism (DCP). DCP assigns each request a context-parallel degree sized to its KV footprint: long requests distribute attention across multiple instances; short requests remain local. This dynamic parallelism effectively liquefies the KV cache across the cluster, balancing both the per-instance KV cache occupancy and batch sizes without unnecessary load-balancing costs. To bridge DCP with static execution, \work introduces an ahead-of-time (AOT) graph engine paired with a custom routing-based communication backend. Experimental results show that \work maintains up to $1.88\times$--$3.27\times$ higher request rates under strict time-per-output-token (TPOT) service level objectives (SLOs). Furthermore, \work significantly mitigates stragglers, reducing P99 tail latency by up to $1.79\times$--$2.12\times$.

cs.DC↗

arXiv2025

TASP: Topology-aware Sequence Parallelism

Long-context large language models (LLMs) face constraints due to the quadratic complexity of the self-attention mechanism. The mainstream sequence parallelism (SP) method, Ring Attention, attempts to solve this by distributing the query into multiple query chunks across accelerators and enable each Q tensor to access all KV tensors from other accelerators via the Ring AllGather communication primitive. However, it exhibits low communication efficiency, restricting its practical applicability. This inefficiency stems from the mismatch between the Ring AllGather communication primitive it adopts and the AlltoAll topology of modern accelerators. A Ring AllGather primitive is composed of iterations of ring-styled data transfer, which can only utilize a very limited fraction of an AlltoAll topology. Inspired by the Hamiltonian decomposition of complete directed graphs, we identify that modern accelerator topology can be decomposed into multiple orthogonal ring datapaths which can concurrently transfer data without interference. Based on this, we further observe that the Ring AllGather primitive can also be decomposed into the same number of concurrent ring-styled data transfer at every iteration. Based on these insights, we propose TASP, a topology-aware SP method for long-context LLMs that fully utilizes the communication capacity of modern accelerators via topology decomposition and primitive decomposition. Experimental results on both single-node and multi-node NVIDIA H100 systems and a single-node AMD MI300X system demonstrate that TASP achieves higher communication efficiency than Ring Attention on these modern accelerator topologies and achieves up to 3.58 speedup than Ring Attention and its variant Zigzag-Ring Attention. The code is available at https://github.com/infinigence/HamiltonAttention.

cs.LG↗

arXiv2025

Efficient and Adaptable Overlapping for Computation and Communication via Signaling and Reordering

Generative models have achieved remarkable success across various applications, driving the demand for multi-GPU computing. Inter-GPU communication becomes a bottleneck in multi-GPU computing systems, particularly on consumer-grade GPUs. By exploiting concurrent hardware execution, overlapping computation and communication latency becomes an effective technique for mitigating the communication overhead. We identify that an efficient and adaptable overlapping design should satisfy (1) tile-wise overlapping to maximize the overlapping opportunity, (2) interference-free computation to maintain the original computational performance, and (3) communication agnosticism to reduce the development burden against varying communication primitives. Nevertheless, current designs fail to simultaneously optimize for all of those features. To address the issue, we propose FlashOverlap, which utilizes a novel signaling mechanism: when part of the output finishes, the computation kernel sends a signal to trigger the communication of that part, while continuing the computation of the remaining part (interference-free computation). Consequently, the communication of the finished part and the computation of the remaining part can be overlapped. On top of the signaling mechanism, FlashOverlap comprises two key components: (1) the determination of the signaling timing to boost the overlap efficiency (tile-wise overlapping), and (2) a pre-communication reordering to create the contiguous address for finished data, enabling communication by simply calling NCCL APIs (communication agnosticism), and a post-communication reordering to correct the data order. Experiments show that FlashOverlap achieves up to 1.65x speedup through overlap, outperforming existing works in most cases. Code is available at https://github.com/infinigence/FlashOverlap.

cs.DC↗

arXiv2025

Zeppelin: Balancing Variable-length Workloads in Data Parallel Large Model Training

