news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

nvidia may suspend the development of the dual-cabinet version gb200 (nvl36×2) ai server.

well-known analyst ming-chi kuo said that in the absence of customer customization requirements, nvidia will no longer provide the dual-cabinet version of the gb200 (nvl36×2) ai server, and will only provide the single-cabinet version of the gb200 nvl72 and nvl36. the reason for nvidia's move is said to be due to limited resources and customer preferences.

initially, nvidia plans to develop three ai server cabinets based on blackwell gpu gb200 for ai and high-performance computing (hpc) workloads, namely nvl36, nvl72 and nvl36×2. however, managing all three projects became challenging, especially with two different 72 gpu versions (nvl72 and nvl36×2) being developed simultaneously. therefore, nvidia is now only focusing on nvl72 and nvl36.

kuo said the move won't affect artificial intelligence or nvidia's long-term trends, but it does raise questions about the company and its supply chain's ability to execute.

he added: “in my opinion, nvidia’s recent frequent changes to its ai server product roadmap reflect their attempts to achieve a better balance between supply chain execution, competitive advantage, and customer needs with limited resources. this is a matter of it's a good thing that nvidia is taking a more pragmatic approach to its product plans, but this shift may confuse some market participants about the supply chain changes."