🐯虎嗅•Aug 20, 2026Freshcollected in 30m

Moore Threads’ Growth Comes with Big Questions

#gpu #domestic-chips #cash-flow #software-ecosystemmoore-threads-gpu-platformmoore-threads musa cambrian cuda

💡Moore Threads’ revenue surged, but subsidies, customer concentration, inventory, and cash flow tell a less comfortable s

⚡ 30-Second TL;DR

What Changed

Cloud-computing products generated RMB 1.693 billion, or 97.5% of first-half revenue.

Why It Matters

For AI infrastructure buyers and founders, Moore Threads’ report highlights the execution risks behind domestic GPU adoption: supply concentration, uncertain inventory absorption, and immature software ecosystems. Hardware availability alone is not enough; workload portability, software compatibility, and vendor financial resilience should be evaluated before committing to large deployments.

What To Do Next

Before adopting MUSA GPUs, benchmark your actual training and inference workloads against CUDA using operator-coverage, performance, migration effort, and fallback-cost checklists.

Who should care:Enterprise & Security Teams

Key Points

•Cloud-computing products generated RMB 1.693 billion, or 97.5% of first-half revenue.
•Customer concentration remains high: the top five customers represented 91.36% of 2025 sales, while related receivables and contract assets reached 99% in the first half of 2026.
•Excluding RMB 139 million in non-recurring gains, including subsidies and financial-asset gains, first-half net loss was still RMB 151 million.
•Inventory and prepayments reached nearly RMB 4.9 billion, while operating cash flow was negative RMB 2.169 billion.
•Compared with Cambrian’s positive cash flow and sustained profitability, Moore Threads illustrates the growing financial divergence among domestic GPU companies.