Introduction
The glass core package substrates market is at an early commercial stage and is moderately consolidated. A small number of companies have built pilot or first commercial production lines for glass cores intended for large, high-performance packages, and these companies account for most of the qualification activity with leading chip designers. Around them sits a wider group of glass makers, substrate producers, display companies and equipment suppliers that provide materials, process technology or early-stage capacity. As of late 2026, no glass core substrate had been publicly confirmed as shipping in a high-volume commercial accelerator or server product, so competitive positions are being shaped by qualification progress, yield and capacity readiness rather than by established revenue.
Competition takes place on several fronts at once. Substrate makers compete on their ability to deliver large, flat, crack-free panels with dense through-glass vias and multilayer build-up at acceptable yield. Glass makers compete on composition control, panel size, flatness and the supply of pre-structured glass. Equipment suppliers compete on via-forming speed, defect control and inspection capability. Across all of these groups, relationships with a few large customers in AI accelerators, data-centre processors and networking chips carry great weight, because a single qualification win can determine the direction of an entire product generation.
Regional strategy also matters. South Korean companies have moved early to build dedicated capacity in both Korea and the United States, Japanese printing and glass companies are using large-panel know-how from displays and photomasks, and Chinese display makers are entering the market with pilot lines and local supply chains. Partnerships between glass suppliers, substrate makers and equipment vendors are common, and capital raising and joint ventures have become important tools for funding the transition to volume production during 2025 and 2026.
Key Player Positioning
The leaders group comprises Absolics (SKC), Samsung Electro-Mechanics, Intel and Corning. Absolics operates a dedicated glass substrate plant in Covington, Georgia, supported by U.S. CHIPS incentives finalised in December 2024, and during 2026 it reported progress in customer reliability evaluations and received further funding from its parent SKC in May and September 2026. Samsung Electro-Mechanics combines a leading position in high-end organic FC-BGA substrates with a glass core pilot line at its Sejong plant and a 2026 joint venture with Sumitomo Chemical's subsidiary Dongwoo Fine-Chem to produce glass cores, which gives it both technical depth and a path to capacity. Intel is the technology pioneer, having announced its glass substrate programme in September 2023 and presented a thick glass-core package with EMIB bridges in January 2026; its position rests on intellectual property, design know-how and its role as a large in-house user. Corning is the leading glass supplier, with broad experience in precision glass for displays and semiconductors and a widening range of partnerships, including a three-year MOU with BOE signed in May 2026 that covers glass-based packaging substrates.
The challengers group includes Dai Nippon Printing, TOPPAN, LG Innotek, BOE, AGC and SCHOTT. DNP and TOPPAN draw on decades of experience in fine patterning and large-panel processing, and both brought glass core pilot lines into operation between late 2025 and mid-2026 in Japan, with DNP targeting mass production in fiscal 2028. LG Innotek has completed product development and runs a pilot line, but reports in January 2026 indicated that it had moved its large-scale commercialisation target to around 2030, which places it behind the leaders in timing. BOE has invested in a glass-based packaging substrate pilot line and has delivered large-format prototypes, supported by its display-scale glass handling expertise. AGC and SCHOTT are established specialty glass producers with packaging-grade glass panels, carriers and structured substrates; they are challengers because they supply materials and semi-finished cores rather than complete substrates, and their success depends on the adoption choices of substrate makers.
The niche players group includes Nippon Electric Glass and LPKF Laser & Electronics. Nippon Electric Glass offers a differentiated glass-ceramic core material, GC Core, introduced in June 2024, which can be drilled with conventional CO2 lasers and targets cost reduction in via formation; its focus on a specific material approach places it in a niche until the material wins broad adoption. LPKF supplies laser-induced deep etching equipment and process know-how for through-glass via formation, and in September 2026 it received a letter of intent from a leading Asian packaging manufacturer for production systems. Its role is specialised but strategically important, because via formation is one of the main yield bottlenecks in glass core production.
Positions within and across these groups remain fluid. The first company to achieve stable high-volume yield with a major AI or networking customer is likely to gain a lasting advantage in qualification and capacity planning, while companies that slip in timing may shift toward material supply, licensing or partnership roles.
Competitor Dashboard
The dashboard rates the twelve companies on four dimensions using a four-level scale of Low, Medium, High and Very High. Product Portfolio reflects the breadth and maturity of glass core and related packaging products; Regional Presence reflects manufacturing and customer reach across regions; Strategic Alliances reflects partnerships, joint ventures and customer collaborations; and Industry Experience reflects the depth of relevant experience in substrates, glass and advanced packaging. The ratings are consistent with the positioning of each company as a leader, challenger or niche player.
The competitive structure of the glass core package substrates market reflects an industry moving from pilot production to its first commercial volumes. Capital intensity is high, process knowledge is scarce, and the number of buyers able to commit to a new substrate technology is small. The five forces are assessed for the period 2026–2032, taking into account that commercial products were still in qualification during 2026 and that competitive behaviour will change as yields mature and second sources are qualified.
