Quick Summary
Covering the latest research from top Wall Street investment banks

Taiwan ODM/Brands 3-month preview: March revenue expected to rise +13% MoM on average, with AI servers and liquid cooling driving YoY growth

Institution
Goldman Sachs
Date
2026-04-02
Authors
Allen Chang, Verena Jeng, Xuan Zhang, Ting Song, Yifan Hu
Company
Taiwan ODM/Brands coverage universe
Ticker
-
Industry
Taiwan tech hardware, AI servers, and PC supply chain
Rating
Buy: Hon Hai, Wiwynn, Wistron, AVC; Neutral: Gigabyte, Quanta, Inventec, Compal, ASUS; Sell: Pegatron
NeutralLow confidenceThe report expects Taiwan AI server and PC supply-chain revenue to recover in March, with solid YoY growth driven by higher AI server rack shipments, greater ASIC AI server penetration, and increased liquid-cooling adoption, while memory price increases continue to weigh on consumer electronics such as PCs and smartphones.
AuthorsAllen Chang, Verena Jeng, Xuan Zhang, Ting Song, Yifan Hu
CoverageAsia-Pacific
Asset classesEquity
Business segmentsAI servers、ASIC AI servers、GPU AI servers、PC and notebook ODM、smartphones and consumer electronics、liquid cooling and thermal components、general servers
Research firm divisions/subsidiariesGoldman Sachs(Other)

AI summary card

Taiwan ODM/Brands 3-month preview: March revenue expected to rise +13% MoM on average, with AI servers and liquid cooling driving YoY growth

Goldman Sachs expects March revenue for 10 Taiwan technology supply-chain companies to recover from a low Lunar New Year base, with the main incremental drivers coming from AI server racks, ASIC AI server penetration, and higher liquid-cooling adoption, while memory price increases continue to pressure PC and smartphone demand.

Overall view is slightly positive but divergent: AI servers and cooling-related names are better positioned than companies with higher PC and consumer electronics exposure.
Artificial intelligenceAI serversTaiwan ODMsliquid coolingPC supply chainmemory price increases
  • Average revenue for the 10 companies is expected to be +13%, -2%, and +8% MoM in March, April, and May 2026, respectively, and +28%, +26%, and +23% YoY.
  • Goldman Sachs remains constructive on the Apple supply chain, AI server component specification upgrades, and rising ASIC AI server penetration.
  • AVC and Wistron are expected to outperform relatively, supported by continued volume growth in GPU and ASIC AI servers and higher liquid-cooling penetration.
  • Compal, ASUS, and Inventec face pressure from higher memory costs and slower PC end-demand growth due to their higher PC exposure.
  • The rating split is clear: Buy for Hon Hai, Wiwynn, Wistron, and AVC; Neutral for Gigabyte, Quanta, Inventec, Compal, and ASUS; Sell for Pegatron.

Report interpretation

Overview

This report previews the next three months of revenue for 10 Taiwan AI server, PC, and consumer electronics supply-chain companies covered by Goldman Sachs, and updates earnings forecasts for Wistron, Quanta, Pegatron, and Wiwynn. The core view is that February revenue was weighed down by fewer working days during Lunar New Year, but March should recover meaningfully; medium-term growth will mainly be driven by AI server rack volume ramp, higher ASIC AI server penetration, and rising demand for liquid cooling.

Core views

The report is constructive on the Apple supply chain because of demand support from form-factor changes such as foldable phones in 2026 and iPhone 20 in 2027, and it is also constructive on AI server component specification upgrades and higher ASIC AI server penetration. At the stock level, Hon Hai, Wiwynn, Wistron, and AVC are rated Buy; Quanta, Gigabyte, ASUS, Compal, and Inventec are Neutral; Pegatron is Sell. The main negative factor is rising memory prices, which will increase PC and smartphone bill-of-materials costs and weigh on end demand and consumer electronics revenue growth.

Analysis framework

The report uses a 3-month revenue preview framework, combining each company’s actual February revenue, Lunar New Year seasonality, AI server shipment cadence, PC and consumer electronics demand, memory price changes, and management guidance to estimate month-on-month and year-on-year revenue changes from March to May 2026. Valuation is mainly based on 12-month target prices and forward-year P/E multiples, with reference to the relationship between EPS growth and valuation for peers in PCs, servers, and technology hardware.

Methodology notes

  • Valuation methodNear-term P/E and peer EPS growth correlation

    Derive 12-month target prices using 2026E or 2027E P/E multiples

    Most target prices use forward-year P/E multiples and refer to the correlation between EPS growth and P/E multiples among peers in PCs, servers, or technology hardware; Wistron and Pegatron roll the base year forward to 2027E.

