The electronic component shortage did not end — it evolved. In 2026, the crisis is quieter but more strategic. Certain components are extremely difficult to source, prices are climbing sharply, and the companies caught off guard are the ones that assumed the worst was behind them.
This article breaks down which components are constrained right now, what’s driving the pressure, which industries are most exposed, and what engineers and procurement teams can do to reduce risk.
2026 Is Not 2021 — But the Shortage Is Real
The 2020–2022 shortage was broad. Almost every component category was constrained at the same time. Buyers were competing for everything from resistors to microcontrollers to logic ICs.
The 2026 shortage is different. It’s selective and SKU-specific. Many standard passives and commodity components are relatively available. But specific product families — automotive-grade MCUs, SiC power devices, high-bandwidth memory — are severely constrained.
Think of it this way. The 2021 shortage was the entire highway gridlocked. The 2026 shortage is one or two critical lanes closed. Most traffic moves, but anyone who depends on those lanes is still stuck.
This distinction matters because companies that assume the market has normalized may stop monitoring lead times. They may stop building buffer stock or qualifying alternative parts. When a constraint hits their specific BOM, they have no protection in place.
Which Components Are Hardest to Source Right Now
MCUs and Processors
Automotive-grade 32-bit MCUs with AEC-Q100 qualification are sitting at lead times of 52 weeks or longer. That’s over a year from order to delivery.
Legacy 8-bit and 16-bit MCU families used in industrial equipment are also at risk. As foundries deprioritize older process nodes, these parts become harder to get — and harder to replace without a redesign.
Power Devices
Silicon carbide (SiC) MOSFETs and modules are among the tightest categories, especially for EV applications. IGBTs, power management ICs (PMICs), and high-voltage analog ICs are also constrained. Lead times for some FPGAs and automotive PMICs range from 30 to 55 weeks or beyond.
Major vendors have responded with significant price increases. Texas Instruments announced increases of 15–85% on certain product lines. Analog Devices raised prices 15–30%. Infineon applied increases across power switches and related ICs.
Memory
DRAM, NAND, DDR5, and high-bandwidth memory (HBM) are in a shortage that analysts describe as unprecedented. Supply growth for DRAM is running at roughly 16% year-over-year and NAND at about 17% — but demand from AI data centers is outpacing both figures significantly.
Prices for HBM and DDR5 are rising sharply. Limited supply is flowing first to hyperscalers and AI infrastructure builders, leaving less for consumer electronics and industrial applications.
Passives and Connectors
Specialty MLCCs for AI servers and EVs are tightening. Automotive-grade resistors, capacitors, and high-speed connectors are also under pressure. Standard commodity MLCCs are more stable than during the peak crisis, but they are not immune to price increases driven by raw material costs.
Copper, gold, silver, and tin prices are at or near record highs, raising the cost base for PCBs and passive components across the board.
Sensors
Automotive and industrial sensors tied to ADAS systems and factory automation remain at risk. These parts often have long qualification cycles, which makes substitution difficult when supply tightens.
What Is Driving These Shortages
AI Infrastructure Build-Out
Hyperscalers are building AI data centers at a pace that is creating a concentrated demand shock across multiple component categories. GPUs get the headlines, but the ripple effect reaches HBM, DDR5, power management ICs, specialty MLCCs, and high-speed interconnects.
When AI companies lock up supply through long-term premium contracts, that capacity is no longer available to other buyers. Industries that share components with AI infrastructure — even loosely — feel the squeeze.
Automotive Electrification and ADAS
EVs require SiC power devices, automotive-grade MCUs, sensors, and specialty passives in large quantities. ADAS systems add further demand for sensors and processors. Both trends were already stressing supply chains before AI demand amplified the pressure.
Consider a practical example: an EV manufacturer designs its onboard charger around a specific SiC MOSFET module. When global SiC supply tightens due to competing demand from AI systems and other EV programs, lead times push past 52 weeks. Without a pre-qualified alternative, the manufacturer must either delay production or undertake a costly redesign and recertification.
