Goldman Sachs expert call notes: bullish on China's memory expansion and benefits to domestic SPE

Overview

This report summarizes Goldman Sachs' China Memory Expert Webinar held on July 24, 2026, with discussion focused on China's DRAM capacity expansion, product development, and overall memory pricing trends. The experts' views are consistent with Goldman Sachs' positive outlook on China's semiconductor equipment sector: domestic wafer fabs and memory makers are expanding capacity over the next few years, while the adoption rate of domestic SPE is rising, which is expected to benefit Chinese semiconductor equipment companies.

Core views

The core view is that China's domestic DRAM expansion cycle is still underway, with leading industry players' annual capacity potentially more than doubling from current levels by 2030; capacity expansion in Hefei, Shanghai, and Beijing is expected to be fully ramped up by 2028, and large bases will continue contributing to domestic DRAM supply after 2028. On products, domestic DRAM suppliers continue to improve with process upgrades; although they still lag global peers, production of HBM3 and HBM3E remains a 2026E target for the domestic leader, and 3D DRAM may see new development progress in 2027E. On pricing, the pace of price increases in 2H26 may be lower than in 1H26, but strong AI demand supports continued price increases in 2027E.

Analysis framework

The report is based on expert call notes and validates trends in China's memory supply chain from three dimensions—capacity, product, and price—then maps them to the investment logic for beneficiaries among Chinese semiconductor production equipment companies. The capacity dimension focuses on expansion milestones and regional layouts of domestic DRAM makers; the product dimension focuses on HBM3/3E, DUV substitution paths, and 3D DRAM; the price dimension focuses on smartphone customer bargaining power, AI demand, and capacity allocation by global suppliers.

Methodology notes

• Industry expert researchExpert call notes

  Use experts' judgments on capacity, product, and price to validate investment views on the industry chain.

  Goldman Sachs uses expert observations on China's DRAM capacity expansion, domestic equipment adoption, HBM/3D DRAM progress, and memory pricing trends to support its positive view on China's semiconductor equipment sector.

• Factor framework disclosureGS Factor Profile

  Goldman Sachs compares stocks with the market and industry peers across four attributes: Growth, Financial Returns, Multiple, and Integrated.

  The disclosure page states that Growth is based on forward sales, EBITDA, and EPS growth; Financial Returns is based on ROE, ROCE, and CROCI; Multiple is based on valuation metrics such as P/E, P/B, P/D, EV/EBITDA, and EV/FCF; Integrated is the composite percentile of growth, returns, and inverse valuation.

• M&A scenario disclosureM&A Rank

  Goldman Sachs uses an M&A rank from 1 to 3 to assess the probability of being acquired, and incorporates M&A factors into some target prices.

  The disclosure page states that M&A rank 1 indicates a 30%-50% probability of being acquired, 2 indicates 15%-30%, and 3 indicates 0%-15%; the main body of this report does not provide specific M&A ranks for the relevant companies.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Naura

  A beneficiary in China's semiconductor equipment space; rated Buy by Goldman Sachs.

  Strengths

  Benefits from capacity expansion at domestic wafer fabs and memory makers, as well as from some plants increasing the use of domestic SPE.

  Weaknesses

  High-end equipment segments remain constrained by EUV, lithography, and similar factors; some premium equipment still depends on global tier-one suppliers.

  Comparison

  Compared with global tier-one SPE, domestic SPE is seeing higher adoption in some new DRAM production lines, but gaps remain in high-end equipment.

  Risks

  Slower capacity expansion, weaker-than-expected progress in domestic substitution, memory price declines, or customer price pressure.

• AMEC

  A beneficiary in China's semiconductor equipment space; rated Buy by Goldman Sachs.

  Strengths

  Benefits alongside the China SPE sector from domestic DRAM capacity expansion and rising adoption of domestic equipment.

  Weaknesses

  The report does not disclose company-specific details on orders, revenue, or product lines.

  Comparison

  The report groups AMEC together with Naura, ACMR, and Kematek as Buy-rated China SPE names.

  Risks

  Pace of equipment localization, memory makers' capex, and constraints from technology node migration.

• ACMR / ACM Research

  A beneficiary in China's semiconductor equipment space; rated Buy by Goldman Sachs.

  Strengths

  Benefits from domestic memory expansion and a rising share of domestic SPE usage.

  Weaknesses

  The report does not provide standalone fundamentals or financial forecasts for ACMR.

  Comparison

  The disclosure page lists ACM Research at Buy with a $88.42 target price, appearing alongside other Buy-rated China SPE names.

  Risks

  Delays in domestic DRAM expansion, a weaker price cycle, and global competitive pressure.

• Kematek

  A beneficiary in China's semiconductor equipment space; rated Buy by Goldman Sachs.

  Strengths

  Included in Goldman Sachs' favored China SPE basket and benefits from capacity expansion by domestic memory makers.

  Weaknesses

  The report does not provide company-specific operating data or contributions by specific equipment categories.

  Comparison

  The disclosure page lists Kematek Buy Rmb94.60, alongside Naura, AMEC, and ACMR.

  Risks

  Domestic equipment penetration below expectations, constrained technology upgrades, and downstream capex volatility.

• Domestic DRAM suppliers

  The core industry-trend participants in this report.

  Strengths

  Capacity expansion is clear, products continue to evolve with process upgrades, and HBM3/3E and 3D DRAM represent premiumization directions.

  Weaknesses

  Products still lag global peers, for example in speed; yields may be lower than those of global tier-one suppliers.

  Comparison

  Prices are close to those of global tier-one suppliers, but yields may be lower; global tier-one suppliers are unlikely in the short term to shift HBM capacity back to traditional DRAM.

  Risks

  Restrictions on equipment such as EUV, uncertainty around DUV substitution paths, smartphone customer price pressure, and slower-than-expected technological progress.

Key data

• Meeting date2026-07-24Goldman Sachs held the China Memory Expert Webinar.
• Capacity outlookLeading industry players' annual capacity could more than double from current levels by 2030Experts believe expansion in Hefei, Shanghai, and Beijing will be fully ramped up by 2028, with large bases continuing to contribute thereafter.
• Domestic equipment adoptionSome plants mainly use domestic SPE except for high-end tools such as lithographyNew DRAM capacity uses both global tier-one SPE and domestic SPE.
• Product targetThe domestic leader targets production of HBM3 and HBM3E in 2026EDue to restrictions on equipment such as EUV, domestic manufacturers need to use DUV or other innovative processes to develop high-end products.
• 3D DRAM milestoneNew product development progress may emerge in 2027EExperts believe 3D DRAM could become one of the breakthrough paths for high-end products.
• Price trendPrice increases in the third and fourth quarters of 2026 will be smaller than in 1H26, while 2027E still has support for further increasesSmartphone brand makers are resisting higher costs, but AI demand in China and overseas remains strong.
• Rating and pricing informationACM Research Buy $88.42; AMEC Buy Rmb388.50; Kematek Buy Rmb94.60; NAURA Buy Rmb757.00From the disclosure page's rating and pricing information.
• Goldman Sachs global coverageAs of 2026-07-01, coverage of 3,104 stocks; rating distribution Buy 50%, Hold 34%, Sell 16%The proportion with investment banking relationships is Buy 65%, Hold 59%, Sell 43%.

Impact & implications

This expert call reinforces the medium- to long-term demand logic for China's semiconductor equipment chain: DRAM capacity expansion drives equipment procurement demand, domestic substitution raises domestic SPE penetration, and AI demand supports memory prices and the sustainability of industry capex. In terms of investment mapping, Goldman Sachs lists Naura, AMEC, ACMR, and Kematek as Buy-rated names; at the same time, EUV constraints, yield gaps, and smartphone customer price pressure indicate that industrial upgrading still faces technical and pricing headwinds.

Risks

• Domestic DRAM products still lag global peers, for example in slower speed, and yields may be lower.
• Advanced lithography equipment such as EUV is restricted, so domestic manufacturers need to rely on DUV or other innovative manufacturing processes to develop high-end products, creating uncertainty in the technology path.
• Smartphone brand makers resist higher memory costs, which may restrain the magnitude of price increases in 2H26.
• If AI demand slows, the assumption of continued memory price increases in 2027E and continued HBM capacity allocation by global suppliers could be challenged.
• If the rise in domestic SPE adoption is below expectations, the magnitude of benefits to Chinese semiconductor equipment companies will be reduced.

What to watch

• Whether DRAM expansion projects in Hefei, Shanghai, and Beijing are fully ramped up by 2028 as planned.
• The actual contribution of large bases to domestic DRAM supply after 2028.
• The progress toward achieving the domestic leader's 2026E production targets for HBM3 and HBM3E.
• Whether 3D DRAM sees new product development progress in 2027E.
• Whether memory price increases in the third and fourth quarters of 2026 are indeed lower than in 1H26 as experts expect.
• Whether AI demand in China and overseas continues to support memory price increases in 2027E.
• Whether global tier-one suppliers continue to maintain HBM capacity rather than shift back to traditional DRAM.