Post

Mulight沐光
Mulight沐光
Show original
长鑫上市首日涨了466%,朋友圈全是截图的。同一时间SK海力士发了Q2财报——营收创纪录,然后说DRAM供应紧张会持续到2028年。 两件事放一起看就有意思了。一边是中国存储芯片的旗帜终于上市,拿回295B人民币的弹药。另一边是韩国人告诉你:不好意思,未来三年DRAM还是不够。 CXMT现在占全球DRAM产能大概5%,主要做DDR4和LPDDR4。但市场在往DDR5和HBM3E推进——这两个品类的毛利率是DDR4的两到三倍。SK海力士Q2的HBM营收同比涨了快200%,美光也刚把HBM产能规划翻了一倍。 CXMT这295B,大头会投到先进制程和HBM产线。但存储芯片不是有钱就能追赶的——设备禁运、人才缺口、良率爬坡,每道坎都不轻松。 短期股价有概念溢价,中长期要看HBM能不能出货。真正的验证节点在2027年——能不能进英伟达的供应链。 接下来看:CXMT下一份财报里的研发支出占比和资本开支方向。SK海力士的供应紧张预警,对CXMT来说既是天花板也是地板——追不上就被锁死在低端,追上了就是全球第四把交椅。
Mulight沐光
Mulight沐光
Profit surged 557%, but the stock price plummeted 10%. SK Hynix just reported the most profitable quarter in the company's history, yet the Korean stock market triggered a circuit breaker. On the same day, Samsung said the shortage of memory chips will continue until 2028 and has already signed five-year long-term contracts, causing its stock price to rise 8%. Two extremes in the same sector. The issue is not demand, but expectations. The market has fully priced in AI storage — record profits still can't support the valuation because everyone is betting on "exceeding expectations again and again." But on the other hand, Samsung's willingness to sign five-year long contracts to lock in capacity indicates that downstream AI data center demand is not just a short-term hype. SK Hynix's drop shattered the expectation that "AI storage will always rise." What remains is the real fundamental worth watching: HBM4 is just beginning to ramp up, downstream cloud providers are still scrambling for capacity, and shortages will last at least until 2028. The short-term dip is actually a good thing in the medium to long term. The above does not constitute investment advice; please make judgments based on your own research.

Disclaimer: OKX Orbit content is provided for informational purposes only. Learn more

Replies

No comments yet. Be the first to reply!