Unverified60% confidenceFactExact time
The upfront investment for building a hyperscale data center is typically $1–3 billion, and increasing GPU utilization from 50% to 80% can reduce the cost per unit of compute by approximately 40%.
2
Sources
60%
Confidence
Long-term
Relevance
8/5/2026
First Seen
Sources
Related Claims
Unverified大模型API的成本结构主要由GPU算力折旧（60-70%）、电力和冷却成本（15-20%）、带宽和运维开销（10-15%）构成76% similarUnverified根据 Ponemon Institute 的研究，数据中心单次非计划停机的平均成本从 2010 年的约 50 万美元上升到 2023 年的超过 90 万美元73% similarUnverifiedBeyond the break-even volume, the actual cost of a 30B-class model on a busy GPU is approximately $0.06–$0.85 per million tokens.73% similarUnverifiedThe communication overhead of large-scale GPU clusters grows approximately quadratically with the number of nodes.71% similarVerifiedElectricity costs for data centers typically account for 40%–60% of total operating costs.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/691013API
curl https://kongchang.com/api/v1/knowledge/claims/691013MCP
get_claim(id=691013)