Elamur Research⌕
☰
home / papers / 2025 / Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity

Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity

Paper · 2025 · Yehui Tang, Xiaosong Li, Fangcheng Liu et al.

Summary

Sparse Mixture-of-Experts LLMs are cheap per token. However, when experts are spread across accelerators, uneven routing overloads some devices, and the slowest device limits overall speed. The Pangu Team at Huawei proposes Mixture of Grouped Experts (MoGE). Experts are split into equal, non-overlapping groups, typically one per device. A global softmax router scores all experts, and each token then takes the same number of top-scoring experts from every group. Device load is therefore balanced by design, with an Imbalance Score of 0. On this basis the team built Pangu Pro MoE, a 72B-parameter model with 16B parameters active per token. Its configuration was tuned through system simulation for Ascend 300I Duo and 800I A2 NPUs. The model was pre-trained on 13 trillion tokens using 4K Ascend NPUs and refined with SFT and RL. For serving, it uses hybrid parallelism, quantization and custom kernels (MulAttention, SwiftGMM). It achieves high inference throughput and beats GLM-Z1-32B, Qwen3-32B and Gemma3-27B on many benchmarks. This shows that co-designing the architecture with the hardware makes sparse models practical on Ascend.

Key points

  • Inference on Ascend 800I A2 reaches 1148 tokens/s per card, or 1528 tokens/s with speculative decoding. This exceeds comparable 32B and 72B dense models.
  • A Monte Carlo study of standard Top-K routing (N = 64, K = 8, M = 8, batch of 16 tokens) shows that some device imbalance (IS > 0) occurs almost certainly.
  • Within each group, every expert receives roughly 12.5% of tokens, the ideal for top-1 selection among 8 experts. By comparison, the busiest expert in DeepSeek-V2 handles up to 30% of tokens (measured on a 0.5% sample of the C4 validation set).
  • Co-activation analysis indicates low redundancy between experts, with somewhat more cross-expert collaboration in shallow and deep layers than in middle ones.
  • The model and code are released via gitcode.com/ascend-tribe/pangu-pro-moe (arXiv:2505.21411).

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksComputer Vision and Pattern Recognition · 2,869 worksHardware and Architecture · 81 works

Cite

Cite as Yehui Tang, Xiaosong Li, Fangcheng Liu, Wei Guo, Hang Zhou, Yaoyuan Wang, Kai Han, Xianzhi Yu, Jinpeng Li, Hui Zang, Fei Mi, Xiaojun Meng, Zhicheng Liu, Hanting Chen, Binfan Zheng, Can Chen, Youliang Yan, Ruiming Tang, Peifeng Qin, Xinghao Chen, Dacheng Tao and Yunhe Wang (and Other Contributors). 2025. Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity. arXiv:2505.21411.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2505.21411 · arXiv 2505.21411

@misc{tang2025pangu,
  title = {{Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity}},
  author = {Yehui Tang, Xiaosong Li, Fangcheng Liu, Wei Guo, Hang Zhou, Yaoyuan Wang, Kai Han, Xianzhi Yu, Jinpeng Li, Hui Zang, Fei Mi, Xiaojun Meng, Zhicheng Liu, Hanting Chen, Binfan Zheng, Can Chen, Youliang Yan, Ruiming Tang, Peifeng Qin, Xinghao Chen, Dacheng Tao and Yunhe Wang (and Other Contributors)},
  year = {2025},
  doi = {10.48550/arxiv.2505.21411},
  eprint = {2505.21411},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2505.21411}
}

