· 个股论点
是的，我同意这种拆解。但我觉得你关注错了重点（外部竞争优势）。 他们全栈软件(software full stack)加上涵盖初创企业、中小企业(SMBs)的多元化客户群，有助于优化自身的运营支出(opex)，这本身就是护城河(moat)。 你提到了像 $IREN 这样的矿企(miners)，但如果看看 $ORCL，他们在利润率以及未能转化为实际效益的建设/转型（如建设后的GPU利用率等）方面曾举步维艰。拥有其他软件层也会降低利润率，$CRWV 就亲身经历了这一点，这也是他们近期斥资数十亿美元进行收购的原因。 因此，拥有原始算力(raw capacity)并不总能转化为良好的利润率，就像 $NBIS 已经拥有的那样（如果我没记错的话，上次财报(ER)显示其毛利率为71.2%），这是全栈+运营效率+多元化中小企业客户带来的优势，而那些转型的矿企将不具备这一点。
Yeah I’d agree with the breakdown. But think you’re focusing on the wrong thing (external competitive advantage). Their full stack software + diversified customer base with startups, SMBs, helps their own opex, which is the moat in and of itself. You mention miners like $IREN but if we look at $ORCL they were struggling with margins and buildout/pivot that didn’t translate (gpu utilization after buildout, etc). Having other software layers also reduces margins and $CRWV found that out firsthand, which is why they spent billions on acquisitions recently. So having raw capacity doesn’t always translate to good profit margins like what $NBIS already has (71.2% gross from last ER, if I remember correctly) being full stack + operational efficiency + diversified smbs that miners that pivot won’t have.