SemiAnalysis responds to Vera SOCAMM controversy: Critics have not visited the SK Hynix booth at Computex

```

After SemiAnalysis’s report on NVIDIA’s Vera Rubin platform SOCAMM memory triggered dramatic market reactions, the latest response shifts the focus of controversy from “whether AI memory demand is cooling” back to the facts about the product itself: the agency says those accusing it of spreading “fake news” have clearly not visited the Hynix booth at Computex.

SemiAnalysis posted on social media, saying its Vera SOCAMM report “caused some turbulence,” but “some people are drawing incorrect conclusions.” The organization emphasized that critics who call the report fake news “clearly haven’t visited the Hynix booth at Computex.”

The report previously claimed that NVIDIA has reduced the modular memory capacity of its next-generation Vera Rubin server rack from 55TB to 28TB, and the market quickly interpreted this as a signal of cooling AI memory demand.

However, the key controversy is that SOCAMM DRAM and HBM are not the same type of products. SemiAnalysis founder Dylan Patel later clarified that his analysis “contains no bearish outlook at all.” According to Barron's, the reduction of modular memory capacity in the Vera Rubin platform may actually reflect the crowding-out effect of HBM on traditional memory production capacity.

Focus of Controversy: SOCAMM Capacity Reduction Interpreted by Market as Cooling Demand

The immediate trigger for this round of market volatility was SemiAnalysis’s report on the change in SOCAMM DRAM capacity in the NVIDIA Vera Rubin server rack. The report said that NVIDIA reduced the modular memory capacity from 55TB to 28TB.

This change was quickly priced in by the market as a signal of cooling AI server memory demand. Micron’s stock price plummeted 13% that day, the semiconductor sector was under pressure, and major indices closed lower across the board.

The reaction in Micron’s stock price also amplified market scrutiny regarding the impact of SemiAnalysis research. Several market figures and internet users questioned the quality of the agency’s research and the way information was disseminated, believing it repeatedly created bearish narratives about Micron under the guise of “in-depth research” and triggered market panic.

However, SemiAnalysis’s response indicates that the agency does not accept allegations of “fake news.” Their latest statement redirects the focus to the Computex live demonstration, implying that the relevant SOCAMM products are not market speculation, but items already physically displayed at the booth.

SemiAnalysis Pushback: If You Haven’t Been to the Hynix Booth, Don’t Dismiss It Easily

SemiAnalysis said in a recent post: “Our Vera SOCAMM note caused some turbulence. As always, some people are drawing incorrect conclusions. Those calling it fake news have clearly not visited the Hynix booth at Computex.”

The core of this response is to shift the controversy from market inferences about demand to whether the product actually exists. SemiAnalysis does not deny that its report triggered market volatility, but explicitly refutes the external claim that it is “fake news.”

Semiconductor Insider also posted on social media supporting this point, saying that SK Hynix has displayed a 192GB SOCAMM2 module at Computex, specifically built for the NVIDIA Vera Rubin platform, based on LPDDR5X, focusing on power efficiency and is ready for mass production.

The account also stated that SOCAMM2 is a plug-in module rather than soldered memory, with up to twice the bandwidth of older server DRAM and better energy efficiency, helping reduce thermal and cost pressures for AI servers running 24/7.

SOCAMM2 Positioning: Vera Rubin Uses Both HBM4 & System Memory

Another key controversial point is the market conflating changes in SOCAMM with overall AI memory demand.

According to Semiconductor Insider, the Vera Rubin Superchip uses two types of memory simultaneously: HBM4 and SOCAMM2. The former is high-speed stacked memory, while the latter serves as system memory. SK Hynix’s Computex display includes the complete memory configuration needed for Rubin GPUs and Vera CPUs.

This means that SOCAMM2 is not a substitute for HBM, but another part of the Vera Rubin platform’s memory architecture. Its role is more about capacity, power consumption, and system efficiency.

Semiconductor Insider also stated that AI training and inference requires a large amount of high-speed memory, SOCAMM2 helps relieve bottlenecks in power consumption and capacity. Vera Rubin is expected to deliver 3.5 to 5 times more performance than Blackwell. The account concludes that SemiAnalysis’s claims about the booth demonstration are accurate, and AI infrastructure construction is ongoing.

HBM Crowding-Out Effect: Capacity Reduction Doesn’t Necessarily Equal Weaker Demand

For investors, the more important question is whether Vera Rubin’s reduction of SOCAMM capacity means AI memory demand is weakening.

SemiAnalysis founder Dylan Patel has clarified that the agency’s analysis “contains no bearish outlook at all.” This is in contrast to the initial market reaction.

According to Barron's, NVIDIA cutting modular memory capacity may actually reflect that the HBM crowding-out effect is occurring. The reason is that HBM production consumes far more wafer capacity—about three times as much per unit as standard memory. As HBM supply expands, capacity for other memory types will be squeezed and may push up prices for the entire memory industry.

Trendforce analysts also point out that as HBM technology evolves through 2027, chip sizes rise in sync with demand, and the crowding-out effect on traditional DRAM capacity is expected to intensify further. This will provide suppliers with justification to raise HBM prices and strengthen their pricing power during next year’s HBM negotiations.

In other words, the adjustment of SOCAMM capacity is not necessarily a signal of weakening AI memory demand, but possibly a result of prioritizing high-value HBM production capacity.

Risk Warning and DisclaimerThe market has risks; investments require caution. This article does not constitute personal investment advice and does not take into account specific investment goals, financial situation, or needs of individual users. Users should consider if any opinions, views, or conclusions in this article suit their particular circumstances. Investing based on this is at your own risk. ```