References

49 references, 29 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Meta AI. The llama 4 herd: The beginning of a new era of natively multimodal ai innovation. https://ai.meta.com/blog/llama-4-multimodal-intelligence/, 2025. Accessed: 2025-04-05. ↗
  2. Jacob Austin, Augustus Odena, Maxwell Nye, Maarten Bosma, Henryk Michalewski, David Dohan, Ellen Jiang, Carrie J. Cai, Michael Terry, Quoc V. Le, and Charles Sutton. Program synthesis with large language models. ArXiv, abs/2108.07732, 2021.
  3. Yonatan Bisk, Rowan Zellers, Ronan Le Bras, Jianfeng Gao, and Yejin Choi. Piqa: Reasoning about physical commonsense in natural language. In AAAI Conference on Artificial Intelligence, 2019.
  4. Weilin Cai, Juyong Jiang, Fan Wang, Jing Tang, Sunghun Kim, and Jiayi Huang. A survey on mixture of experts. arXiv preprint arXiv:2407.06204, 2024. find
  5. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Pondé, Jared Kaplan, Harrison Edwards, Yura Burda, Nicholas Joseph, Greg Brockman, Alex Ray, Raul Puri, Gretchen Krueger, Michael Petrov, Heidy Khlaaf, Girish Sastry, Pamela Mishkin, Brooke Chan, Scott Gray, Nick Ryder, Mikhail Pavlov, Aleth… — Evaluating Large Language Models Trained on Code
  6. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168, 2021.
  7. Yiming Cui, Ting Liu, Wanxiang Che, Li Xiao, Zhipeng Chen, Wentao Ma, Shijin Wang, and Guoping Hu. A span-extraction dataset for Chinese machine reading comprehension. In Kentaro Inui, Jing Jiang, Vincent Ng, and Xiaojun Wan, editors, Proceedings of the 2019 Conference on Empirical Methods in Natura…
  8. Nan Du, Yanping Huang, Andrew M Dai, Simon Tong, Dmitry Lepikhin, Yuanzhong Xu, Maxim Krikun, Yanqi Zhou, Adams Wei Yu, Orhan Firat, et al. Glam: Efficient scaling of language models with mixture-of-experts. In International Conference on Machine Learning, pages 5547–5569. PMLR, 2022.
  9. Dheeru Dua, Yizhong Wang, Pradeep Dasigi, Gabriel Stanovsky, Sameer Singh, and Matt Gardner. Drop: A reading comprehension benchmark requiring discrete reasoning over paragraphs. In North American Chapter of the Association for Computational Linguistics, 2019.
  10. Suyu Ge, Yunan Zhang, Liyuan Liu, Minjia Zhang, Jiawei Han, and Jianfeng Gao. Model tells you what to discard: Adaptive kv cache compression for llms. arXiv preprint arXiv:2310.01801, 2023.
  11. Team GLM, Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, Jing Zhang, Juanzi Li, Lei Zhao, Lindong Wu, Lucen Zhong, Mingdao Liu, Minlie Huang, Peng Zhang, Qinka… — ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools
  12. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. Measuring massive multitask language understanding. arXiv preprint arXiv:2009.03300, 2020.
  13. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. Measuring mathematical problem solving with the math dataset. arXiv preprint arXiv:2103.03874, 2021.
  14. Yuzhen Huang, Yuzhuo Bai, Zhihao Zhu, Junlei Zhang, Jinghan Zhang, Tangjun Su, Junteng Liu, Chuancheng Lv, Yikai Zhang, Jiayi Lei, Fanchao Qi, Yao Fu, Maosong Sun, and Junxian He. C-eval: A multi-level multi-discipline chinese evaluation suite for foundation models. ArXiv, abs/2305.08322, 2023.
  15. Naman Jain, King Han, Alex Gu, Wen-Ding Li, Fanjia Yan, Tianjun Zhang, Sida Wang, Armando Solar-Lezama, Koushik Sen, and Ion Stoica. Livecodebench: Holistic and contamination free evaluation of large language models for code. arXiv preprint arXiv:2403.07974, 2024.
  16. Albert Q Jiang, Alexandre Sablayrolles, Antoine Roux, Arthur Mensch, Blanche Savary, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Emma Bou Hanna, Florian Bressand, et al. Mixtral of experts. arXiv preprint arXiv:2401.04088, 2024. find