Threat of New Entrants: Medium
The barriers to entering glass core substrate manufacturing are substantial. A new entrant needs a cleanroom panel line, laser via formation and wet-etch tools, copper metallisation equipment, build-up lamination capability and inspection systems able to find defects inside transparent material. Absolics' Covington plant, the most advanced dedicated facility, has been supported by a CHIPS incentive of up to USD 75 million finalised in December 2024 and by repeated capital injections from SKC, including the board-approved investment of about KRW 281 billion in September 2026. Beyond capital, the entrant must build yield know-how that has taken incumbents several years, and it must pass customer qualification programmes that can last more than a year.
Several factors nonetheless keep this threat at a medium rather than low level. Display panel makers already handle large, thin glass sheets and operate panel-format cleanrooms, which gives them a shorter route into the business. BOE has built a pilot line for glass-based packaging substrates and, according to 2026 industry reporting, more than ten Chinese companies have entered the glass substrate field, with firms such as WG Tech reporting high-aspect-ratio TGV capability. Government programmes in the United States, Japan, Korea, China and India actively lower the cost of entry through grants, tax credits and subsidised pilot lines; India's August 2025 cabinet approval of a 3D Glass Solutions unit in Odisha is one example. Equipment makers are also packaging their processes as turnkey systems, so a well-funded entrant can buy capability that incumbents had to develop internally. The most likely entrants are therefore not small start-ups but large display, glass or substrate groups with government backing.
The threat also varies by segment of the value chain. Entering as a glass supplier is harder than entering as a substrate fabricator, because few companies possess fusion-draw or precision down-draw capacity and the compositions involved are proprietary. Entry into via processing and metallisation services is somewhat easier, since equipment can be purchased and some display and printed circuit board firms already run large-panel wet processes. Over 2026–2032 the most visible new entrants are likely to be Chinese display and glass groups supported by domestic funds, Japanese firms connected to panel-level packaging consortia, and Indian projects linked to the national semiconductor programme. Their arrival could add capacity after 2028, but each will need several years to reach the yields required by leading AI chip designers.
Bargaining Power of Suppliers: High
Suppliers hold strong bargaining power. The specialty glass used for package cores comes from a handful of producers, chiefly Corning, AGC, SCHOTT and Nippon Electric Glass, each of which offers proprietary compositions and forming processes. Once a substrate maker has qualified its via, plating and build-up processes on a particular glass, changing to another composition requires repeating much of the qualification work, which ties the substrate maker to its glass supplier for the life of a programme. Glass makers are also extending downstream: Nippon Electric Glass offers a glass-ceramic core product, AGC exhibits microvia glass substrates, and Corning has signed a strategic framework with BOE, which gives them leverage as both suppliers and potential competitors.
Equipment supply is similarly concentrated. Laser-induced deep etching and comparable via processes are offered by a small number of firms, and the September 2026 letter of intent that LPKF received for NEXAR LIDE systems from an Asian packaging manufacturer that had already fully qualified the process shows how a qualified tool becomes embedded in a customer's production plan. Inspection and metrology tools for transparent substrates, high-uniformity copper plating systems and build-up films adapted to glass add further single- or dual-source dependencies. Countervailing factors exist, including the entry of Chinese laser equipment makers and the scale of large substrate makers as customers, but over 2026–2032 the combination of few qualified sources and high switching costs keeps supplier power high.
The power of suppliers is also visible in timing. When glass makers or tool vendors allocate scarce development resources, they favour customers with the clearest path to volume, which can delay smaller substrate makers. Glass suppliers set minimum order sizes and composition-specific pricing, and equipment vendors may require prepayment for systems built to order. Because glass compositions and via processes are tightly coupled, a substrate maker cannot easily play suppliers against each other; switching would demand renewed reliability testing by its own customers. Large substrate makers mitigate this through joint development agreements and, in the case of the Samsung Electro-Mechanics joint venture with Dongwoo Fine-Chem, by bringing glass handling expertise into a jointly owned entity.
Bargaining Power of Buyers: High
The buyer side is concentrated. The early volume for glass cores is expected to come from AI accelerators, data centre CPUs, networking ASICs and co-packaged optics, a set of products designed by a small number of companies and assembled by an even smaller number of IDMs, foundries and OSATs. These buyers control the package design, decide whether glass is used at all, and set the reliability criteria that suppliers must meet. Because organic substrates remain available and familiar, a buyer can delay or cancel a glass programme with limited penalty, which gives it leverage in price and allocation negotiations.
Buyers also use multi-sourcing deliberately. Industry reporting during 2026 describes AMD evaluating Absolics material while Samsung Electro-Mechanics sampled glass cores to several large chip designers, and Intel has signalled a willingness to license its glass substrate technology and work with external producers. This allows a designer to compare yield, cost and delivery across suppliers before committing. Large buyers can also provide funding or capacity commitments in return for priority supply, as has occurred in the organic substrate market, which shifts value toward them. Buyer power is moderated only where a supplier has a clear yield or technology lead and where the buyer needs glass to make a specific product feasible; during the early commercial years, those cases are expected to be the exception rather than the rule.
