Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/kla-at-goldman-sachs-communicopia-technology-conference-september-10-2025.23669/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

KLA at Goldman Sachs Communicopia + Technology Conference, September 10, 2025

hist78

Well-known member
The audio replay is available at:

https://ir.kla.com/


The following summary was generated by Google NotebookLM AI, with some minor corrections I made for misspellings.

Executive Summary​

This document summarized the key themes and strategic insights from KLA's presentation at the Goldman Sachs Communicopia + Technology Conference. The discussion, led by CFO Brent Higgins and President of Semiconductor Products and Solutions Ahmed Khan, outlines a period of significant outperformance for KLA, driven by powerful secular trends in the semiconductor industry.

The central thesis is that the "AI inflection" is fundamentally tilting the industry towards greater process control intensity, benefiting KLA disproportionately. This is manifesting through several key drivers: the dramatic increase in chip die sizes, the multi-layered complexity of High Bandwidth Memory (HBM), and the evolution of advanced packaging into a front-end-like, high-stakes integration process. As a result, KLA is tracking significantly ahead of its long-term goals for market share, expecting to capture nearly 8% of Wafer Fab Equipment (WFE) spending in 2025, well above its 2026 target of 7.25%.

Financially, the company is executing with high precision, delivering gross margins of 62.5% and operating margins above its target model, even while absorbing tariff impacts. The services business continues to show resilient double-digit growth, underpinning a robust capital return program that includes a consistently growing dividend and systematic share buybacks. KLA's strategic priorities are focused on innovation, customer collaboration at the leading edge, and driving internal productivity to maintain its differentiated business model and capitalize on the growing complexity of high-performance computing.

I. Business Outlook and Financial Performance​

KLA's financial outlook for 2025 is shaping up as expected, with incremental improvements reflected in recent results and guidance. The business is characterized by strong execution and momentum in key growth areas.

Key Financial Metrics & Business Trends:

  • Quarterly Performance: The June quarter was described as a "good quarter," leading to a raise in guidance for the September quarter.
  • Gross Margins: The guidance for the year is 62.5%, which includes a 50 to 100 basis point negative impact from tariffs that the company is actively working to mitigate.
  • Operating Margins: KLA is currently performing above the top end of its long-term target model, driven by a 40% to 50% incremental operating margin on revenue growth.
  • Service Business: Despite US export controls denying access to certain fabs, the service business is performing in the double-digits. The long-term growth target remains 12-14%.
  • Advanced Packaging Revenue: This segment shows significant momentum. After generating over $500 million in 2024, the forecast for 2025 has been repeatedly raised, with the latest guidance targeting $925 million.
Corporate Priorities:

  1. Customer Support: Supporting customers at the leading edge is the primary focus, ensuring collaboration to meet the demands of new technologies like the 2-nanometer node.
  2. Innovation: Described as the "lifeblood of KLA," investment in R&D is critical for creating the differentiation that drives the company's financial model. R&D spending is expected to remain in the 12-13% range of revenue.
  3. Productivity & Leverage: The company is actively seeking opportunities to "bend the cost curves" by leveraging new productivity tools in both technical and non-technical areas to enhance its business model.

II. The AI Inflection and Rising Process Control Intensity​

The core driver of KLA's recent outperformance is the AI-driven shift in semiconductor manufacturing, which inherently demands more sophisticated process control. This trend has caused KLA's share of WFE to rise from the low 6% range in 2021-2022 to a projected ~8% in 2025. This increase is attributed to several technical inflections.

The Die Size Effect (D0)​

As explained by Ahmed Khan, the increasing size of individual chip dies is a major factor. Larger dies mean fewer chips per wafer, making each die more valuable and a single "killer defect" far more impactful on overall yield.

"If you had a wafer and you had two dies on it only and you had one killer defect that that killed the die your yield would be 50%. And if you had a wafer that had 10 dies and you had one killer defect your yield would be 90%. So as die sizes increases...process control intensity goes up because of the criticality of die size." - Ahmed Khan

This effect is not isolated to the leading edge; it is occurring at mature nodes (5nm, 4nm, 3nm) as well as the advanced 2nm node.

High Bandwidth Memory (HBM) Complexity​

HBM architecture is fundamentally more complex than traditional DDR memory, driving process control needs at multiple stages:

  • Logic Base Die: HBM stacks begin with a logic base die (currently 7-10nm, moving toward 3nm), which introduces logic-level process control intensity into memory manufacturing.
  • Stacked Dies: Multiple DRAM dies are stacked on top of the logic base. The value of the entire package increases with each added layer, making it critical to inspect each step to avoid wasting previously stacked, high-value dies.
  • Integration: The final integration of HBM with a logic die (e.g., a GPU) in a CoWoS (Chip-on-Wafer-on-Substrate) fab is a high-risk process. Losing an HBM stack during integration is a catastrophic and costly failure.
  • Performance & Reliability: High-performance compute (HPC) applications have rigorous reliability requirements. The performance of an entire HBM stack is limited by its weakest die, and there is little opportunity to "down-sell" a slightly underperforming device, unlike in consumer markets.

Advanced Packaging Integration​

Advanced packaging has transformed from a back-end process into a highly complex, front-end-like system integration challenge.