Training large language models (LLMs) with increasingly long and varying sequence lengths introduces severe load imbalance challenges in large-scale data-parallel training. Recent frameworks attempt to mitigate these issues through data reorganization or hybrid parallel strategies. However, they often overlook how computational and communication costs scale with sequence length, resulting in suboptimal performance. We identify three critical challenges: (1) varying computation-to-communication ratios across sequences of different lengths in distributed attention, (2) mismatch between static NIC-GPU affinity and dynamic parallel workloads, and (3) distinct optimal partitioning strategies required for quadratic attention versus linear components. To address these challenges, we present Zeppelin, a novel training system that integrates three key techniques: (1) a hierarchical sequence partitioning method for the attention module that reduces communication overhead and balances computation, supported by an efficient attention engine that applies divergent parallel strategies; (2) a routing layer that orchestrates inter-node transfers to fully utilize NIC bandwidth; and (3) a remapping layer that transforms sequence layouts between attention and linear modules, ensuring high computational efficiency across both. Comprehensive evaluations across diverse configurations show that Zeppelin delivers an average 2.80x speedup over state-of-the-art methods.

cs.DC↗

arXiv2025

Past-Future Scheduler for LLM Serving under SLA Guarantees

The exploration and application of Large Language Models (LLMs) is thriving. To reduce deployment costs, continuous batching has become an essential feature in current service frameworks. The effectiveness of continuous batching relies on an accurate estimate of the memory requirements of requests. However, due to the diversity in request output lengths, existing frameworks tend to adopt aggressive or conservative schedulers, which often result in significant overestimation or underestimation of memory consumption. Consequently, they suffer from harmful request evictions or prolonged queuing times, failing to achieve satisfactory throughput under strict Service Level Agreement (SLA) guarantees (a.k.a. goodput), across various LLM application scenarios with differing input-output length distributions. To address this issue, we propose a novel Past-Future scheduler that precisely estimates the peak memory resources required by the running batch via considering the historical distribution of request output lengths and calculating memory occupancy at each future time point. It adapts to applications with all types of input-output length distributions, balancing the trade-off between request queuing and harmful evictions, thereby consistently achieving better goodput. Furthermore, to validate the effectiveness of the proposed scheduler, we developed a high-performance LLM serving framework, LightLLM, that implements the Past-Future scheduler. Compared to existing aggressive or conservative schedulers, LightLLM demonstrates superior goodput, achieving up to 2-3$\times$ higher goodput than other schedulers under heavy loads. LightLLM is open source to boost the research in such direction (https://github.com/ModelTC/lightllm).

cs.DC↗

arXiv2025

MxMoE: Mixed-precision Quantization for MoE with Accuracy and Performance Co-Design

Mixture-of-Experts (MoE) models face deployment challenges due to their large parameter counts and computational demands. We explore quantization for MoE models and highlight two key insights: 1) linear blocks exhibit varying quantization sensitivity, and 2) divergent expert activation frequencies create heterogeneous computational characteristics. Based on these observations, we introduce MxMoE, a mixed-precision optimization framework for MoE models that considers both algorithmic and system perspectives. MxMoE navigates the design space defined by parameter sensitivity, expert activation dynamics, and hardware resources to derive efficient mixed-precision configurations. Additionally, MxMoE automatically generates optimized mixed-precision GroupGEMM kernels, enabling parallel execution of GEMMs with different precisions. Evaluations show that MxMoE outperforms existing methods, achieving 2.4 lower Wikitext-2 perplexity than GPTQ at 2.25-bit and delivering up to 3.4x speedup over full precision, as well as up to 29.4% speedup over uniform quantization at equivalent accuracy with 5-bit weight-activation quantization. Our code is available at https://github.com/cat538/MxMoE.

cs.LG↗

arXiv2025

semi-PD: Towards Efficient LLM Serving via Phase-Wise Disaggregated Computation and Unified Storage