Information also favours buyers. Chip designers see results from several suppliers' samples and can benchmark yields, electrical performance and reliability across them, while each supplier sees only its own data. Buyers decide when to publish or withhold confirmation of a programme, which affects suppliers' ability to raise capital; no glass core customer had been officially confirmed by a supplier as of the third quarter of 2026, even though several were in evaluation. Assemblers such as foundries and OSATs add a second layer of buyer power, because they control the packaging line into which the substrate must fit and can favour suppliers whose panels match their equipment formats. In the early commercial years, suppliers are therefore expected to accept demanding terms on price, reliability and change control in order to secure reference customers.
Threat of Substitutes: High
The main substitute is the organic build-up substrate, which supports nearly all high-performance packages in production during 2026. Organic substrate makers continue to improve core materials, line widths and warpage control, and their capacity, supply chains and reliability records are established. For many applications, especially smartphones, consumer electronics and mid-range processors, an advanced organic substrate is likely to remain the lower-cost and lower-risk choice throughout the forecast period. Silicon interposers used in chip-on-wafer-on-substrate packages, embedded silicon bridges on organic substrates, and panel-level fan-out packages with organic redistribution layers are further alternatives that solve parts of the same interconnect problem.
The threat is strongest where glass offers only incremental benefits and weakest in very large AI packages and optical integration, where glass flatness, stiffness and optical transparency are harder to replicate. Industry timelines also show the substitute's strength: LG Innotek reportedly moved its large-scale commercialisation target for glass substrates from around 2028 to 2030 in January 2026, citing insufficient demand to justify the investment, and SEMI's May 2026 report placed initial production around 2028 in selected high-performance applications. Over 2026–2032 the threat of substitutes therefore stays high, even as glass gains ground in the most demanding packages.
Substitution also works in the opposite direction within packaging technologies. Organic substrate makers such as Ibiden, Unimicron and Shinko have continued to invest in larger, thicker organic cores and finer build-up layers, and some have also entered glass development, which allows them to offer whichever technology a customer prefers. Silicon interposers are mature in high-bandwidth memory integration, and silicon bridge approaches reduce the area of costly silicon while keeping organic cores. For buyers, these options mean that glass must justify its cost and qualification effort with clear performance gains. The degree of substitution threat will decline over time as glass yields improve, but it remains the most important constraint on how quickly glass reaches mainstream products.
Competitive Rivalry: Medium
Rivalry is currently a race on technology and timing rather than on price. Absolics, Samsung Electro-Mechanics, LG Innotek, DNP, TOPPAN and BOE are each trying to be the first qualified high-volume supplier for a flagship package, and each has announced pilot lines or sampling programmes between late 2024 and 2026. Competition shows up as public commitments to schedules, capacity announcements, joint ventures and customer sampling claims. The formation of the Samsung Electro-Mechanics and Dongwoo Fine-Chem glass core joint venture in 2026 and SKC's decision in May 2026 to allocate about KRW 589.6 billion of a capital increase to glass substrates show how rivals are committing resources to secure an early position.
Several features keep rivalry at a medium level for now. The number of credible suppliers is limited, demand in 2026 is still at the sample and qualification stage, and buyers want more than one source, so early movers are not yet competing for the same fixed volume. Regional and customer alignment also separates rivals to some degree: U.S. designers seeking domestic supply look to Georgia and to Intel's planned New Mexico capacity, while Japanese and Korean makers work closely with Asian packaging houses. Rivalry is expected to intensify after 2028 as additional plants reach volume, Chinese entrants scale up and price becomes a larger part of procurement decisions.
The basis of competition is also broader than manufacturing. Rivals compete on access to glass supply, on intellectual property positions in via formation and metallisation, on proximity to major packaging lines and on their ability to finance years of pre-revenue operation. Intel's January 2026 demonstration of a thick glass core with embedded EMIB bridges and its stated openness to licensing add another dimension, since technology owners may compete by enabling others rather than by producing at scale. Japanese suppliers such as DNP and TOPPAN emphasise panel-level processing and existing relationships with Japanese and Taiwanese assemblers, while Korean suppliers lean on their positions in organic flip-chip substrates and their ties to U.S. chip designers. This diversity of strategies limits head-to-head price competition in the near term.
Taken together, the forces describe a market in which value is concentrated at the two ends of the chain, with specialty glass and equipment suppliers on one side and a small group of chip designers and assemblers on the other, while substrate makers carry most of the capital and yield risk in between.
Threat of New Entrants: Medium; High capital and yield-learning barriers, partly offset by display makers' glass handling skills and government-funded pilot lines.
Bargaining Power of Suppliers: High; Few qualified specialty glass and TGV equipment sources; process qualification ties substrate makers to one glass and tool set.
Bargaining Power of Buyers: High; A small group of chip designers and assemblers control package design, set qualification terms and multi-source deliberately.
Threat of Substitutes: High; Advanced organic substrates, silicon interposers and bridges remain established, lower-risk alternatives for most packages through 2032.
Competitive Rivalry: Medium; Rivals compete on qualification timing and capacity commitments; price competition is limited while demand remains at the sampling stage.