  • Fundamental forecast3-month revenue preview

    Estimate each company’s March to May 2026 revenue trends on a MoM and YoY basis

    The forecast framework incorporates post-Lunar New Year workday normalization, AI server rack volume ramp, ASIC AI server penetration, liquid-cooling adoption, PC demand, and memory cost pressure.

  • Disclosure frameworkGS Factor Profile

    Compare stock attributes across Growth, Financial Returns, Multiple, and Integrated dimensions

    Goldman Sachs Factor Profile uses standardized rankings of sales, EBITDA, and EPS growth, ROE, ROCE, and CROCI, as well as P/E, P/B, and EV/EBITDA, to provide investment context relative to the market and peers.

Asset mapping & comparison

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

  • Hon Hai (2317.TW)
    Buy on CL
    Strengths
    Leading AI server market share, March revenue expected at +39% MoM / +50% YoY, and seasonal recovery in consumer electronics also provides support.
    Weaknesses
    February revenue was below expectations and is still affected by consumer electronics EMS competition and global capacity ramp-up.
    Comparison
    Among the 10 companies, March YoY growth is expected to be among the strongest, driven by AI server rack volume ramp.
    Risks
    Slower-than-expected AI server business ramp, weaker-than-expected EV overall solutions, slower-than-expected global capacity ramp, and more intense consumer electronics EMS competition.
  • Quanta (2382.TW)
    Neutral
    Strengths
    New AI server racks started ramping in 4Q25, and 1Q26E revenue is expected to be NT$688bn, or +42% YoY / +8% QoQ.
    Weaknesses
    Valuation is considered relatively fair, limiting upside to the rating.
    Comparison
    AI server momentum is strong, but the valuation is less attractive than that of Buy-rated names.
    Risks
    PC market recovery strength, AI server ramp speed, and changes in general server demand.
  • AVC (3017.TW)
    Buy
    Strengths
    Higher liquid-cooling penetration, AI server rack ramp, new ASIC AI server models, and higher adoption of VC in smartphones together drive growth.
    Weaknesses
    February revenue declined MoM due to fewer working days during Lunar New Year and the off-season for consumer electronics.
    Comparison
    The report expects AVC to be one of the relative outperformers, supported by the dual themes of AI servers and liquid cooling.
    Risks
    Slower-than-expected liquid-cooling adoption, lower-than-expected AI server demand, stronger competition, and slower-than-expected recovery in general servers.
  • Wiwynn (6669.TW)
    Buy
    Strengths
    Leading position in ASIC AI servers, concentrated exposure to CSP general servers, and diversification into GPU AI servers.
    Weaknesses
    Revenue growth is sensitive to server demand and the pace of client expansion.
    Comparison
    Compared with companies with higher PC exposure, Wiwynn has a higher server purity and is more directly supported by AI compute demand.
    Risks
    AI server demand growth weaker than expected, general server demand recovery slower than expected, and intensified competition among server ODMs.
  • Wistron (3231.TW)
    Buy
    Strengths
    February revenue rose 177% YoY, with a significant ramp in AI server racks; 2026-28E revenue and net profit have both been raised.
    Weaknesses
    Rack-scale AI server assembly margins are usually lower than those of other product lines, and 2026E gross margin is cut to 5.9%.
    Comparison
    The report expects Wistron and AVC to outperform relatively, driven by continued ramp-up in GPU and ASIC AI servers.
    Risks
    PC market recovery weaker than expected, AI server demand ramp slower than expected, and general server demand weaker than expected.
  • Gigabyte (2376.TW)
    Neutral
    Strengths
    AI servers continue to expand, enterprise customers are accelerating AI deployment, and the company plans to expand its customer base and global capacity in 2026E.
    Weaknesses
    PC end demand is pressured by rising memory prices, and 2Q26 revenue is expected to be only +2% YoY.
    Comparison
    The AI server direction is positive, but PC demand and memory costs limit near-term growth elasticity.
    Risks
    PC market recovery strength, gaming PC growth pace, and AI server ramp speed.
  • ASUS (2357.TW)
    Neutral
    Strengths
    The product mix is upgrading toward AI PCs, gaming PCs, and commercial PCs, while gradually diversifying into AI servers.
    Weaknesses
    The end of the Win10 replacement cycle, rising memory costs, and a small AI server shipment share limit near-term growth.
    Comparison
    Compared with more server-pure companies, ASUS is still more exposed to PC demand and the consumer electronics cycle.
    Risks
    PC market growth strength, AI and gaming PC growth pace, and AI server ramp speed.
  • Compal (2324.TW)
    Neutral
    Strengths
    AI server business visibility is relatively high, and management hopes to mitigate memory-price pressure through PC product-mix upgrades.
    Weaknesses
    Consumer electronics still account for most revenue, February revenue was below expectations, and PC end demand remains weak.
    Comparison
    AI server growth is offset by the high consumer electronics mix, making it weaker than more server-pure names.
    Risks
    PC market recovery strength, AI server ramp speed, and tablet demand strength.
  • Inventec (2356.TW)
    Neutral
    Strengths
    AI servers are expanding, and the ramp-up of US L10/L11 capacity will help shipments and client expansion.
    Weaknesses
    2Q26 growth is expected to be only +3% YoY, and rising memory costs may pressure PC end demand.
    Comparison
    There is an AI server growth trend, but overall revenue growth is weaker than that of companies with higher AI exposure.
    Risks
    PC market recovery strength, AI server ramp speed, and general server demand strength.
  • Pegatron (4938.TW)
    Sell
    Strengths
    AI server customers are expanding from Neo Cloud to CSP, and global capacity plans in the US, Vietnam, and Mexico are advancing.
    Weaknesses
    Smartphone and PC businesses are pressured by memory shortages and price increases, weighing on consumer electronics growth.
    Comparison
    Although AI server revenue forecasts have been raised, consumer electronics pressure keeps the rating below other AI-server beneficiaries.
    Risks
    For a Sell rating, upside risks include stronger-than-expected smartphone recovery, faster-than-expected AI server ramp, and stronger-than-expected consumer electronics demand.