  17. Guokun Lai, Qizhe Xie, Hanxiao Liu, Yiming Yang, and Eduard H. Hovy. Race: Large-scale reading comprehension dataset from examinations. ArXiv, abs/1704.04683, 2017.
  18. Dmitry Lepikhin, HyoukJoong Lee, Yuanzhong Xu, Dehao Chen, Orhan Firat, Yanping Huang, Maxim Krikun, Noam Shazeer, and Zhifeng Chen. {GS}hard: Scaling giant models with conditional computation and automatic sharding. In International Conference on Learning Representations, 2021. find
  19. Haonan Li, Yixuan Zhang, Fajri Koto, Yifei Yang, Hai Zhao, Yeyun Gong, Nan Duan, and Timothy Baldwin. Cmmlu: Measuring massive multitask language understanding in chinese. arXiv preprint arXiv:2306.09212, 2023.
  20. Tianle Li, Wei-Lin Chiang, Evan Frick, Lisa Dunlap, Banghua Zhu, Joseph E. Gonzalez, and Ion Stoica. From live data to high-quality benchmarks: The arena-hard pipeline, April 2024. find
  21. Wenhao Li, Fanchao Qi, Maosong Sun, Xiaoyuan Yi, and Jiarui Zhang. Ccpm: A chinese classical poetry matching dataset, 2021. find
  22. Xing Li, Zeyu Xing, Yiming Li, Linping Qu, Hui-Ling Zhen, Wulong Liu, Yiwu Yao, Sinno Jialin Pan, and Mingxuan Yuan. Kvtuner: Sensitivity-aware layer-wise mixed precision kv cache quantization for efficient and nearly lossless llm inference, 2025. find
  23. Heng Liao, Jiajin Tu, Jing Xia, and Xiping Zhou. Davinci: A scalable architecture for neural network computing. In 2019 IEEE Hot Chips 31 Symposium (HCS), pages 1–44. IEEE Computer Society, 2019. find
  24. Aixin Liu, Bei Feng, Bin Wang, Bingxuan Wang, Bo Liu, Chenggang Zhao, Chengqi Dengr, Chong Ruan, Damai Dai, Daya Guo, et al. Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model. arXiv preprint arXiv:2405.04434, 2024. find
  25. Zirui Liu, Jiayi Yuan, Hongye Jin, Shaochen Zhong, Zhaozhuo Xu, Vladimir Braverman, Beidi Chen, and Xia Hu. Kivi: A tuning-free asymmetric 2bit quantization for kv cache. arXiv preprint arXiv:2402.02750, 2024.
  26. MAA. Codeforces. American Invitational Mathematics Examination - AIME 2024, 2024. https://maa.org/math-competitions/american-invitational-mathematics-examination-aime. ↗
  27. MAA. Codeforces. American Invitational Mathematics Examination - AIME 2025, 2025. https://maa.org/math-competitions/american-invitational-mathematics-examination-aime. ↗
  28. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R Bowman. Gpqa: A graduate-level google-proof q&a benchmark. In First Conference on Language Modeling, 2024.
  29. Keisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavatula, and Yejin Choi. Winogrande: An adversarial winograd schema challenge at scale, 2019. find
  30. Noam Shazeer, Azalia Mirhoseini, Krzysztof Maziarz, Andy Davis, Quoc Le, Geoffrey Hinton, and Jeff Dean. Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. arXiv preprint arXiv:1701.06538, 2017. find
  31. Freda Shi, Mirac Suzgun, Markus Freitag, Xuezhi Wang, Suraj Srivats, Soroush Vosoughi, Hyung Won Chung, Yi Tay, Sebastian Ruder, Denny Zhou, Dipanjan Das, and Jason Wei. Language models are multilingual chain-of-thought reasoners. In The Eleventh International Conference on Learning Representations,…
  32. Kai Sun, Dian Yu, Dong Yu, and Claire Cardie. Investigating prior knowledge for challenging chinese machine reading comprehension, 2019.
  33. Mirac Suzgun, Nathan Scales, Nathanael Scharli, Sebastian Gehrmann, Yi Tay, Hyung Won Chung, Aakanksha Chowdhery, Quoc V. Le, Ed H. Chi, Denny Zhou, and Jason Wei. Challenging big-bench tasks and whether chain-of-thought can solve them. In Annual Meeting of the Association for Computational Linguist…
