· 供应链分析
这与 $NVDA 及其 GPU/ASIC（谷歌云 TPU、AWS Trainium）的 4T+ 论点相同，即超大规模云服务商能否取代英伟达。同样的逻辑也适用于 $NBIS。如果微软能如此快速地自建，他们不会签署 170 亿美元、为期 5 年的合同。如果你阅读了我的深度研究(DD)，NBIS 和 $CRWV 是英伟达保持护城河的战略解决方案。CRWV 和 NBIS 有助于防止超大规模云服务商掌握全部议价权，并阻止 Azure 等未来整合算力并压缩利润率。NBIS 有助于维持英伟达的护城河，这也是数百亿美元的合同收入开始从“七巨头”流向新云提供商的原因。250 亿美元的市场资本化与远期收入不匹配。
That's the same 4T+ thesis with $NVDA and their GPUs/ASICs (Google Cloud TPUs, AWS Trainium) when hyperscalers could replace Nvidia. Same applies to $NBIS. MSFT wouldn't sign a 17B 5 year contract if they could build it out themselves this fast. If you read my DD, NBIS, $CRWV are Nvda's strategic solution to keeping their moat. CRWV, NBIS helps keep hyperscalers from having all the leverage + prevent Azure, others from consolidating compute and compressing margins down the road. NBIS helps maintain NVDA's moat and why hundreds of billions contracted revenue is already starting to flow from mag7 to the new neocloud providers. 25B marketcap does not align to forward revenue.