$MICRON
相关推文
共 2 条 · 按时间倒序 · 滚到底自动加载更多
-
易胜达受益于成熟存储涨价且估值偏低，建议做多易胜达而非美光。
是的，易胜达(ESMT, 3006) 涉足多个成熟制程(legacy)存储细分市场。 其产品组合包括：SLC/SPI NAND、NOR Flash、DDR2/DDR3、MCP/eMCP，也就是所有有趣的成熟制程产品。 如果你想关注易胜达所涉猎的更广泛的行业涨价趋势： 1. SLC NAND：2026年下半年预计上涨约120-170%。 2. DDR2：2026年第三季度预计上涨约35-40%。 3. DDR4：第三季度+30–40%（伴随DDR3供应收紧） 4. NOR：下半年+60–65%+（高密度） 5. MCP/eMCP：据我估算，易胜达环比可能上涨约+15–25% 官方收入结构拆解： DDR2 + DDR3：约占收入的45% DDR4/其他DRAM ~ 10% 的收入 eMCP/MCP ≈30% 模拟及其他 ≈15% 无论如何，持续时间和涨价幅度是关键： -> “没有任何一家全球主要SLC NAND供应商计划在近期增加新产能。” -> TrendForce特别预计，库存耗尽将使SLC NAND在第四季度供应更加紧张。 -> 华邦电子(Winbond)退出DDR2，将使库存更加紧张（他们正在分配订单至2029-2030年） -> 易胜达管理层表示，市场的主要问题在于供应而非需求；公司的晶圆代工产能分配仅覆盖了客户订单的大约60–70% -> 南亚科(Nanya)：“内存短缺可能持续到2027年底，2028年供需依然紧张” -> 南亚科明确表示，产能限制可能导致DDR5、LPDDR5、DDR4、LPDDR4和DDR3全面短缺。 -> 易胜达将eMCP列为因成熟制程资源被转移至其他领域而获得订单转移红利的产品之一 ______ 如果我要再次陈述我的观点： 我认为，尽管已计入力积电(PSMC) 45%的晶圆涨价因素，易胜达约1.9倍的7月年化运行市盈率(P/E)可能会进一步压缩，因为成熟存储价格仍在上涨。 行业评论支持这种趋势将持续至2027年。 锦上添花的是，如果我们开始看到长期协议(LTAs)以及对2029-2030年的可见度（但这尚未计入估值）。 我认为市场要么不知道这一点，要么错误定价了持续时间，要么高估了力积电晶圆涨价的影响。 也许玩经典LOL中排名第一的挑战者“易胜达”的双排，比玩卡在低分段的“美光(Micron)”更有趣？
英文原文
Yep, ESMT (3006) has exposure to multiple legacy memory segments. Has a portfolio of: SLC/SPI NAND, NOR Flash, DDR2/DDR3, MCP/eMCP, so all the fun legacy stuff. If you want to look at broader industry price hikes coming up that ESMT has exposure to: 1. SLC NAND: ~120-170% for H2 2026. 2. DDR2: ~35-40% for Q3 2026. 3. DDR4: +30–40% Q3 (with DDR3 tightening) 4. NOR +60–65%+ for H2 (high density) 5. MCP/eMCP: maybe ~+15–25% QoQ for ESMT from my own estimate For official breakdown: DDR2 + DDR3: ~45% of revenue DDR4/other DRAM ~ 10% of revenue eMCP/MCP ≈30% Analog + other ≈15% Regardless, duration/hikes is the key: -> "none of the major global SLC NAND suppliers plan to add new production capacity in the near term." -> TrendForce specifically expects depleted inventories to make SLC NAND even tighter in Q4. -> Winbond withdrawing DDR2, would make inventory even tighter (and they're doing allocations into 2029-2030) -> ESMT management said the market's main problem was supply, not demand; the company's foundry allocations covered only roughly 60–70% of customer orders -> Nanya: "memory shortages could last through the end of 2027, with supply and demand still tight in 2028" ->Nanya explicitly said capacity constraints could cause shortages across DDR5, LPDDR5, DDR4, LPDDR4 and DDR3. -> ESMT included eMCP among the products receiving order-transfer benefits as mature node resources were being pulled elsewhere ______ If I had to state my thoughts again: I think ESMT's ~1.9x July annualized run-rate P/E could compress further, despite factoring in PSMC's 45% wafer hike due to further legacy memory price hikes. Industry commentary supports durability throughout 2027. Cherry on top is if we start seeing LTAs and visibility into 2029-2030 (but not factored in). I think markets either don't know about this, are mispricing duration, or overestimating PSMC's wafer hike effects. Maybe playing duoQ with Rank 1 Challenger "ESMT" in LoL classic... Is more fun than playing with hardstuck "Micron" in League?
