Mizuho Securities: The long-term trend for CPO remains unchanged, but the real alpha lies with the beneficiaries during the transition period, not just the eventual winners.
BlockBeats News, June 17 — In its latest report, Mizuho Securities responded to market concerns regarding the delays in CPO (Co-Packaged Optics) and 800VDC high-voltage direct current technologies. The report asserts that the long-term trend for CPO technology remains unchanged, but the real alpha lies with the beneficiaries during the transitional period, rather than focusing solely on the ultimate winners. Both CPO and 800VDC are key upgrades for AI servers over the next five years—the former significantly reducing power consumption and latency, the latter increasing single rack power from 60kW to over 600kW. However, large-scale implementation will take time, and the market's anxiety is overstated.
The report reiterates the medium- and long-term advantages of stocks such as LITE, AVGO, AAOI, and CRDO.
Overall, CPO and 800VDC are not "delayed," but are instead being "phased in stepwise." The key in 2026-2027 is to seize the transitional dividends of NPO and +/-400VDC, and to switch to the final technologies afterward. On the demand side, the power density of AI racks continues to accelerate. The crucial point is to see the "ladder" rather than bet everything on a single point in time.
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
OpenAI is expected to burn over $278 billion in cash by 2030, with a revenue target of $350 billion over the same period.
OpenAI is expected to generate up to $278 billion in negative free cash flow over the next five years, driven solely by computing power and infrastructure, with cumulative spending projected to reach approximately $856 billion by 2030. The company anticipates revenue of $350 billion in 2030 and $36 billion in 2026, with total revenue from 2026 to 2030 amounting to around $840 billion.
Garrett Jin's ZEC short position has an unrealized loss of $33.422 million.
Big iron ore miners deplete reserves faster than they replace
