What the 3+N+3 Notation Actually Means
The notation describes a build-up construction. The number before the plus sign is the count of laser-drilled build-up layers on one side, the middle value is the number of conventional core layers, and the number after the plus sign is the build-up count on the other side. A three plus N plus three board therefore has three sequential build-up layers on each face, laminated around a conventional multilayer core.
That construction exists because routing density is limited by via geometry. On an ordinary multilayer board all holes pass through the entire stack, so each hole consumes routing area on every layer it crosses. Build-up layers use small laser-drilled vias that connect only the layers they need, which frees the layers underneath. The more build-up layers, the more freedom the designer gains, and the more process steps the factory must perform.
Understanding this is the key to understanding the price. The cost of an HDI board is driven primarily by the number of sequential laminations and the number of laser drilling and plating operations, not by the quantity of laminate it contains. A three plus N plus three structure requires those operations to be repeated three times per side, with imaging, plating, inspection and lamination between each one.
There is also a yield arithmetic that customers rarely see. If each build-up layer step has a ninety-five percent probability of being defect-free, the probability that all six steps succeed on a given panel falls below seventy-five percent. High-yield HDI manufacturing is therefore a matter of making each individual step extremely reliable, because the effect of six sequential operations multiplies rather than adds.
Why Each Additional Build-Up Layer Compounds
The first build-up layer is the most disruptive, because entering build-up technology at all requires laser drilling capability, fine-line imaging, and the handling discipline for thin panels. Once that capability exists, adding a second layer is largely a matter of repeating controlled steps. In practice, however, each additional layer still increases cost, because the number of process operations grows and the accumulated registration budget tightens.
Registration is where the compounding becomes severe. Every lamination cycle introduces dimensional movement, and every laser drilling step must land within tolerance of the features created in the previous step. With three build-up layers per side, the cumulative error tolerance is smaller than the sum of the individual tolerances, because the requirement is that the first layer still aligns with the last. The process window is therefore narrower and the yield risk higher.
This is why manufacturers rarely quote an HDI price purely from a stackup drawing. The achievable yield depends on the factory’s specific process capability, and two factories with apparently similar equipment can deliver materially different results on the same stackup. The quoted price reflects that difference.
Drill capacity is a practical constraint as well. Laser drilling is slower than mechanical drilling per hole, and a dense HDI panel can contain hundreds of thousands of microvias. Capacity planning for HDI is therefore dominated by laser time rather than by lamination, which is one reason lead times for HDI work behave differently from lead times for conventional multilayer boards.
Laser Vias and the Cost of a Hole
A laser-drilled microvia is typically between 0.075 and 0.15 millimetres in diameter, with a depth limited by the aspect ratio the process can achieve reliably. Creating it requires removing dielectric precisely without damaging the copper beneath, which is achieved by using a copper surface that stops the laser or by controlling energy delivery tightly. Both approaches impose constraints on the materials and the process.
Filling and plating the via is a second operation with its own requirements. Vias that will be stacked must be filled and planarised so the next layer can be built on a flat surface; vias that will not be stacked can be left open, which is cheaper. The decision to stack vias is therefore not only an electrical choice, it is a cost and yield choice, and design decisions about via placement propagate directly into price.
Testing adds further cost, and it belongs to the same quality management system as fabrication. Microvias are difficult to inspect optically because they are small and often filled, and electrically they can only be verified through the nets they connect. Factories that build HDI in volume typically invest in dedicated inspection and electrical test capability for this reason.
Supply of build-up materials has also been less predictable than supply of standard laminate. Several advanced dielectric films are produced by a small number of suppliers, and demand from AI and high-speed programmes competes with demand from mobile and automotive work. A stackup that depends on a single build-up material with a long lead time carries a risk that no manufacturing skill can remove.
Materials Add Their Own Premium
Build-up dielectric is not the same as core laminate. It is applied as a resin-coated copper or as a film, and it must flow correctly during lamination, adhere to the copper beneath, and survive laser drilling and plating. The material selection affects how small a via can be produced reliably and how well the layers tolerate thermal cycling.
Impedance requirements push material choice further. HDI boards are frequently used for high-speed designs, which means that the build-up dielectric’s electrical properties matter as much as its mechanical ones. Low-loss build-up materials are more expensive and are available from fewer suppliers, adding both cost and supply risk to the quote.
Combining these factors explains why an HDI board can cost several times a conventional multilayer board of the same dimensions. The customer is not paying for material volume, but for process steps, tight tolerances and the yield risk that accompanies them.
A useful test is to count the layers that actually need the freedom. If a design uses build-up layers only to escape one dense component while the rest of the board could be routed conventionally, a smaller device footprint or a different escape pattern may recover the density at lower cost. That analysis is best done while the layout can still change.
When 3+N+3 Is Genuinely Necessary
The first question a buyer should ask is whether the routing density actually requires three build-up layers. Common reasons for needing them include a very high ball count package with fine pitch, a large number of high-speed nets that must escape a dense area, or a requirement for controlled impedance on multiple layers within a small outline.
Cases where the structure is chosen without justification include designs that inherited a stackup from an earlier product, or that added build-up layers because the layout was congested rather than because the routing demanded it. In those situations, a redesign that relocates components or reorganises the escape strategy can allow a simpler stackup and a materially lower price.
This is a conversation worth having with the manufacturer before the design is frozen. A fabricator who understands both the electrical requirement and the process cost can often suggest a two plus N plus two structure, or a hybrid approach using buried vias in the core, that meets the specification at a lower cost and with better yield.
Deliverable definition matters too. Specifications that leave microvia appearance, resin recession or solder mask registration vaguely described will produce disputes when volume production begins, because the acceptable band becomes a matter of interpretation. Defining those criteria before the first build is cheaper than negotiating them during the third.
How to Evaluate an HDI Quote
Comparing HDI quotations on price alone is unreliable, because the quote is a statement about yield as much as about cost. The useful information includes the microvia diameter and aspect ratio the factory can produce reliably, the registration capability it holds across the build, and the inspection and test methods it applies to each layer.
Equally important is the definition of acceptance. Impedance tolerance, via reliability under thermal cycling and the permitted level of cosmetic variation around microvias should all be specified, because those criteria determine how much material is scrapped and therefore how the price behaves in volume.
A supplier who can provide production yield data for a comparable stackup, rather than a capability statement, is offering a much stronger basis for a sourcing decision. That data also makes it possible to identify which aspects of the design are driving rejections, and to correct them before volume production rather than during it. Working with a manufacturer who supports that analysis from the capability discussion stage onwards is what turns an HDI decision into a predictable programme.
Cost, Density and Reliability Trade Together
The practical conclusion is that HDI pricing is not arbitrary and is not primarily a negotiation outcome. It reflects measurable process content: the number of laser operations, the number of lamination cycles, the registration accuracy required, the material performance specified and the inspection depth applied.
Buyers who understand that structure can ask better questions and make better trade-offs. They can choose to relax a via stacking requirement, or to accept a slightly larger board, in exchange for a simpler stackup and better yield. Those choices are often worth more than a price concession, and they are only available to customers who engage with the manufacturing reality rather than treating the stackup as fixed.
Choosing where to spend on density and where to accept a constraint is the central engineering decision in an HDI programme, and it is best made before the layout is finalised rather than after quotations arrive. Engaging a fabrication partner during the design phase, with process planning discussions in parallel, is what makes that trade available.