Industrial Automation, 5G, and Satellite Communications
Industrial automation, robotics, 5G infrastructure, and satellite communications all add demand for MCUs, analog ICs, and RF components. These aren’t niche markets — they represent a large base of ongoing component consumption that compounds pressure on already-strained supply chains.
Supply-Side Structural Limits
Semiconductor fabs take three to five years to build. Capacity expansion always lags demand, which means imbalances persist far longer than buyers expect.
Geopolitical risk adds another layer. US-China tensions, Taiwan risk, and specific export controls have restricted supply from certain regional producers, particularly for wafers and advanced nodes. TSMC and a small number of large foundries produce a very high share of the world’s advanced chips. Any policy change or disruption at that level has global consequences.
Which Industries Are Most Exposed
Automotive teams are managing some of the longest lead times in the market, especially for AEC-Q100-qualified MCUs, SiC power modules, and automotive-grade sensors. Delays in sourcing these parts can hold up EV and ADAS production lines entirely.
Consumer electronics companies are facing rising BOM costs as memory prices climb. A smartphone OEM planning a mid-range model with 8 GB DRAM and 256 GB NAND may find that memory prices have risen enough to require higher retail prices, reduced memory configurations, or a delayed launch.
Industrial and embedded OEMs face a specific vulnerability: legacy MCUs and analog parts used in long-lifecycle equipment. When a foundry moves a part to NRND (Not Recommended for New Designs) and lead times jump past 40 weeks, an OEM with no qualified alternative faces either a production halt or an expensive redesign that includes firmware changes, EMC re-testing, and customer requalification.
AI and datacenter builders are, in some ways, both the cause and the most insulated group. Large hyperscalers sign long-term supply agreements at premium prices, which gives them priority access but further crowds out other buyers.
How Long Will This Last
Some constrained categories may ease by late 2026 or 2027 as capacity investments come online. But certain segments are projected to remain tight well beyond that.
Automotive SiC devices, high-voltage analog ICs, and defense-grade passives may stay constrained into 2028, according to industry outlook data. Memory shortages tied to AI demand are expected to persist at least through 2027, with analysts seeing no quick relief given the pace of demand growth.
The honest answer is that there is no single end date. Different categories will ease at different times, and new demand drivers may create fresh pressure in categories that appear stable today.
What Engineers and Procurement Teams Can Do Now
Design for Availability
Any component with a lead time over 26 weeks should be treated as a shortage risk in your BOM. That’s the point at which you need an action plan — not when the purchase order comes back rejected.
Where possible, specify two or three sources for critical components during the design phase. Prefer JEDEC- or IPC-standardized parts over exotic or single-sourced options. Document pin-compatible alternatives and pre-qualify them through pilot runs so you can substitute quickly without a full redesign cycle.
Track Product Change Notices (PCNs) and Product Discontinuation Notices (PDNs) using tools such as Octopart, SiliconExpert, or Z2Data. Catching an NRND designation early gives you time to respond; catching it after you’ve placed a large order does not.
Rethink Inventory Planning
Ultra-lean just-in-time purchasing made sense when lead times were short and predictable. It does not make sense for components with 30-to-55-week lead times.
For high-risk parts, move to a model that combines planned inventory with safety buffers. Place purchase orders earlier — covering 9 to 12 months of demand instead of a single quarter. Blanket POs and long-term forecasts give suppliers visibility and often improve your position in their allocation queue.
Maintain strict Approved Vendor Lists (AVLs) with a primary and at least one validated alternate for every high-risk component. Diversifying your freight providers and building in emergency shipping arrangements reduces the risk that a logistics disruption compounds a supply disruption.
Strengthen Communication Across Teams
Shortages rarely surprise companies that have strong communication between engineering, procurement, manufacturing, and their suppliers. Sharing demand forecasts with key suppliers — even rough ones — helps them plan capacity allocation in your favor.
Read Also:
