AI drives high growth in HPC CCL/PCB market size

Overview

This report covers the Asia PCB, CCL, substrate, test equipment, and passive component supply chain, with the core theme being rapid expansion of the total addressable market for HPC CCL/PCB driven by AI. The report evaluates supply chain opportunities from the perspectives of AI servers, regular servers, switches, LEO, migration in high-grade material tiers, substrate demand, data center power architecture, and test equipment, and supplements it with J.P. Morgan coverage companies and peer valuation framework notes.

Core views

The core view is that AI-related computing infrastructure is increasing the demand and value content of HPC CCL/PCB, especially in core AI server projects, higher-grade materials such as M8/M9, and data center power testing. The TAM chart presented in the report indicates total size is expected to grow from US$3.89bn in 2025 to US$9.98bn in 2027e; the share of major AI server projects is expected to rise from 43% to 54%, and combined M8 and M9 share from 41% to 73%. In testing equipment, Chroma is described as the leader in AI data center power testing, with product coverage including data center power equipment testing, DC converter testing, BBU testing, and CPO test flow.

Analysis framework

The report mainly uses TAM decomposition, year-on-year growth rates, CAGR, demand share composition, material-grade migration, and peer valuation comparison. TAM is split into major AI server projects, regular servers, switches, and LEO. Material grades are split into M9, M8, and M7-and-below. The testing equipment section combines data center power structure and product coverage breadth in its analysis. For some covered companies, earnings forecasts are based on J.P. Morgan estimates, while other companies use Bloomberg Finance L.P. consensus expectations.

Methodology notes

• Market sizingTAM decomposition

  Segmenting HPC CCL/PCB demand by use case

  The report decomposes HPC CCL/PCB TAM into major AI server projects, regular servers, switches, and LEO, and compares market size, year-over-year growth, and structural shares for 2025, 2026e, and 2027e.

• Growth estimationCAGR and year-over-year growth

  Compound and annual growth for 2025-2027e

  The report uses 2026e and 2027e YoY growth rates and 2025-2027 CAGR to gauge growth intensity across subsegments; total TAM CAGR is about 60%.

• Product structure analysisMaterial-grade migration

  Changes in M9, M8, M7-and-below mix

  The report splits demand by material grade, showing rising shares of M8 and M9 and declining shares of M7 and below, indicating a shift toward higher-grade HPC materials.

• Valuation comparisonPeer valuation table

  Comparing forecasts of covered companies and peers

  The report discloses valuation table share prices as of April 9; covered companies under J.P. Morgan use internal estimates, while other companies use Bloomberg Finance L.P. consensus expectations.

Asset mapping & comparison

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

• Elite Material Co (2383.TW)

  PCB/CCL company covered by J.P. Morgan

  Strengths

  Potential supply chain candidate benefiting from expansion in HPC CCL/PCB demand and the shift to higher-grade materials.

  Weaknesses

  The input materials do not disclose company-level earnings forecasts or rating conclusions.

  Comparison

  Compared with peers such as Isupetasys, Tripod Tech, and Flexium in the valuation table.

  Risks

  AI server demand, pace of material upgrades, and volatility in industry valuation could affect performance.

• Tripod Technology Corp (3044.TW)

  PCB-related company covered by J.P. Morgan

  Strengths

  Belongs to the PCB supply chain and may benefit from growth in AI server, regular server, and switch demand.

  Weaknesses

  The input materials do not provide a standalone company-level fundamental summary.

  Comparison

  Compared with peers such as Elite Material, Isupetasys, and Flexium in the valuation table.

  Risks

  Lower-than-expected downstream demand, pricing competition, and slower-than-expected product mix upgrades.

• Flexium Interconnect Inc (6269.TW)

  PCB-related company covered by J.P. Morgan

  Strengths

  Falls within the PCB chain covered by the report.

  Weaknesses

  The input materials do not specify its specific share or competitive advantage in HPC CCL/PCB.

  Comparison

  Compared with peers such as Elite Material, Isupetasys, and Tripod Tech in the valuation table.

  Risks

  Order mix, capacity utilization, and valuation changes may drive volatility.

• Unimicron (3037.TW)

  Substrate-related company covered by J.P. Morgan

  Strengths

  Within the global substrate demand analysis set, potentially linked to expansion in high-performance computing demand.

  Weaknesses

  The input materials do not disclose separate rating, target price, or earnings leverage.

  Comparison

  The report notes it is compared with substrate peers such as Ibiden and Samsung Electro.

