Phantora: Live GPU Cluster Simulation for Machine Learning System Performance Estimation
- URL: http://arxiv.org/abs/2505.01616v1
- Date: Fri, 02 May 2025 22:36:24 GMT
- Title: Phantora: Live GPU Cluster Simulation for Machine Learning System Performance Estimation
- Authors: Jianxing Qin, Jingrong Chen, Xinhao Kong, Yongji Wu, Liang Luo, Zhaodong Wang, Ying Zhang, Tingjun Chen, Alvin R. Lebeck, Danyang Zhuo,
- Abstract summary: Phantora is a live GPU cluster simulator for performance estimation.<n>It overcomes several research challenges in integrating an event-driven network simulator with live system execution.<n>Our evaluation results show that Phantora can deliver similar estimation accuracy to the state-of-the-art workload simulation approach with only one GPU.
- Score: 11.48166268734119
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: To accommodate ever-increasing model complexity, modern machine learning (ML) systems have to scale to large GPU clusters. Changes in ML model architecture, ML system implementation, and cluster configuration can significantly affect overall ML system performance. However, quantifying the performance impact before deployment is challenging. Existing performance estimation methods use performance modeling or static workload simulation. These techniques are not general: they requires significant human effort and computation capacity to generate training data or a workload. It is also difficult to adapt ML systems to use these techniques. This paper introduces, Phantora, a live GPU cluster simulator for performance estimation. Phantora runs minimally modified ML models and frameworks, intercepting and simulating GPU-related operations to enable high-fidelity performance estimation. Phantora overcomes several research challenges in integrating an event-driven network simulator with live system execution, and introduces a set of techniques to improve simulation speed, scalability, and accuracy. Our evaluation results show that Phantora can deliver similar estimation accuracy to the state-of-the-art workload simulation approach with only one GPU, while reducing human effort and increasing generalizability.
Related papers
- MoE-Lens: Towards the Hardware Limit of High-Throughput MoE LLM Serving Under Resource Constraints [7.287566040274871]
MoE-Lens is an inference system designed through holistic performance modeling for resource-constrained environments.<n>It captures the system execution mechanisms to identify the key hardware bottlenecks and accurately predict the achievable throughput.<n> evaluated on diverse MoE models and datasets, MoE-Lens outperforms the state-of-the-art solution by 4.6x on average (up to 25.5x)
arXiv Detail & Related papers (2025-04-12T21:26:56Z) - Fake Runs, Real Fixes -- Analyzing xPU Performance Through Simulation [4.573673188291683]
We present xPU-Shark, a fine-grained methodology for analyzing ML models at the machine-code level.<n>xPU-Shark captures traces from production deployments running on accelerators and replays them in a modified microarchitecture simulator.<n>We optimize a common communication collective by up to 15% and reduce token generation latency by up to 4.1%.
arXiv Detail & Related papers (2025-03-18T23:15:02Z) - AutoHete: An Automatic and Efficient Heterogeneous Training System for LLMs [68.99086112477565]
Transformer-based large language models (LLMs) have demonstrated exceptional capabilities in sequence modeling and text generation.<n>Existing heterogeneous training methods significantly expand the scale of trainable models but introduce substantial communication overheads and CPU workloads.<n>We propose AutoHete, an automatic and efficient heterogeneous training system compatible with both single- GPU and multi- GPU environments.
arXiv Detail & Related papers (2025-02-27T14:46:22Z) - DSMoE: Matrix-Partitioned Experts with Dynamic Routing for Computation-Efficient Dense LLMs [70.91804882618243]
This paper proposes DSMoE, a novel approach that achieves sparsification by partitioning pre-trained FFN layers into computational blocks.<n>We implement adaptive expert routing using sigmoid activation and straight-through estimators, enabling tokens to flexibly access different aspects of model knowledge.<n>Experiments on LLaMA models demonstrate that under equivalent computational constraints, DSMoE achieves superior performance compared to existing pruning and MoE approaches.
arXiv Detail & Related papers (2025-02-18T02:37:26Z) - DoMINO: A Decomposable Multi-scale Iterative Neural Operator for Modeling Large Scale Engineering Simulations [2.300471499347615]
DoMINO is a point cloudbased machine learning model that uses local geometric information to predict flow fields on discrete points.<n>DoMINO is validated for the automotive aerodynamics use case using the DrivAerML dataset.
arXiv Detail & Related papers (2025-01-23T03:28:10Z) - The Artificial Scientist -- in-transit Machine Learning of Plasma Simulations [33.024345484180024]
We demonstrate a streaming workflow in which simulation data is streamed directly to a machine-learning (ML) framework.<n>With the presented workflow, data operations can be performed in common and easy-to-use programming languages.
arXiv Detail & Related papers (2025-01-06T20:58:27Z) - GausSim: Foreseeing Reality by Gaussian Simulator for Elastic Objects [55.02281855589641]
GausSim is a novel neural network-based simulator designed to capture the dynamic behaviors of real-world elastic objects represented through Gaussian kernels.<n>We leverage continuum mechanics and treat each kernel as a Center of Mass System (CMS) that represents continuous piece of matter.<n>In addition, GausSim incorporates explicit physics constraints, such as mass and momentum conservation, ensuring interpretable results and robust, physically plausible simulations.
arXiv Detail & Related papers (2024-12-23T18:58:17Z) - Numerical Pruning for Efficient Autoregressive Models [87.56342118369123]
This paper focuses on compressing decoder-only transformer-based autoregressive models through structural weight pruning.<n>Specifically, we propose a training-free pruning method that calculates a numerical score with Newton's method for the Attention and modules, respectively.<n>To verify the effectiveness of our method, we provide both theoretical support and extensive experiments.
arXiv Detail & Related papers (2024-12-17T01:09:23Z) - In Situ Framework for Coupling Simulation and Machine Learning with
Application to CFD [51.04126395480625]
Recent years have seen many successful applications of machine learning (ML) to facilitate fluid dynamic computations.
As simulations grow, generating new training datasets for traditional offline learning creates I/O and storage bottlenecks.
This work offers a solution by simplifying this coupling and enabling in situ training and inference on heterogeneous clusters.
arXiv Detail & Related papers (2023-06-22T14:07:54Z) - Accelerated Policy Learning with Parallel Differentiable Simulation [59.665651562534755]
We present a differentiable simulator and a new policy learning algorithm (SHAC)
Our algorithm alleviates problems with local minima through a smooth critic function.
We show substantial improvements in sample efficiency and wall-clock time over state-of-the-art RL and differentiable simulation-based algorithms.
arXiv Detail & Related papers (2022-04-14T17:46:26Z) - SimNet: Computer Architecture Simulation using Machine Learning [3.7019798164954336]
This work describes a concerted effort, where machine learning (ML) is used to accelerate discrete-event simulation.
A GPU-accelerated parallel simulator is implemented based on the proposed instruction latency predictor.
Its simulation accuracy and throughput are validated and evaluated against a state-of-the-art simulator.
arXiv Detail & Related papers (2021-05-12T17:31:52Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.