"You don't take a final wafer cut it and put it in a package and put it on a PCB you start building from there onwards with memory with other things and this inflection will continue and that's that's why the packaging inflection is continuing." - Ahmed Khan

The high cost of the components being integrated (GPUs, HBMs) and the extreme difficulty of repairing a faulty package on a CoWoS wafer make the return on investment for inspection extremely high. This dynamic has fueled demand for KLA's front-end capabilities in the back-end packaging environment.

Proliferation of Leading-Edge Designs​

The economics and performance benefits of the latest nodes are attracting a wider customer base. The 2-nanometer (N2) node is seeing around 15 customers engaged in designs, including roughly 10 in high-performance computing. This creates several challenges that benefit KLA:

  • High-Mix Fabs: Managing numerous different designs and process flows simultaneously is highly complex for foundries.
  • Constrained Capacity: With the N3 node highly utilized, customers have less flexibility to migrate or reuse capacity for N2, necessitating new investment.
  • Market Windows: The pressure to deliver multiple designs within tight market windows increases the need for robust process control to optimize yields and wafer starts. KLA projects that the N2 node could be the largest node ever in its first three years from a design perspective.

III. Market Segment Analysis and WFE Outlook​

KLA provided a qualitative outlook for Wafer Fab Equipment (WFE) spending and detailed the dynamics across key market segments.

WFE Forecast:

  • 2025: Expected to grow in the "mid-single digits" compared to 2024. This growth is partly driven by advanced packaging, which KLA notes may need to be analyzed as a separate market going forward due to its size ($10-11 billion).

  • 2026: While not quantified, the company expects WFE to see further growth, driven by continued momentum at the leading edge and a broadening of investment.
Market Segment Dynamics:

1758584518067.png


IV. Strategic Portfolio and Services

Non-Process Control Businesses (EPC)​

The Electronics, Packaging, and Component (EPC) business segment, which accounts for roughly 30% of revenue, is undergoing a strategic shift to align with market trends.

  • Shift to Packaging: The entire EPC portfolio is being reoriented towards the high-growth advanced packaging market.
  • Value Chain Alignment: KLA is aligning its products to follow the migration of value from traditional PCBs to advanced substrates and ultimately to direct silicon integration.

  • Component Businesses:
    • The specialty/MEMS business ($500-$600 million) leverages technology for applications like die singulation, which is transitioning from specialty uses to dicing memory and logic chips.
    • The PCB business is showing "signs of life" after weakness related to mobile markets.
    • Component inspection is growing in line with overall semiconductor revenue expansion toward a potential $1 trillion industry.
  • Organizational Integration: The formerly separate EPC group has been consolidated with the core semiconductor business to better align with customers who now manage back-end packaging with front-end-like discipline.

Services Business​

The services business is a key pillar of KLA's financial model, demonstrating stability and growth.

  • Resilient Growth: The business is growing at a double-digit rate despite near-term headwinds from export controls restricting access to certain fabs. The long-term growth target is 12-14%.

  • Unique Model: 75% of service revenue is contract-based. The business correlates more closely with long-term semiconductor revenue growth than with cyclical WFE spending.

  • Value Proposition: KLA sells system performance, including spec performance, matching across the fleet, and high availability. Process control tools rarely idle, as optimizing yield is always a priority for fabs, ensuring high demand for service contracts.

  • Dividend Underpinning: The predictable, recurring revenue from the services business provides the company with the confidence to maintain its policy of consistently growing its dividend.

V. Capital Allocation and Financial Strategy​

KLA maintains a disciplined and shareholder-friendly capital allocation strategy focused on funding innovation while returning significant capital.

  • Capital Return Framework:The company is committed to returning 85% of the free cash flow it generates to shareholders.
    • Dividends: The dividend is a priority, having been raised for 16 consecutive years. The target is to grow the dividend at approximately 15% annually, with a long-term payout ratio of 25-30% of free cash flow.
    • Buybacks: The remaining capital is allocated to an ongoing, systematic share buyback program.
  • Opex Management: Opex is managed to deliver 40-50% incremental operating margins. Investment priorities are focused on R&D to drive differentiation, while leverage is sought in non-technical spending areas.

  • M&A Philosophy: The company's primary focus is on investing in its current businesses where it sees significant opportunities. While KLA has made small, augmenting acquisitions (e.g., in software and chemical process control), the bar for M&A is high and is evaluated against the return available from buying back its own stock. The regulatory backdrop is also a key consideration.
 
Last edited:
KLA was represented by Brent Higgins, its CFO, and Ahmad Khan, who oversees KLA’s systems business and channels, at this Goldman Sachs conference.

I believe they unintentionally revealed that TSMC’s N2 node has about 15 customers, including roughly 10 in high performance computing.

Two other interesting points they mentioned are:
  1. 1. Due to the already high utilization of TSMC’s N3 node, there isn’t much equipment that can be moved to N2 production. It benefits KLA.
2. Since N7 (around 2019–2020), for a period of time the number of new designs adopting the newest nodes had decreased due to high development costs. However, KLA noted that development costs have come down significantly, and with tight market windows and high performance requirements, they expect many more designs to adopt the latest leading edge nodes more quickly.
 
Last edited:
Back
Top