Existing large language model (LLM) serving systems fall into two categories: 1) a unified system where prefill phase and decode phase are co-located on the same GPU, sharing the unified computational resource and storage, and 2) a disaggregated system where the two phases are disaggregated to different GPUs. The design of the disaggregated system addresses the latency interference and sophisticated scheduling issues in the unified system but leads to storage challenges including 1) replicated weights for both phases that prevent flexible deployment, 2) KV cache transfer overhead between the two phases, 3) storage imbalance that causes substantial wasted space of the GPU capacity, and 4) suboptimal resource adjustment arising from the difficulties in migrating KV cache. Such storage inefficiency delivers poor serving performance under high request rates. In this paper, we identify that the advantage of the disaggregated system lies in the disaggregated computation, i.e., partitioning the computational resource to enable the asynchronous computation of two phases. Thus, we propose a novel LLM serving system, semi-PD, characterized by disaggregated computation and unified storage. In semi-PD, we introduce a computation resource controller to achieve disaggregated computation at the streaming multi-processor (SM) level, and a unified memory manager to manage the asynchronous memory access from both phases. semi-PD has a low-overhead resource adjustment mechanism between the two phases, and a service-level objective (SLO) aware dynamic partitioning algorithm to optimize the SLO attainment. Compared to state-of-the-art systems, semi-PD maintains lower latency at higher request rates, reducing the average end-to-end latency per request by 1.27-2.58x on DeepSeek series models, and serves 1.55-1.72x more requests adhering to latency constraints on Llama series models.

cs.CL↗

arXiv2025

MBQ: Modality-Balanced Quantization for Large Vision-Language Models

Vision-Language Models (VLMs) have enabled a variety of real-world applications. The large parameter size of VLMs brings large memory and computation overhead which poses significant challenges for deployment. Post-Training Quantization (PTQ) is an effective technique to reduce the memory and computation overhead. Existing PTQ methods mainly focus on large language models (LLMs), without considering the differences across other modalities. In this paper, we discover that there is a significant difference in sensitivity between language and vision tokens in large VLMs. Therefore, treating tokens from different modalities equally, as in existing PTQ methods, may over-emphasize the insensitive modalities, leading to significant accuracy loss. To deal with the above issue, we propose a simple yet effective method, Modality-Balanced Quantization (MBQ), for large VLMs. Specifically, MBQ incorporates the different sensitivities across modalities during the calibration process to minimize the reconstruction loss for better quantization parameters. Extensive experiments show that MBQ can significantly improve task accuracy by up to 4.4% and 11.6% under W3 and W4A8 quantization for 7B to 70B VLMs, compared to SOTA baselines. Additionally, we implement a W3 GPU kernel that fuses the dequantization and GEMV operators, achieving a 1.4x speedup on LLaVA-onevision-7B on the RTX 4090. The code is available at https://github.com/thu-nics/MBQ.

cs.CV↗

arXiv2024

SKVQ: Sliding-window Key and Value Cache Quantization for Large Language Models

Large language models (LLMs) can now handle longer sequences of tokens, enabling complex tasks like book understanding and generating lengthy novels. However, the key-value (KV) cache required for LLMs consumes substantial memory as context length increasing, becoming the bottleneck for deployment. In this paper, we present a strategy called SKVQ, which stands for sliding-window KV cache quantization, to address the issue of extremely low bitwidth KV cache quantization. To achieve this, SKVQ rearranges the channels of the KV cache in order to improve the similarity of channels in quantization groups, and applies clipped dynamic quantization at the group level. Additionally, SKVQ ensures that the most recent window tokens in the KV cache are preserved with high precision. This helps maintain the accuracy of a small but important portion of the KV cache.SKVQ achieves high compression ratios while maintaining accuracy. Our evaluation on LLMs demonstrates that SKVQ surpasses previous quantization approaches, allowing for quantization of the KV cache to 2-bit keys and 1.5-bit values with minimal loss of accuracy. With SKVQ, it is possible to process context lengths of up to 1M on an 80GB memory GPU for a 7b model and up to 7 times faster decoding.

cs.LG↗

arXiv2024

A Survey on Efficient Inference for Large Language Models

Large Language Models (LLMs) have attracted extensive attention due to their remarkable performance across various tasks. However, the substantial computational and memory requirements of LLM inference pose challenges for deployment in resource-constrained scenarios. Efforts within the field have been directed towards developing techniques aimed at enhancing the efficiency of LLM inference. This paper presents a comprehensive survey of the existing literature on efficient LLM inference. We start by analyzing the primary causes of the inefficient LLM inference, i.e., the large model size, the quadratic-complexity attention operation, and the auto-regressive decoding approach. Then, we introduce a comprehensive taxonomy that organizes the current literature into data-level, model-level, and system-level optimization. Moreover, the paper includes comparative experiments on representative methods within critical sub-fields to provide quantitative insights. Last but not least, we provide some knowledge summary and discuss future research directions.

