Chipset sales continue to break records, but several paradigm shifts will determine who gets the biggest slice of the semiconductor industry pie.
Key Insights
- Semiconductor growth in 2026 is being reshaped by a few structural shifts, not just rising demand. Memory constraints, faster product cycles, new security requirements, and robotics are changing where vendors need to place their bets.
- The memory shortage is becoming a strategic fault line for the industry. Chipmakers that improve memory efficiency and align with high-value AI and data center demand will be in a stronger position than those tied too closely to volume consumer markets.
- Wi-Fi 8 is not set to completely replace Wi-Fi 7 in a clean generational handoff. Vendors need product roadmaps that support coexistence and target the specific use cases where Wi-Fi 8 brings clear value.
- Security is moving deeper into the semiconductor stack. OEMs now expect chipset vendors to deliver integrated hardware, firmware, and software security rather than standalone IP blocks.
- Robotics is emerging as a major semiconductor growth engine, but the market is becoming more ecosystem-driven. Vendors need to align with dominant platforms like NVIDIA while still carving out differentiation in sensing, processing, and power efficiency.
The World Semiconductor Trade Statistics (WSTS) projects global semiconductor sales to reach more than US$1.5 trillion in 2026. Behind this growth is record-shattering demand for new silicon designs that can accommodate Artificial Intelligence (AI) and other compute-hungry workloads across both the consumer and enterprise spaces.
Semiconductors are the lifeblood of the modern technological landscape. From smartphones and connected vehicles to Internet of Things (IoT) devices and data center infrastructure, chipsets act as the brains for virtually every application in use today. To meet this demand, leading chipset suppliers are increasingly designing their products in ways we have never seen before.
ABI Research is embedded deep in the trenches of these paradigm shifts. Our analysts have recently attended marquee technology conferences in 2026 to meet directly with the world’s leading semiconductor firms. Through the first quarter of the year, our team has synthesized deep insights from the Consumer Electronics Show (CES), Mobile World Congress (MWC) Barcelona, embedded world, and NVIDIA GTC. In tandem with conversations with semiconductors, the team has identified several transformative trends that will redefine what it means to be an innovative foundry, fabless chipmaker, or software provider.
This analyst evaluation provides a glimpse into the latest trends reshaping the semiconductor industry. We also share strategic next steps for industry participants to meet these new challenges.
Table of Contents:
The AI Data Center Boom Is Causing Semiconductor Memory Shortages
Semiconductor memory shortages were on everyone’s mind at MWC26 Barcelona. One Chief Technology Officer (CTO) at the show reported seeing up to a 26X price increase in DDR4 Random-Access Memory (RAM) chips over the past year.
However, the memory crunch came as no surprise to our team.
ABI Research’s CES 2026 attendance in early January had already revealed widespread concerns over an upcoming memory shortage that could cripple semiconductor accessibility. A global pandemic caused the last shortage. This time, the shortage is being caused by an economic concept recognized for centuries: supply and demand.
For memory manufacturers like Samsung Electronics and Micron, the big prize is in enterprise AI and data centers. NVIDIA, Google, and other tech titans have an unwavering appetite for high-bandwidth Dynamic Random-Access Memory (DRAM). These premium products generate higher profit margins for memory suppliers compared to high-volume products. Consequently, standard memory hardware is becoming scarce, and its cost is skyrocketing.
Industry analyst Georgia Cooke sounded the alarm in the ABI Research report Impacts of Memory Shortage on the eSIM Smartphone Market published in January. “Foundry capacity is already very limited, with stable historical shipment figures demonstrating the almost ‘fixed’ nature of memory supply in the absence of massive capacity changes.” She continues, “Most foundries are already operating at very high utilization, and building and readying new foundries takes many years and significant investment, so changes to the technological distribution of demand essentially results in reallocation of capacity.”
The semiconductor industry’s greatest fears were confirmed in the halls of Fira Gran Via at MWC26. Chipmakers are increasingly sidelining Personal Computer (PC) and smartphone manufacturers to prioritize more lucrative clients. Meanwhile, hyperscalers and AI-focused firms are monopolizing the available inventory.
ABI Research contends that the current supply constraints represent a fundamental shift in the industry's “status quo” rather than a short-lived cycle. With this in mind, the semiconductors that sustain commercial success will re-architect their chipsets to be smarter and more memory efficient.
Strategic Priorities for Semiconductor Companies
- Design memory-efficient chipset architectures to reduce dependency on constrained RAM supply.
- Prioritize high-margin AI and data center segments, while managing exposure to consumer markets.
- Invest in advanced packaging and integration (e.g., High-Bandwidth Memory (HBM), chiplets) to maximize performance.
- Establish long-term supply agreements and ecosystem partnerships with hyperscalers.
Wi-Fi 8 Chipsets Will Complement Wi-Fi 7 Rather Than Fully Replace It
Semiconductor giants like MediaTek and Qualcomm drove the Wi-Fi 8 narrative at CES and MWC26. But according to ABI Research, the path to widespread adoption remains uncertain due to looming market challenges.
