PeopleofSuperintelligence

Builder

Chenggang Zhao

Chenggang Zhao works on AI infrastructure for large language models. In 2024, he earned a master's degree at Tsinghua under Mingyu Gao, joined DeepSeek, and coauthored DeepSeek-V2 and DeepSeek-V3. His 2025 work included DeepSeek-R1, the DeepEP communication library, and a study of the hardware behind V3 and R1. As of October 2026, he is a project lead for DeepGEMM-Ascend, a matrix-multiplication kernel library for Huawei Ascend NPUs.

Now at
DeepSeek
Works
24
Citations
2,571
h-index
12
Badges

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2024

    A Tsinghua master's degree

    In 2024, he completes a master's degree at Tsinghua with Mingyu Gao and moves to DeepSeek.

  2. 2024

    Coauthoring DeepSeek-V2

    He coauthors DeepSeek-V2 in 2024, a mixture-of-experts model with 236 billion total parameters and 21 billion activated for each token.

  3. 2024

    The DeepSeek-V3 report

    In 2024, he coauthors DeepSeek-V3, a mixture-of-experts model with 671 billion parameters and 37 billion activated for each token.

  4. 2025

    Reinforcement learning for DeepSeek-R1

    He coauthors the 2025 DeepSeek-R1 paper, which uses reinforcement learning to develop reasoning patterns such as self-reflection and verification.

  5. 2025

    Building the DeepEP library

    In 2025, he coauthors DeepEP, a communication library that supplies high-throughput and low-latency expert-parallel GPU kernels for mixture-of-experts training and inference.

  6. 2025

    DeepSeek-V3 hardware analysis

    A 2025 paper he coauthors analyzes the architecture and infrastructure behind DeepSeek-V3 and R1, including training on 2,048 NVIDIA H800 GPUs and hardware-aware model co-design.

  7. As of October 2026

    Leading DeepGEMM-Ascend

    In 2026, he is one of the project leads for DeepGEMM-Ascend, a matrix-multiplication kernel library for Huawei Ascend NPUs; he also contributes to its mHC kernel.

Career

  1. 2024–2025DeepSeek

Education

  1. 2024Tsinghua UniversityMaster's degree

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning2025 · Nature · 1,112 citationsWith Bingxuan Wang, Bochao Wu, Daya Guo and Zhibin Gou
  2. Can Open Large Language Models Catch Vulnerabilities?2025 · 538 citationsWith Bingxuan Wang, Bochao Wu and Zhibin Gou
  3. DeepSeek-V3 Technical Report2024 · 292 citationsWith Bingxuan Wang, Bochao Wu and Daya Guo
  4. DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models2024 · Annual Meeting of the Association for Computational Linguistics (ACL) · 282 citationsWith Zhenda Xie
  5. DeepSeek LLM: Scaling Open-Source Language Models with Longtermism2024 · 98 citationsWith Bingxuan Wang and Daya Guo

From OpenAlex, as of 19 October 2026.

  1. The median h-index is 54 at universities and 19 at frontier AI labsIncludes Chenggang Zhao. The median h-index is 54 at universities and 19 at frontier AI labs. We compare publishing careers, citations and moves to estimate how universities and frontier labs contribute to AI research.
  2. The company share of AI researchers’ top papers rose from 2% to 32%Includes DeepSeek. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 32% in 2020–2024.
  3. Google alumni founded or led 81 other organizations, the most among the labs we compareIncludes DeepSeek. We trace 194 organizations founded or led by AI lab alumni, and 202 career moves between the labs.
  4. Stanford University trained or employed 26 of the 108 AI major-prize winners, more than any other universityIncludes Tsinghua University. Stanford University trained or employed 26 of the 108 AI major-prize winners, more than any other university. We compare doctorate and employment links for major-prize and best-paper winners.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Chenggang Zhao yet. Send us a free-licensed photo.

  1. Mingyu Gao | Students people.iiis.tsinghua.edu.cn. Checked 19 October 2026.
  2. DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model arxiv.org. Checked 19 October 2026.
  3. DeepSeek-V3 Technical Report arxiv.org. Checked 19 October 2026.
  4. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning nature.com. Checked 19 October 2026.
  5. GitHub - deepseek-ai/DeepEP: DeepEP: an efficient expert-parallel communication library github.com. Checked 19 October 2026.
  6. Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures arxiv.org. Checked 19 October 2026.
  7. GitHub - deepseek-ai/DeepGEMM-Ascend: DeepGEMM-Ascend: clean and efficient matrix multiplication kernel library for Huawei Ascend NPUs github.com. Checked 19 October 2026.