cs.CL↗

arXiv2024

MuxServe: Flexible Spatial-Temporal Multiplexing for Multiple LLM Serving

Large language models (LLMs) have demonstrated remarkable performance, and organizations are racing to serve LLMs of varying sizes as endpoints for use-cases like chat, programming and search. However, efficiently serving multiple LLMs poses significant challenges for existing approaches due to varying popularity of LLMs. In the paper, we present MuxServe, a flexible spatial-temporal multiplexing system for efficient multiple LLM serving. The key insight behind is to colocate LLMs considering their popularity to multiplex memory resources, and leverage the characteristics of prefill and decoding phases to separate and flexibly colocate them to multiplex computation resources. MuxServe formally formulates the multiplexing problem, and proposes a novel placement algorithm and adaptive batch scheduling strategy to identify optimal colocations and maximize utilization. MuxServe designs a unified resource manager to enable flexible and efficient multiplexing. Evaluation results show that MuxServe can achieves up to $1.8\times$ higher throughput or processes $2.9\times$ more requests within $99\%$ SLO attainment. The code is available at: \url{https://github.com/hao-ai-lab/MuxServe}.

cs.DC↗

arXiv2024

FlashDecoding++: Faster Large Language Model Inference on GPUs

As the Large Language Model (LLM) becomes increasingly important in various domains. However, the following challenges still remain unsolved in accelerating LLM inference: (1) Synchronized partial softmax update. The softmax operation requires a synchronized update operation among each partial softmax result, leading to ~20% overheads for the attention computation in LLMs. (2) Under-utilized computation of flat GEMM. The shape of matrices performing GEMM in LLM inference is flat, leading to under-utilized computation and >50% performance loss after padding zeros in previous designs. (3) Performance loss due to static dataflow. Kernel performance in LLM depends on varied input data features, hardware configurations, etc. A single and static dataflow may lead to a 50.25% performance loss for GEMMs of different shapes in LLM inference. We present FlashDecoding++, a fast LLM inference engine supporting mainstream LLMs and hardware back-ends. To tackle the above challenges, FlashDecoding++ creatively proposes: (1) Asynchronized softmax with unified max value. FlashDecoding++ introduces a unified max value technique for different partial softmax computations to avoid synchronization. (2) Flat GEMM optimization with double buffering. FlashDecoding++ points out that flat GEMMs with different shapes face varied bottlenecks. Then, techniques like double buffering are introduced. (3) Heuristic dataflow with hardware resource adaptation. FlashDecoding++ heuristically optimizes dataflow using different hardware resource considering input dynamics. Due to the versatility of optimizations in FlashDecoding++, FlashDecoding++ can achieve up to 4.86x and 2.18x speedup on both NVIDIA and AMD GPUs compared to Hugging Face implementations. FlashDecoding++ also achieves an average speedup of 1.37x compared to state-of-the-art LLM inference engines on mainstream LLMs.

cs.LG↗

arXiv2023

Proteus: Simulating the Performance of Distributed DNN Training

DNN models are becoming increasingly larger to achieve unprecedented accuracy, and the accompanying increased computation and memory requirements necessitate the employment of massive clusters and elaborate parallelization strategies to accelerate DNN training. In order to better optimize the performance and analyze the cost, it is indispensable to model the training throughput of distributed DNN training. However, complex parallelization strategies and the resulting complex runtime behaviors make it challenging to construct an accurate performance model. In this paper, we present Proteus, the first standalone simulator to model the performance of complex parallelization strategies through simulation execution. Proteus first models complex parallelization strategies with a unified representation named Strategy Tree. Then, it compiles the strategy tree into a distributed execution graph and simulates the complex runtime behaviors, comp-comm overlap and bandwidth sharing, with a Hierarchical Topo-Aware Executor (HTAE). We finally evaluate Proteus across a wide variety of DNNs on three hardware configurations. Experimental results show that Proteus achieves $3.0\%$ average prediction error and preserves order for training throughput of various parallelization strategies. Compared to state-of-the-art approaches, Proteus reduces prediction error by up to $133.8\%$.

