arXiv subjects

Zizhao Mo

Publications and source records attributed to Zizhao Mo.

6 recordsLinked to original sources

arXiv2026

Poseidon: DAG-Guided Parallelism Search for LLM Pre-Training on Heterogeneous Clusters

With the rapid advancement of accelerator technologies, pre-training large language models (LLMs) on heterogeneous accelerator clusters has become increasingly crucial for maximizing hardware utilization. Existing systems, however, suffer from inaccurate training time modeling, which undermines the parallelization optimizations built upon it. Moreover, for current approaches, the vast configuration search space makes exhaustive exploration infeasible, forcing a trade-off between search time and training efficiency. To overcome these limitations, we introduce Poseidon, an efficient and scalable LLM training framework designed with heterogeneity awareness. Its core is an explicit training time model based on a directed acyclic graph. Building on this graph, Poseidon employs two efficient, theoretically grounded strategies: stage-level pruning via early stopping with partial estimation, and layer-to-stage mapping exploiting a ridge-like distribution pattern. These strategies reduce the search space without sacrificing optimal training efficiency. Experiments on heterogeneous clusters show that Poseidon improves training throughput by up to $2.76\times$ over state-of-the-art systems.

cs.DC↗

arXiv2026

psRL: Efficient Training for Agentic AI via Training-Time Prefix Sharing

In modern agentic AI training, the system bottleneck is shifting from rollout to update. Emerging sampling strategies such as tree-structured and step-wise RL greatly increase training sample volume while incurring relatively low marginal rollout cost, causing the update phase to dominate the end-to-end execution time. Crucially, this shift exposes a new optimization opportunity, as production traces reveal substantial prefix redundancy across training samples. In this paper, we propose psRL (prefix sharing for RL), a new training system for agentic AI designed to exploit prefix redundancy among training samples. Leveraging the global visibility and data immutability inherent to the update phase, psRL achieves efficient workload scheduling and memory management for distributed training. Specifically, psRL introduces two novel prefix-sharing mechanisms that enable flexible, fine-grained workload distribution across GPU workers, simultaneously optimizing prefix reuse and achieving load balancing. Moreover, psRL implements a new underlying KV cache manager that facilitates adaptable block-size allocation and dynamic KV caching, maximizing memory utilization while maintaining a high prefix hit rate. Evaluations using production traces demonstrate that psRL outperforms existing systems by up to 5.2x in throughput. The source code will be publicly available soon.

cs.DC↗

arXiv2026

Cloud-native and Distributed Systems for Efficient and Scalable Large Language Models -- A Research Agenda

The rapid rise of Large Language Models (LLMs) has revolutionized various artificial intelligence (AI) applications, from natural language processing to code generation. However, the computational demands of these models, particularly in training and inference, present significant challenges. Traditional systems are often unable to meet these requirements, necessitating the integration of cloud-native and distributed architectures. This paper explores the role of cloud platforms and distributed systems in supporting the scalability, efficiency, and optimization of LLMs. We discuss the complexities of LLM deployment, including data management, resource optimization, and the need for microservices, autoscaling, and hybrid cloud-edge solutions. Additionally, we examine emerging research trends, such as serverless inference, quantum computing, and federated learning, and their potential to drive the next phase of LLM innovation. The paper concludes with a roadmap for future developments, emphasizing the need for continued research, standardization, and cross-sector collaboration to sustain the growth of LLMs in both research and enterprise applications.

cs.DC↗

arXiv2026

Serving Hybrid LLM Loads with SLO Guarantees Using CPU-GPU Attention Piggybacking

Nowadays, service providers often deploy multiple types of LLM services within shared clusters. While the service colocation improves resource utilization, it introduces significant interference risks for latency-sensitive (LS) services-which have strict SLO requirements for inference latency-and severely constrain the service capacity of best-effort (BE) services due to limited available memory. To address interference, existing systems typically rely on reserving headroom to constrain BE resource usage. However, this approach's coarse granularity compromises the SLO compliance of the latency-sensitive service and unnecessarily restricts the generation potential of the best effort service. In this paper, we propose OmniServe, a novel LLM serving system that efficiently harnesses both CPU and GPU resources to mitigate interference and improve throughput. Central to OmniServe is the Attention Piggybacking mechanism, which effectively offloads the Attention computation of BE services to CPUs on the fly. This mechanism also facilitates asynchronous communication between CPU and GPU streams, preventing GPUs from being blocked while aggregating Attention results. Additionally, OmniServe incorporates a dynamic batching control policy to adapt to fluctuating request arrivals, facilitating Dense module computation using layer-wise batching. Experimental results show that OmniServe improves the SLO attainment rate for LS services by up to $1.48\times$ while enhancing BE serving throughput by up to $9.85\times$ compared to state-of-the-art systems.

cs.DC↗

arXiv2025

Hetis: Serving LLMs in Heterogeneous GPU Clusters with Fine-grained and Dynamic Parallelism

The significant resource demands in LLM serving prompts production clusters to fully utilize heterogeneous hardware by partitioning LLM models across a mix of high-end and low-end GPUs. However, existing parallelization approaches often struggle to scale efficiently in heterogeneous environments due to their coarse-grained and static parallelization strategies. In this paper, we introduce Hetis, a new LLM system tailored for heterogeneous GPU clusters. Hetis addresses two critical challenges: (1) memory inefficiency caused by the mismatch between memory capacity and computational power in heterogeneous devices, and (2) computational inefficiency arising from performance gaps across different LLM modules. To tackle these issues, Hetis employs a fine-grained and dynamic parallelism design. Specifically, it selectively parallelizes compute-intensive operations to reduce latency and dynamically distributes Attention computations to low-end GPUs at a head granularity, leveraging the distinct characteristics of each module. Additionally, Hetis features an online load dispatching policy that continuously optimizes serving performance by carefully balancing network latency, computational load, and memory intensity. Evaluation results demonstrate that Hetis can improve serving throughput by up to $2.25\times$ and reduce latency by $1.49\times$ compared to existing systems.

cs.DC↗

arXiv2024

Optimal Resource Efficiency with Fairness in Heterogeneous GPU Clusters

Ensuring the highest training throughput to maximize resource efficiency, while maintaining fairness among users, is critical for deep learning (DL) training in heterogeneous GPU clusters. However, current DL schedulers provide only limited fairness properties and suboptimal training throughput, impeding tenants from effectively leveraging heterogeneous resources. The underlying design challenge stems from inherent conflicts between efficiency and fairness properties. In this paper, we introduce OEF, a new resource allocation framework specifically developed for achieving optimal resource efficiency and ensuring diverse fairness properties in heterogeneous GPU clusters. By integrating resource efficiency and fairness within a global optimization framework, OEF is capable of providing users with maximized overall efficiency, as well as various guarantees of fairness, in both cooperative and non-cooperative environments. We have implemented OEF in a cluster resource manager and conducted large-scale experiments, showing that OEF can improve the overall training throughput by up to 32% while improving fairness compared to state-of-the-art heterogeneity-aware schedulers.

cs.DC↗