-
InP成2026新瓶颈，市场重估AAOI与LITE在超大规模云ASIC中的价值。
$AAOI 自今日发布论点以来上涨 24%，$LITE 上涨 5%。 从物料清单(BOM)分析来看，LITE（市值 270 亿美元）因光路交换(OCS)技术而向 TPU Ironwood 倾斜，同时也受益于 NVDA 及所有专用集成电路(ASIC)。 AAOI（市值 25 亿美元）则主要受益于 MSFT MAIA 的量产爬坡和 Amazon Trainium。 磷化铟(InP)就像高带宽内存(HBM)一样，将成为 2026 年的瓶颈，因为它们是这些部署中激光器使用的基础材料。 类似于美光和 SK 海力士的内存瓶颈，市场注意力可能会转向 InP 晶圆厂，例如 $AAOI，它是美国少数几家此类工厂之一（还有 COHR, Macom）。 但相比之下，$LITE 由于 Google TPU 的成功（来自 Meta 和 Anthropic 的采购订单），今年迄今(YTD)已上涨 362%，而 $AAOI 今年迄今仅上涨 7%。 我们主要看到这种情况是因为散户或媒体对 $AMZN Trainium 或 $MSFT Maia 部署缺乏关注，这些部署预计将在 2026-2027 年大规模量产。 然而，由于每个超大规模云服务商都希望降低其自有云平台的推理成本，它们很可能都会成功。 如果我们看到其他超大规模云服务商采用 OCS 以实现 TPU 达到的优化性能，鉴于 $LITE 在该特定领域的垄断地位，预计其估值将比现在进一步提升。 然而，如果我们看到 $MSFT Maia 量产（鉴于 $AAOI 可能正在为他们开发新架构），以及 $AMZN Trainium 量产（40 亿美元权证 + 采购订单），预计 $AAOI 将重新估值。 光电子学和 InP 将成为像内存一样的新瓶颈。 我们可能会看到投资流向下游玩家，如 $COHR、中际旭创(Innolight)、$LITE，以及 2026 年针对特定超大规模云服务商 ASIC 的隐藏杠杆标的如 $AAOI 这一主题。 市场目前正在奖励 Google TPU 供应链，但可能错过了其他超大规模云服务商 ASIC 的量产机会。
英文原文
$AAOI is up 24% and $LITE is 5% since my thesis today. From BOM analysis, LITE ($27B) is levered toward TPU Ironwood due to OCS but benefits from NVDA + all ASICs. AAOI ($2.5B), is levered toward MSFT MAIA ramp and Amazon Trainium. InP like HBM, will be a bottleneck for 2026 as they’re the foundational materials used for lasers in these deployments. Similar to memory bottlenecks with Micron and SK Hynix, we’ll likely see attention drawn to InP fabs, such as $AAOI, which happens to be one of the sole ones in America (COHR,Macom) But compared to $LITE that is up 362% YTD due to the success of Google’s TPU (from Meta and Anthropic purchase orders), $AAOI is only up 7% YTD. We’re largely seeing this because there’s a lack of retail or media attention on the $AMZN Trainium or $MSFT Maia deployments, which are largely expected to ramp up in 2026-2027. However they’re all likely to succeed due to each hyperscaler wanting to lower costs of inference for their own cloud platform. If we see other hyperscalers adopt OCS for optimized performance that the TPU achieved, expect $LITE to re-rate more than they have now given their monopoly in that specific segment. However, if we see $MSFT Maia ramp up (given $AAOI is likely developing a new architecture for them), and $AMZN Trainium ramp up ($4B warrant + purchase orders), expect $AAOI to rerate. Photonics and InP will be the new bottleneck like memory. We’ll likely see investments pour down stream to players like $COHR, Innolight, $LITE, and hidden levered plays on specific hyperscaler ASICs like $AAOI as a theme in 2026. The market is currently rewarding the Google TPU supply chain but might be missing other hyperscaler ASIC ramps.