ServeGen: Workload Characterization and Generation of Large Language Model Serving in Production
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917478459768832 |
|---|---|
| author | Xiang, Yuxing Li, Xue Qian, Kun Yu, Wenyuan Zhai, Ennan Jin, Xin |
| author_facet | Xiang, Yuxing Li, Xue Qian, Kun Yu, Wenyuan Zhai, Ennan Jin, Xin |
| contents | With the widespread adoption of Large Language Models (LLMs), serving LLM inference requests has become an increasingly important task, attracting active research advancements. Practical workloads play an essential role in this process: they are critical for motivating and benchmarking serving techniques and systems. However, the existing understanding of real-world LLM serving workloads is limited due to the lack of a comprehensive workload characterization. Prior analyses remain insufficient in scale and scope, thus failing to fully capture intricate workload characteristics.
In this paper, we fill the gap with an in-depth characterization of LLM serving workloads collected from our worldwide cloud inference serving service, covering not only language models but also emerging multimodal and reasoning models, and unveiling important new findings in each case. Moreover, based on our findings, we propose ServeGen, a principled framework for generating realistic LLM serving workloads by composing them on a per-client basis. A practical use case in production validates that ServeGen avoids 50% under-provisioning compared to naive workload generation, demonstrating ServeGen's advantage in performance benchmarking. ServeGen is available at https://github.com/alibaba/ServeGen. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_09999 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ServeGen: Workload Characterization and Generation of Large Language Model Serving in Production Xiang, Yuxing Li, Xue Qian, Kun Yu, Wenyuan Zhai, Ennan Jin, Xin Distributed, Parallel, and Cluster Computing With the widespread adoption of Large Language Models (LLMs), serving LLM inference requests has become an increasingly important task, attracting active research advancements. Practical workloads play an essential role in this process: they are critical for motivating and benchmarking serving techniques and systems. However, the existing understanding of real-world LLM serving workloads is limited due to the lack of a comprehensive workload characterization. Prior analyses remain insufficient in scale and scope, thus failing to fully capture intricate workload characteristics. In this paper, we fill the gap with an in-depth characterization of LLM serving workloads collected from our worldwide cloud inference serving service, covering not only language models but also emerging multimodal and reasoning models, and unveiling important new findings in each case. Moreover, based on our findings, we propose ServeGen, a principled framework for generating realistic LLM serving workloads by composing them on a per-client basis. A practical use case in production validates that ServeGen avoids 50% under-provisioning compared to naive workload generation, demonstrating ServeGen's advantage in performance benchmarking. ServeGen is available at https://github.com/alibaba/ServeGen. |
| title | ServeGen: Workload Characterization and Generation of Large Language Model Serving in Production |
| topic | Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2505.09999 |