cs.DC↗

arXiv2023

Exploring Multimodal Sentiment Analysis via CBAM Attention and Double-layer BiLSTM Architecture

Because multimodal data contains more modal information, multimodal sentiment analysis has become a recent research hotspot. However, redundant information is easily involved in feature fusion after feature extraction, which has a certain impact on the feature representation after fusion. Therefore, in this papaer, we propose a new multimodal sentiment analysis model. In our model, we use BERT + BiLSTM as new feature extractor to capture the long-distance dependencies in sentences and consider the position information of input sequences to obtain richer text features. To remove redundant information and make the network pay more attention to the correlation between image and text features, CNN and CBAM attention are added after splicing text features and picture features, to improve the feature representation ability. On the MVSA-single dataset and HFM dataset, compared with the baseline model, the ACC of our model is improved by 1.78% and 1.91%, and the F1 value is enhanced by 3.09% and 2.0%, respectively. The experimental results show that our model achieves a sound effect, similar to the advanced model.

cs.CV↗

arXiv2018

CuLDA_CGS: Solving Large-scale LDA Problems on GPUs

Latent Dirichlet Allocation(LDA) is a popular topic model. Given the fact that the input corpus of LDA algorithms consists of millions to billions of tokens, the LDA training process is very time-consuming, which may prevent the usage of LDA in many scenarios, e.g., online service. GPUs have benefited modern machine learning algorithms and big data analysis as they can provide high memory bandwidth and computation power. Therefore, many frameworks, e.g. Ten- sorFlow, Caffe, CNTK, support to use GPUs for accelerating the popular machine learning data-intensive algorithms. However, we observe that LDA solutions on GPUs are not satisfying. In this paper, we present CuLDA_CGS, a GPU-based efficient and scalable approach to accelerate large-scale LDA problems. CuLDA_CGS is designed to efficiently solve LDA problems at high throughput. To it, we first delicately design workload partition and synchronization mechanism to exploit the benefits of mul- tiple GPUs. Then, we offload the LDA sampling process to each individual GPU by optimizing from the sampling algorithm, par- allelization, and data compression perspectives. Evaluations show that compared with state-of-the-art LDA solutions, CuLDA_CGS outperforms them by a large margin (up to 7.3X) on a single GPU. CuLDA_CGS is able to achieve extra 3.0X speedup on 4 GPUs. The source code is publicly available on https://github.com/cuMF/ CuLDA_CGS.

cs.DC↗

arXiv2017

Microfocus small-angle X-ray scattering at SSRF BL16B1

Offering high-brilliance X-ray beams on micrometer length scales, the microfocus-SAXS at SSRF BL16B1 was established with a KB mirror system for studying small sample volumes, or probing micro-scopic morphologies. The SAXS minimum q value was 0.1nm-1 with a flux of 1.5 * 10^10 photons/s. Two position-resolved scanning experimental methods were combined with microfocus-SAXS that include STXM and CT. To improve the significant smearing effect in the horizontal direction, an effective and easy-to-use desmearing procedure for two-dimensional SAXS pattern based on the blind deconvolution was developed and the deblurring results demonstrated the good restoration effect for the defocus image. Finally, a bamboo sample was selected for SAXS-CT experiment which illustrated the performance of the microfocus-SAXS method.

physics.ins-det↗

arXiv2013

Hard X-ray optics simulation using the coherent mode decomposition of Gaussian Schell model

The propagation of hard X ray beam from partially coherent synchrotron source is simulated by using the novel method based on the coherent mode decomposition of Gaussian Schell model and wave front propagation. We investigate how the coherency properties and intensity distributions of the beam are changed by propagation through optical elements. Here, we simulate and analyze the propagation of the partially coherent radiation transmitted through an ideal slit. We present the first simulations for focusing partially coherent synchrotron hard X ray beams using this novel method. And when compared with the traditional method which assumes the source is a totally coherent point source or completely incoherent, this method is proved to be more reasonable and can also demonstrate the coherence properties of the focusing beam. We also simulate the double slit experiment and the simulated results validate the academic analysis.

physics.acc-ph↗