Morgan Stanley responds to SemiAnalysis note: Acknowledges CPO delay, does not acknowledge 800V delay.

Morgan Stanley has made a cautious assessment regarding a market-shaking report issued by a well-known semiconductor research institute.

According to Chase Wind Trading Desk, on June 9, SemiAnalysis released a report stating that the large-scale shipment of Nvidia’s native 800V DC products will be delayed until after 2028, and also lowered expectations for CPO-related switch shipments, directly causing a slump in U.S. optical communications stocks. Nvidia reportedly has denied these claims.

Morgan Stanley’s Greater China hardware and semiconductor team responded immediately, agreeing basically with SemiAnalysis on the CPO issue, but holding an opposite stance on the 800V timeline.

For CPO, Morgan Stanley estimates global optical engine shipments in 2027 will only be 6 to 7 million units, far below the market consensus of 20 to 30 million units, believing short-term sentiment will remain under pressure; nonetheless, they maintain an overweight rating on core CPO-related stocks, judging that explosive growth for CPO will start from 2028, with the long-term outlook unchanged.

As for 800V, Morgan Stanley clearly stated that this news does not match its supply chain research, and the mass production progress of 800V DC power cabinets has not been interrupted.

CPO: Recognizing Delay, But Long-Term Logic Unchanged

Morgan Stanley’s Greater China semiconductor team issued a systematic forecast on the recent ramp-up rhythm of CPO in its June 10 report.

Morgan Stanley expects, taking into account TSMC’s capacity expansion plans and current yields, that global optical engine shipments in 2027, including scale-up and scale-out solutions, will only be 6 to 7 million units.

Manufacturing bottlenecks are the key limiting factor.

TSMC plans to expand photonic integrated circuit (PIC) capacity to 10,000 wafers per month in Q1 2027, but the system-on-integrated-chip (SoIC) yield is currently only 50-60%, with downstream assembly yield even lower at 20-50%. The overlap of these two bottlenecks drastically reduces the final shipment scale.

Morgan Stanley points out that current market expectations for optical engine shipments are as high as 20 to 30 million units, significantly different from its projections, and this gap will keep CPO sector sentiment weak in the short term.

2026-2028 will be a transitional phase where pluggable optical modules, CPO/NPO optical solutions, and copper interconnect solutions coexist. The industry mainstream will still be 1.6T/3.2T products, and CPO penetration will require more time.

Despite this, Morgan Stanley maintains an overweight rating on core CPO-related stocks such as TSMC, judging that the long-term growth logic for CPO remains unchanged, with explosive growth expected to start from 2028.

800V: Supply Chain Research Contradicts SemiAnalysis

On the issue of 800V DC power architecture, Morgan Stanley’s Greater China hardware team’s position is clearer.

According to Morgan Stanley, Nvidia stated at the Taipei GTC conference that its 800V DC technology R&D is progressing steadily, and its 800V DC power cabinets will be ready for mass production in Q3 2026. This is not entirely contrary to SemiAnalysis’s claim that “large-scale shipments will be delayed until after 2028”—"ready for mass production" and "large-scale ramp-up" are inherently different milestones.

An additional noteworthy detail from Morgan Stanley's report is: Delta Electronics is expected to become the first manufacturer to mass-produce independent 800V DC power cabinets, with delivery of the first batch to a top North American hyperscale cloud service provider in Q4 2026.

However, Morgan Stanley also notes that initial shipment volumes will be limited, and that R&D, UL certification, and the establishment of industry safety standards for 800V DC protection devices will all take time to materialize.

The real substantive disagreement between the two reports lies in the direction of the ±400V DC solution.

SemiAnalysis believes ±400V will proceed as planned, continuing to serve cloud providers’ own ASIC deployments and running parallel long-term with 800V; whereas Morgan Stanley’s supply chain research indicates that research efforts of major hyperscale cloud providers have shifted from ±400V to 800V. For the same group of cloud providers and deployment window, both scenarios cannot be true at the same time. This divergence will directly affect the power supply chain landscape and supplier positioning.

Market Impact: Expectation Gap and Verification Points

The downturn triggered by the SemiAnalysis report reflects the market's high sensitivity to the ramp-up pace of the two key AI infrastructure tech routes.

Morgan Stanley's response has to some extent provided support for 800V-related stocks, but the confirmed expectation gap for CPO means it will be hard for sentiment in that sector to quickly recover.

There are three key points for subsequent verification:

First, whether Delta’s first batch of 800V power cabinets in Q4 2026 is shipped as scheduled and to which customer;

Second, whether the year-end ±400V sidecar order lands, and to which supplier—this will directly test the accuracy of both sides’ judgment on the cloud providers' tech path;

Third, Nvidia’s next public update on the 800V and Rubin Ultra/Kyber product generation timeline.

With differing information currently, the status of orders and certification progress will ultimately determine the real pace of the two tech routes.

~~~~~~~~~~~~~~~~~~~~~~~~

The above brilliant content comes from Chase Wind Trading Desk.

For more detailed explanations, including real-time interpretation and frontline research, please join 【Chase Wind Trading Desk ▪ Annual Membership】

Risk Warnings & DisclaimerThe market involves risks, investment requires caution. This article does not constitute individual investment advice, nor does it consider the particular investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their specific circumstances. Investing based on this is at your own risk.