NVDA | Ming-Chi Kuo Predicts Nvidia LPU Shipments of 5 Million Units, Showing Over 10x Growth

robot
Abstract generation in progress

Tianfeng International Securities analyst Guo Mingchi stated that after Nvidia invested in the U.S. AI chip startup Groq, it significantly raised its shipment forecast for its Language Processing Unit (LPU). The total shipments are expected to reach 4 to 5 million units between 2026 and 2027, representing more than a tenfold increase compared to previous years.

Guo Mingchi posted on his social media that the rapid growth in LPU demand is mainly due to two factors. First, the high integration with Nvidia’s ecosystem, such as CUDA, greatly reduces the barriers to application development and deployment. Second, the demand for ultra-low latency inference is rapidly increasing, including AI agents like coding agents and emerging applications like real-time and physical AI.

Nvidia plans to increase the number of LPUs per rack from the current 64 to 256 to expand memory capacity and maintain ultra-low latency performance. The new architecture racks are expected to be mass-produced from the fourth quarter of this year to the first quarter of next year. Shipments of racks are projected to be about 3,000 to 5,000 units in 2026 and approximately 15,000 to 20,000 units in 2027.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments