Goldman Sachs Semiconductor Earnings Preview: After SOX Surges, Can AMD and AMAT Keep Going Up?

TL;DR
· Goldman Sachs expects most subsectors of semiconductors to have upside in Q2, but SOX has already risen by approximately 87.8% in the quarter.
· AI capital spending, DRAM/HBM, advanced packaging, and EDA tools are the key drivers of this round of earnings upside.
· Goldman Sachs favors AMD, AMAT, and ON, exercises caution with ARM and KLAC, and recommends prudent profit-taking post-Qnity surge.

Prior to the semiconductor earnings season, stock prices had already run up significantly. A review of the Nasdaq in Q2 showed the PHLX Semiconductor Index rose by 87.8%, marking its best quarter since its establishment in 1994 and outperforming the broader market. In a semiconductor Q2 outlook, Goldman Sachs indicated that there is still room for upside in fundamentals but the post-rally sector is no longer a "buy everything together" opportunity. AI servers, memory, advanced packaging, and design software remain the strongest themes, while challenges such as softening smartphone demand, evolving capex structures, and export restrictions will amplify stock divergences.

Semiconductors Still Have Room for Upside, but the AI Chain is Already Fragmenting

According to the Goldman Sachs report, companies in computing, memory storage, semiconductor equipment, and some analog chip sectors may still report Q2 results or offer subsequent guidance that exceeds market expectations.

The main theme in computing continues to be AI servers. Demand for server CPUs, cloud provider ASIC projects, and AI accelerator cards are driving expectations for data center revenue for companies like AMD. Memory and storage benefit from tight DRAM/HBM supply-demand dynamics, improving HDD pricing, and expectations for the NAND cycle, with limited short-term supply pressures.

The equipment chain is more focused on the medium term. AI servers require more HBM and advanced packaging, expansions and technological upgrades by memory manufacturers will drive demand for deposition, etching, and other process steps. Visibility on orders for some equipment companies has already extended to 2028.

Analog semiconductors are not experiencing a broad-based recovery. Goldman Sachs prefers companies with high exposure to industrial, aerospace defense, and data center markets, exercising caution with targets more reliant on smartphones or traditional automotive cycles.

This differentiation is also reflected in Goldman Sachs' tactical choices. The report favors Applied Materials, AMD, and Applied Materials, while exercising caution with ARM and KLA. For semiconductor materials and electronic solutions company Qnity, Goldman Sachs remains optimistic about wafer start rates and execution, but believes that the risk-return profile has become more balanced after the stock price increase, mainly due to reporting considerations.

AMD and Applied Materials Are Goldman Sachs' Two Most Convincing Long Position Examples

AMD is one of the most explicit examples of a bullish case in the compute stack for Goldman Sachs. The Goldman model shows that AMD's 2027 EPS forecast has been raised to $14.50, about 13% above market consensus. The 2027 data center revenue forecast is $66.682 billion, approximately 18% higher than market expectations.

Supporting this view is strong server CPU demand, improvement in data center business margins, and operational leverage from the upcoming AI chip ramp. AMD has previously announced that the Advancing AI 2026 event will be held in San Francisco on July 23, 2026, both in-person and virtually. Beyond earnings season, the market will be looking to see if AMD can provide a clearer AI server roadmap, customer progress, and revenue trajectory at this event.

The Goldman model shows that AMD's 2027E EPS is $14.50, above the market's $12.87. The 2027E data center revenue is $66.682 billion, higher than the market's $56.59 billion, with server CPU and MI450 ramp being key drivers.

Applied Materials, on the other hand, represents the end of the equipment chain with stronger order visibility. Goldman has raised Applied Materials' price target from $520 to $645, based on a 32x multiple and normalized EPS of $20. The key assumption in the report is that strong DRAM investment will drive the company to achieve best-in-class growth in 2026, with WFE demand visibility extending into 2028.

DRAM is the focal point here. The rising demand for HBM and high-performance memory in AI servers will drive memory manufacturers to expand production and upgrade processes. Equipment companies have the advantage of longer order cycles and higher revenue visibility. The risk is also direct, as any slowdown in capital spending by cloud or memory fabs will quickly lead to downward revisions in mid-term revenue expectations.

The Goldman model shows that AMAT's CY2027E total revenue is $45.972 billion, a 25% year-over-year growth. The DRAM segment is expected to reach $12.4 billion, up 41% year-over-year, serving as the primary driver for equipment upside.

Applied Materials is placed in a relatively positive mix, with the logic not being a significant upward revision, but rather a reduction in short-term expectations. The company announced on June 25 a proposed all-stock acquisition of Synaptics, with a transaction enterprise value of around $7 billion, expected to be completed in mid-2027 pending approval from Synaptics shareholders, among others. Goldman believes that after the pullback in investor expectations, the focus is more on Applied Materials' potential to slightly outperform expectations for the quarter.

ARM and ASML Explain: The Larger the Upsurge, the Harder the Financial Reporting Flaws Are to Overlook

Goldman Sachs maintains a Sell rating on ARM with a 12-month target price of $150, corresponding to 50 times normalized $3 EPS. The pressure mainly comes from two aspects: continuous softness in smartphone demand and higher-than-expected operating expenses.

ARM is still seen by the market as a potential beneficiary of AI and high-performance computing. However, the more direct revenue and profit pressure in the short term still comes from mobile licensing revenue and cost expansion. For stocks that have already been boosted by the AI narrative, the market will pay more attention to whether recent revenue, profit margins, and guidance can be achieved.

ASML's pressure comes from the equipment spending structure. Goldman Sachs expects its quarterly performance and guidance to slightly improve, but it may still underperform peers as WFE spending is skewed towards DRAM. Compared to logic chips and foundry, the strength of equipment demand for DRAM inspection tools is lower. An overall upturn in the equipment cycle does not mean all equipment segments will benefit equally.

Qnity falls between the two. The company's Q1 announcement showed that Q1 2026 net sales were $1.315 billion, and the full-year guidance was raised. Goldman Sachs maintains a positive view on wafer fab utilization improvement and company execution, but the report assesses that after the stock price has risen, the further upside potential and downside risks are closer. For stocks that have already seen an early recovery in trading, financial reports not only need to deliver good results but also provide strong enough guidance for the next phase.

EDA and Qualcomm Also Dragged into the AI Financial Reporting Topic

The AI outlook is not only confined to the GPU, CPU, and memory chains but is also spreading to chip design software and data center chips.

Cadence is one of the companies in the EDA chain that Goldman Sachs favors. Public information shows that the company has raised its 2026 revenue outlook to around 17% year-over-year growth post-Q1 and has partnered with NVIDIA to launch engineering solutions for agentic AI chips and system design. Furthermore, Goldman Sachs expects that with the monetization of agentic AI tools, IP business, and core EDA demand driving growth, the company's 2026 revenue guidance may still be revised upwards.

Qualcomm's data center business is also back in focus. The company previously mentioned in its Investor Day materials that the data center business will contribute billions of dollars in revenue starting from FY27. Goldman Sachs models Qualcomm's FY27 and FY28 data center revenue at $5 billion and $8.2 billion, respectively. For Qualcomm, this represents a growth narrative expanding from smartphone chips to data centers, but orders, customers, and gross margins still need to be continually realized.

The key question this earnings season is straightforward: Can AI capital spending continue to drive semiconductor companies to raise earnings guidance? In the past quarter, the stock prices have already priced in optimistic expectations. Going forward, the strength of server CPU, ASIC, HBM, EDA, and equipment orders must feed into the revenue and profit models for 2026 and 2027 to sustain the elevated valuations.

Whether the Upside Can Catch Up with Stock Prices Will Determine the Degree of Earnings Season Divergence

The highlight of the semiconductor second-quarter earnings season is not whether the market still holds upward expectations, but whether the magnitude of the upside can cover the already increased stock prices.

Goldman Sachs' outlook provides a more nuanced answer. AI capital spending, DRAM/HBM, advanced packaging, and EDA tool monetization are still boosting some companies' profit expectations. After a significant outperformance of the SOX compared to the broader market, the market's tolerance for imperfections is decreasing.

Soft smartphone demand will suppress ARM, WFE tilting towards DRAM will weaken KLA's relative advantage, and supply chain constraints, export restrictions, and geopolitical risks could also affect order fulfillment. Companies like AMD and Applied Materials, which still have room for upward revisions in their models, will face questions about the pace of realization. Companies that have already surged significantly but have limited short-term fundamental resilience will be more susceptible to pressure during earnings season if the guidance is not robust enough.