DeepSeek expands Inner Mongolia data center operations ahead of V4 AI model launch
DeepSeek, a Chinese AI startup based in Hangzhou, is expanding its computing infrastructure with new hiring in Inner Mongolia ahead of its V4 model release. The company is recruiting server maintenance engineers and delivery managers in Ulanqab, Inner Mongolia, marking its first public job postings for on-site data center roles. This expansion comes more than a year after the company gained prominence with its R1 model. Industry speculation surrounds the chip sourcing strategy for V4, with conflicting reports suggesting potential use of Huawei's Ascend 950PR processor or Nvidia's Blackwell chips. The Information reported last week about Huawei chip usage, while Reuters indicated last month that training occurred on Nvidia's advanced processors. Previous recruitment focused on software engineers and AI researchers in Hangzhou and Beijing.
Key facts
- DeepSeek is hiring server maintenance engineers and delivery managers in Ulanqab, Inner Mongolia
- This marks the company's first public on-site data center infrastructure roles
- The hiring precedes the anticipated launch of DeepSeek's V4 model
- DeepSeek gained prominence with its R1 model release over a year ago
- The Information reported V4 might use Huawei's Ascend 950PR processor
- Reuters reported V4 training potentially used Nvidia's Blackwell chips
- Previous recruitment focused on software engineers and AI researchers in Hangzhou and Beijing
- DeepSeek is headquartered in Hangzhou
Entities
Institutions
- DeepSeek
- The Information
- Reuters
- Huawei
- Nvidia
Locations
- Inner Mongolia
- China
- Ulanqab
- Hangzhou
- Beijing
- United States