· 个股论点

NBIS是英伟达应对超大规模厂商自研芯片威胁的关键,短期增长潜力最大。

涉及标的:

中文翻译

我同意,超大规模云计算厂商(Hyperscaler)自研芯片是对 $NVDA 4万亿美元帝国及GPU供应商的最大威胁。但 $NBIS 和 $CRWV 是英伟达保护其护城河(Moat)并防止GPU利润率压缩的解决方案。 因此,他们有意支持像 NBIS 这样的公司,以分散来自超大规模云计算厂商的需求,并通过将依赖关系分散/限制计算资源给这些较小的 GPU 即服务(GPU as a Service)公司,来防止来自 Azure、AWS 的利润率压缩。 我们已经看到 $MSFT 在 OpenAI 的计算资源耗尽,因此他们不得不转向 GOOGL、CRWV、NBIS 等公司并建立多年期合同,即使他们也在构建定制 ASIC。 $NBIS 也不依赖于单一合同,因为其他超大规模云计算厂商正在接入,但那项170亿美元的大合同确实给了它更多的保障。我们也尚未看到其被完全定价,因为在此之前估值约为160亿美元+,而在稀释后现在为240亿美元。 我提到1年的时间框架是因为其上行空间最大,三次降息也有助于此。即使只有两次降息,论点依然成立,因为 NVDA 支持 NBIS + 超大规模云计算厂商的资本支出(Capex)流向这些 GPU 供应商。 3-4年后,我可能会更担心,这就是为什么我一直维持长期持有 $RKLB、$IBIT 等作为5年期的头寸,但 NBIS 在未来一年具有最大的增长潜力。

英文原文

I agree, hyperscaler chips is the biggest threat to $NVDA's 4T+ empire + GPU providers. But, $NBIS and $CRWV are NVIDIA's solution to protecting it's moat and GPU margin compression. So they're supporting companies like NBIS on purpose to diversify demand from hyperscalers and prevent margin compression from Azure, AWS by spreading out dependencies/limit compute to these smaller GPU as a service companies. We already saw $MSFT run out of compute for OpenAI so they had to go to GOOGL, CRWV, NBIS and others and build out multi-year contracts, even while building out custom ASICs. $NBIS also doesn't hinge on one contract since other hyperscalers are on its way but that 17B big contract does give it a lot more assurance. We also haven't seen it fully priced in yet since valuation was ~16B+ before then and now it's $24B after dilution. I mentioned 1 year timeframe in terms of heaviest upside with triple rate cut helping too. Double rate cut, thesis still stands with NVDA supporting NBIS + capex spend from hyperscalers flowing down to these GPU providers. 3-4 years out, I might be a bit more worried, which is why I always maintained I'm long $RKLB, $IBIT and others on a 5 year timeframe but NBIS has the biggest growth potential over the next year.

在 X 上查看原推 ↗