Post by eCloud Technology Co., Ltd.

Mixed stackup used to be a BOM cost trick. In 2026, it's how your design survives the next CCL shortage. Premium low-loss laminate on the high-speed layers, standard FR-4 elsewhere. The old reason was BOM cost. That still works. The bigger reason now is structural, and we keep seeing RD teams stay on the old logic. M7/M8 grade CCL is quoting 16+ weeks. Low-Df copper foil is on regional quota. Spread-weave glass is constrained. AI server demand has reshaped the material supply chain inside 24 months. Here's where mixed stackup quietly wins: → When only 2–4 layers use the constrained material, the search space for a drop-in equivalent shrinks fast. Two Dk-matched alternates with similar press behavior is realistic. → When all 12 layers run Megtron-grade, you're hunting for a full-stack substitute. That hunt has no answer in a shortage market. → A second AVL on the high-speed layers only is a 2-week SI re-verification. A full-stack swap is a re-spin. The rule we design toward now: fewer layers tied to the scarce material means a smaller replacement-search surface. The new test isn't only "does this hit the insertion loss budget?" It's also "if this CCL goes to 20 weeks next quarter, how many layers do I have to re-qualify?" If the answer is more than three, the stackup is over-committed. How is your team handling CCL grade selection when the AVL keeps shrinking? Genuinely interested in what's working right now. #PCBMaterials #CCL #Megtron #DFM #PCBDesign