  34. Yehui Tang, Yichun Yin, Yaoyuan Wang, Hang Zhou, Yu Pan, Wei Guo, Ziyang Zhang, Miao Rang, Fangcheng Liu, Naifu Zhang, Binghan Li, Yonghan Dong, Xiaojun Meng, Yasheng Wang, Dong Li, Yin Li, Dandan Tu, Can Chen, Youliang Yan, Fisher Yu, Ruiming Tang, Yunhe Wang, Botian Huang, Bo Wang, Boxiao Liu, Cha… — Pangu Ultra MoE: How to Train Your Big MoE on Ascend NPUs
  35. Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, et al. Gemma 3 technical report. arXiv preprint arXiv:2503.19786, 2025. find
  36. M-A-P Team, Xinrun Du, Yifan Yao, Kaijing Ma, Bingli Wang, Tianyu Zheng, Kang Zhu, Minghao Liu, Yiming Liang, Xiaolong Jin, Zhenlin Wei, Chujie Zheng, Kaixing Deng, Shuyue Guo, Shian Jia, Sichao Jiang, Yiyan Liao, Rui Li, Qinrui Li, Sirun Li, Yizhi Li, Yunwen Li, Dehua Ma, Yuansheng Ni, Haoran Que,… find
  37. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971, 2023.
  38. Yubo Wang, Xueguang Ma, Ge Zhang, Yuansheng Ni, Abhranil Chandra, Shiguang Guo, Weiming Ren, Aaran Arulraj, Xuan He, Ziyan Jiang, et al. Mmlu-pro: A more robust and challenging multi-task language understanding benchmark. In The Thirty-eight Conference on Neural Information Processing Systems Datase…
  39. Yunhe Wang, Hanting Chen, Yehui Tang, Tianyu Guo, Kai Han, Ying Nie, Xutao Wang, Hailin Hu, Zheyuan Bai, Yun Wang, et al. Pangu-π: Enhancing language model architectures via nonlinearity compensation. arXiv preprint arXiv:2312.17276, 2023.
  40. Tianwen Wei, Jian Luan, Wei Liu, Shuang Dong, and Bin Wang. Cmath: Can your language model pass chinese elementary school math test?, 2023.
  41. Guangxuan Xiao, Yuandong Tian, Beidi Chen, Song Han, and Mike Lewis. Efficient streaming language models with attention sinks. arXiv preprint arXiv:2309.17453, 2023. find
  42. Liang Xu, Hai Hu, Xuanwei Zhang, Lu Li, Chenjie Cao, Yudong Li, Yechen Xu, Kai Sun, Dian Yu, Cong Yu, et al. Clue: A chinese language understanding evaluation benchmark. arXiv preprint arXiv:2004.05986, 2020.
  43. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. Qwen3 technical report. arXiv preprint arXiv:2505.09388, 2025.
  44. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, et al. Qwen2. 5 technical report. arXiv preprint arXiv:2412.15115, 2024. find
  45. Qingyue Yang, Jie Wang, Xing Li, Zhihai Wang, Chen Chen, Lei Chen, Xianzhi Yu, Wulong Liu, Jianye Hao, Mingxuan Yuan, et al. Attentionpredictor: Temporal pattern matters for efficient llm inference. arXiv preprint arXiv:2502.04077, 2025. — [arXiv:2502.04077](https://arxiv.org/abs/2502.04077) find
  46. Mingjia Yin, Chuhan Wu, Yufei Wang, Hao Wang, Wei Guo, Yasheng Wang, Yong Liu, Ruiming Tang, Defu Lian, and Enhong Chen. Entropy law: The story behind data compression and llm performance. arXiv preprint arXiv:2407.06645, 2024. find
  47. Yichun Yin, Wenyong Huang, Kaikai Song, Yehui Tang, Xue-Fei Wu, Wei Guo, Peng Guo, Yaoyuan Wang, Xiaojun Meng, Yasheng Wang, Dong Li, Can Chen, Dandan Tu, Yin Li, Fisher Yu, Ruiming Tang, Yunhe Wang, Baojun Wang, Bin Wang, Bo Wang, Boxiao Liu, Changzheng Zhang, Duyu Tang, Fei Mi, Hui Jin, Jiansheng… — Pangu Ultra: Pushing the Limits of Dense Large Language Models on Ascend NPUs
  48. Rowan Zellers, Ari Holtzman, Yonatan Bisk, Ali Farhadi, and Yejin Choi. Hellaswag: Can a machine really finish your sentence? In Annual Meeting of the Association for Computational Linguistics, 2019. find
  49. Jeffrey Zhou, Tianjian Lu, Swaroop Mishra, Siddhartha Brahma, Sujoy Basu, Yi Luan, Denny Zhou, and Le Hou. Instruction-following evaluation for large language models. arXiv preprint arXiv:2311.07911, 2023.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/tang-2025-pangu-pro-moe-mixture/