  Risks

  Substrate demand, capacity expansion, customer project progress, and peer competition.

• Chroma ATE (2360.TW)

  Test equipment company covered by J.P. Morgan

  Strengths

  The report explicitly says Chroma is a leader in AI data center power testing and has a comprehensive product portfolio for data center power equipment testing.

  Weaknesses

  Specific earnings forecasts and valuation multiples are not disclosed in the input materials.

  Comparison

  The report compares Chroma ATE with Advantest, Teradyne, Keysight, and Delta Elec.

  Risks

  AI data center capex pace, conversion of test equipment orders, and intensifying competition.

• YAGEO (2327.TW)

  Passive components company covered by J.P. Morgan

  Strengths

  Included in passive component peer comparison and may be affected by changes in electronics supply chain demand.

  Weaknesses

  Details for the passive components section in the input materials are limited.

  Comparison

  Compared with peers such as Murata, Taiyo Yuden, and Samsung Electro.

  Risks

  Passive component cycle risk, price volatility, and uncertainty around end-demand recovery.

Key data

• Total HPC CCL/PCB TAM2025 US$3.89bn; 2026e US$6.33bn; 2027e US$9.98bnCorresponding to 63% 2026e YOY growth, 58% 2027e YOY growth, and a 2025-2027 CAGR of 60%.
• Major AI server project TAM2025 US$1.69bn; 2026e US$3.07bn; 2027e US$5.39bn2025-2027 CAGR is 79%, and share is expected to rise from 43% to 54%.
• Regular server TAM2025 US$0.73bn; 2026e US$0.99bn; 2027e US$1.28bn2025-2027 CAGR is 32%, and share is expected to decline from 19% to 13%.
• Switch TAM2025 US$1.17bn; 2026e US$1.62bn; 2027e US$2.24bn2025-2027 CAGR is 39%, and share is expected to fall from 30% to 22%.
• LEO TAM2025 US$0.30bn; 2026e US$0.65bn; 2027e US$1.07bn2025-2027 CAGR is 89%, the highest growth among the subsegments.
• M8 material demand2025 US$1.59bn; 2026e US$3.66bn; 2027e US$6.49bnShare is expected to rise from 41% in 2025 to 65% in 2027e.
• M9 material demand2025 0; 2026e US$0.24bn; 2027e US$0.77bnShare is expected to rise from 0% in 2025 to 8% in 2027e.
• M7-and-below material demand2025 US$2.30bn; 2026e US$2.43bn; 2027e US$2.72bnShare is expected to decline from 59% in 2025 to 27% in 2027e.
• Chroma positioningLeader in AI data center power testingThe report states Chroma has a comprehensive product portfolio in data center power equipment testing.

Impact & implications

If the report’s TAM and structure migration views materialize, the Asia PCB, CCL, and substrate supply chain is likely to benefit from AI server and high-performance computing upgrades, and suppliers of higher-grade materials may have stronger revenue leverage than lower-grade material segments. On testing equipment, the increasing complexity of data center power architecture is favorable to vendors with a full test product portfolio. The passive components segment is also included in supply-chain monitoring, but the disclosed details in the input materials are less extensive than for HPC CCL/PCB and testing equipment.

Risks

• AI server and data center capital expenditure below expectations could weaken HPC CCL/PCB demand growth.
• If the migration to higher-grade materials M8/M9 is slower than expected, improvements in product mix and TAM expansion could be reduced.
• Volatility in LEO, switch, and regular server demand could cause segment forecasts to miss.
• Testing equipment orders may be affected by changes in data center power architecture, customer validation cycles, and competition.
• Some peer data in the report comes from Bloomberg Finance L.P. consensus expectations, so forecast frameworks may differ from those used for J.P. Morgan-covered companies.
• J.P. Morgan discloses possible business relationships with covered companies; investors should consider potential conflicts of interest.

What to watch

• Whether main AI server project TAM for 2026e and 2027e materializes with YoY growth rates of 82% and 75%.
• Whether M8 share rises from 41% in 2025 to 65% in 2027e and M9 rises from 0% to 8%.
• Whether LEO TAM maintains a high 89% CAGR for 2025-2027.
• Order progress for Chroma in AI data center power testing, DC converter testing, BBU testing, and CPO test flows.
• Updates in ratings, target prices, and earnings forecasts for covered companies such as Elite Material, Tripod Tech, Flexium, Unimicron, and YAGEO in future company reports.