Key data

  • Average revenue MoM forecast for 10 companiesMarch 2026 +13%, April -2%, May +8%March is supported by the recovery in workdays after Lunar New Year and AI server shipments.
  • Average revenue YoY forecast for 10 companiesMarch 2026 +28%, April +26%, May +23%YoY growth is driven by higher AI server rack shipments, ASIC AI server penetration, and liquid-cooling adoption.
  • Hon Hai March revenue forecast+39% MoM / +50% YoYSupported by continued AI server rack shipments and seasonal recovery in consumer electronics; maintain Buy on CL, 12-month target price NT$400.
  • Quanta 1Q26E revenue forecastNT$688bn, +42% YoY / +8% QoQNew AI server racks have ramped since 4Q25; maintain Neutral due to relatively fair valuation, 12-month target price NT$299.
  • AVC target priceNT$2,259Target price is based on 21.2x 2027E P/E; growth is driven by liquid cooling and higher smartphone VC adoption, with a Buy rating.
  • Wiwynn target priceNT$6,400Target price is based on 22.0x 2026E P/E; exposure to ASIC AI servers and CSP general servers is strong, with a Buy rating.
  • Wistron earnings forecast revision2026-28E net profit raised by 4% / 12% / 10%, revenue raised by 14% / 8% / 7%Mainly driven by stronger AI server rack shipments and scale benefits; 12-month target price lowered to NT$246 but Buy is maintained.
  • Pegatron target price revision12-month target price raised to NT$66.2, previous NT$61Despite higher AI server revenue forecasts, consumer electronics remains pressured by memory costs, so the Sell rating is maintained.

Impact & implications

The investment implication is that, within Taiwan’s technology hardware chain, AI server and cooling-related names have stronger revenue visibility and YoY growth, while companies with higher PC and smartphone exposure are more vulnerable to memory price increases and slower end-demand growth. The report recommends focusing on structural opportunities from AI server racks, ASIC AI servers, liquid cooling, and CSP customer expansion, while staying cautious on companies with a high consumer electronics mix or valuation that is already relatively full.

Risks

  • AI server demand or rack-level shipments ramp slower than expected.
  • ASIC AI server penetration increases less than expected.
  • Liquid-cooling adoption rises slower than expected.
  • Memory price increases further pressure PC and smartphone demand.
  • General server demand recovery is slower than expected.
  • Competition among server ODMs and consumer electronics EMS players intensifies.
  • Global capacity expansion or the ramp-up of new capacity in the US, Vietnam, and Mexico is slower than expected.

What to watch

  • Whether each company’s monthly revenue from March to May 2026 follows the average +13%, -2%, and +8% MoM path.
  • The shipment cadence of AI server racks, GPU AI servers, and ASIC AI servers.
  • Changes in liquid-cooling penetration and the revenue contribution of cooling-chain companies such as AVC.
  • The degree to which memory prices pressure demand for PCs, smartphones, and consumer electronics.
  • CSP customer expansion, Neo Cloud customer demand, and the pace of enterprise AI deployment.
  • The subsequent delivery of revised earnings forecasts for Wistron, Pegatron, Quanta, and Wiwynn.
Zhejiang ICP No. 2022035445-5
Disclaimer: Market data, charts, indicators, research views, and other information provided on this website are intended solely for information display, research communication, and educational reference. They should not be regarded as personalized investment advice, securities recommendations, trading instructions, solicitations, or guarantees of return. While we strive to improve the reliability of our data and content, such information may still be subject to delays, errors, incompleteness, or untimely updates due to source differences, methodological limitations, system processing, or market volatility. Users should exercise independent judgment based on their own circumstances and bear all risks and responsibilities arising from the use of this website.

Settings

Sign in to view recent logins