· 供应链分析

SIVE 终局映射很夸张

涉及标的:

中文翻译

$SIVE 可能在 2027-2028 年映射到的所有超大规模云厂商,光按 9 亿美元市值来看就已经夸张了。 市场根本不懂接下来会发生什么。 从推测性映射来看: > $SIVE -> $POET -> $MRVL -> 1. $AMZN(来自 Celestial 的光子 fabric 采购协议 / warrants) 2. $MSFT(maia) 3. $GOOGL(最近和 Marvell 的开发对话) $SIVE 为 Poet Starlight / 光学 interposer 提供激光,而且几天前 Poet 的 CFO 也确认他们正在给 Marvell 供货。 > $SIVE -> $POET -> “其他超大规模云供应商的 NDA” 1. 西方超大规模云厂商 > $SIVE -> $JBL(1.6T LRO)-> 1. $META(Jabil 承接了 $INTC 的硅光遗产,对应 Meta LRO 项目) 2. $NVDA(NVIDIA 可能把光模块 OEM 出去)-> $MSFT | AWS | 其他超大规模云厂商 $SIVE 是 $JBL 1.6T 光模块已确认的激光源。 > $SIVE -> Ayar(上个月为了量产爬坡融资 5 亿美元) -> 1. Alchip(联合 CPO) 2. Intel 3. GUC/Wiwynn -> $AMZN(Alchip) -> $AMD(来自 $GFS 合作的 CPO)也有可能。 $SIVE 已知是 Ayar 的激光供应商,而且 Ayar 最近把 $MTSI / $LITE 从网站上删掉了。现在只显示 $GFS + $SIVE,这很可能说明 Sivers 是主要激光供应商。 既然 $GFS x $AMD 最近合作了,那如果 AMD 走 Ayar 路线,Sivers 就可能成为其 CPO 项目的核心激光供应商。 > $SIVE -> Enablence -> O-Net(大型亚洲 OEM)-> 亚洲超大规模云厂商 1. $AVGO ELS(可能) 2. $META 和 $GOOGL ELS 3. ByteDance(可能)-> ELS 4. Tencent(可能)-> ELS 5. Alibaba(可能)-> ELS $SIVE 和 O-Net / Enablence 在 OFC 附近有 ELS 合作关系。 Sivers 激光由 Win Semi 之类的代工厂大规模生产……而且在最近的图片展示里,它们也验证在 $GFS 的 CPO 供应链中。 这不是在看 Sivers 今天靠认证收入能指引多少,而是未来每一个西方 / 亚洲超大规模云厂商在 2027、2028、2029 及以后对 CPO / 1.6T 的需求所对应的营收。 $SIVE 看起来是市场上最不被了解、但还没被真正反映出来的光子股票之一。

英文原文

All the hyperscalers $SIVE likely ends up in 2027-2028 is staggering at a $900m MC. Markets don't understand what's coming. From speculative mapping: > $SIVE -> $POET -> $MRVL -> 1. $AMZN (purchase agreement/warrants with photonic fabric from celestial) 2. $MSFT (maia) 3. $GOOGL (recent development talks with Marvell) $SIVE powers Poet Starlight/optical interposers, and Poet's CFO confirmed they're supplying to Marvell few days ago. > $SIVE -> $POET -> "NDAs other hyperscaler suppliers" 1. Western Hyperscalers > $SIVE -> $JBL (1.6T LRO)-> 1. $META (Jabil $INTC SiPH inheritance, maps to Meta LRO program) 2. $NVDA (NVIDIA possibly OEMs optical transceivers) -> $MSFT | AWS | hyperscalers $SIVE is the confirmed laser source for $JBL 1.6T optical transceivers. > $SIVE -> Ayar ($500m fundraiser last month for volume ramp) -> 1. Alchip (Joint CPO) 2. Intel 3. GUC/Wiwynn -> $AMZN (Alchip) -> $AMD (CPO from $GFS partnership) possible. $SIVE is known laser supplier to Ayar, and Ayar removed $MTSI / $LITE from their website recently. Only showing $GFS + $SIVE, likely showing Sivers was primary laser supplier. As $GFS x $AMD partnered up recently, that makes Siver a possible core laser supplier for $AMD's CPO program if they go with Ayar. > $SIVE -> Enablence -> O-Net (massive Asian OEM)-> Asian Hyperscalers 1. $AVGO ELS (possible) 2. $META and $GOOGL ELS 3. ByteDance (possible) -> ELS 4. Tencent (possible) -> ELS 5. Alibaba (possible) -> ELS $SIVE ELS partnership with O-Net/Enablence around OFC. Sivers lasers is mass produced by foundries like Win Semi... and they're validated in $GFS CPO supply chains too from their recent image presentations. It's not about what Sivers is forecasting today from qualification revenue that everyone models off of. Alpha comes from future revenue proportional to demand from every Western/Asian hyperscaler for CPO/1.6T in 2027, 2028, 2029, and onward. $SIVE looks like one of the most unknown photonic stocks on the market that's yet to come.

在 X 上查看原推 ↗