For starters, most Wi-Fi 8 innovations are found at the Medium Access Control (MAC) and Physical (PHY) layers. Therefore, the underlying hardware requirements are essentially the same as Wi-Fi 7. The differentiation between the two wireless connectivity standards is inherently more limited than in previous standards.
Accelerated innovation cycles mean that new Wi-Fi protocols don’t offer the level of perceived value we have seen in years past. The traditional 5-year cycle between Wi-Fi generations is slowly but surely disappearing. Consequently, there is a high risk that the residential and enterprise markets will simply skip over Wi-Fi 8 completely. Not helping matters is the fact that Wi-Fi 8 relies on expensive DDR4 and DDR5 chips, which are seeing steep price increases due to the memory crisis.
Going forward, chipset manufacturers must closely monitor how Internet Service Providers (ISPs) decide to move forward. Some vendors, notably Vantiva, claim that their Wi-Fi 7 Customer Premises Equipment (CPE) can be remotely upgraded to Wi-Fi 8 later.
Most network equipment vendors launched new Wi-Fi 7 solutions less than 2 years ago. They are not going to go all-in on Wi-Fi 8 before they have squeezed the most Return on Investment (ROI) possible from Wi-Fi 7. As a result, the semiconductor industry should be prepared for Wi-Fi 8 to serve as a complementary technology alongside Wi-Fi 7.
Vendors are more likely to introduce Wi-Fi 8 solutions for specific use cases. For instance, high-density deployments benefit from Wi-Fi 8’s Multi-Access Point Coordination (MAPC) features. Moreover, the standard’s Single Mobility Domain (SMD) capabilities are a solid match for robotics applications. However, the idea that Wi-Fi 8 chipsets will completely replace Wi-Fi 7 anytime soon is at odds with current market realities and hardware cycles.
The speed of Wi-Fi 8’s arrival is threatening the ability for many equipment vendors to secure strong Return on Investment (ROI) on their Wi-Fi 7 products, which, in some cases, were launched less than 2 years ago. In response, many vendors are viewing Wi-Fi 8 not as a replacement for Wi-Fi 7, but as a complementary standard. Whereas in the past, vendors have built a complete range of solutions based off the latest standard for all segments of the market, with Wi-Fi 8, many vendors plan to selectively introduce the technology into their portfolio to target specific use cases for which it is optimized.
Andrew Spivey, Principal Analyst ABI Research
Strategic Priorities for Semiconductor Companies
- Align product roadmaps to support Wi-Fi 7 and Wi-Fi 8 coexistence, not rapid replacement.
- Target high-value, use-case-driven deployments (e.g., robotics, high-density environments).
- Develop upgradable and modular chipset platforms to extend product lifecycles.
- Closely track ISP and Original Equipment Manufacturer (OEM) adoption signals to time investments effectively.
Semiconductor Security Is Transforming from a Feature Into a Core, Platform Requirement
Intellectual Property (IP) security is increasingly being embedded at the chipset level. Government regulations, such as the European Union (EU) Cyber Resilience Act (CRA), are driving this shift beyond critical infrastructure/national defense and into more traditional enterprise verticals.
Consequently, semiconductors and chipmakers are now responsible for delivering foundational security that spans from internal architecture to full system integration. For example, OEMs expect features like Root of Trust (RoT) and cryptography to be native to Microcontroller Units (MCUs) and Systems-on-Chip (SoCs). In this new landscape, non-compliance means losing customers.
OEMs will seek semiconductor partners that bundle hardware, firmware, and software security into one platform. This includes:
- Cryptographic libraries
- Firmware stacks
- Integration tools
- Certification support
Essentially, silicon security is being positioned as a one-stop solution. Forward-looking semiconductor players recognize the need to strengthen their security capabilities to meet evolving customer expectations. Recent examples include Synopsys acquiring Elliptic Technologies and Candence acquiring Secure-IC. Those that remain pure play IP providers will be on the outside looking in.
This trend toward “platformized” semiconductor security will directly influence how security IP solutions are sold. ABI Research’s Security IP: Design Integration & Pricing Strategies report evaluates these changes and profiles industry leaders like Arm, FortifyIQ, Rambus, and Xiphera.
One of the main gaps in the market currently is for customization, especially on the IP software side. OEMs want specific versions of cryptography IPs that need to be ported, tested, and optimized. This is the differentiation opportunity for core IP providers today.
Michela Menting, Vice President ABI Research
Semiconductor security is also increasingly shifting toward quantum-safe architectures. Although Post-Quantum Cryptography (PQC) standards are still pending in some markets, leading chipset manufacturers (e.g., Infineon, NXP, STMicroelectronics) are developing products that are resistant to quantum threats. Or at the very least, they are developing crypto-agile chipsets that offer a clear path toward PQC once customers are ready. ABI Research continues to assess quantum-readiness in lucrative semiconductor markets such as Subscriber Identity Module (SIM) cards, government ID, and payments. Each market is evolving at a different pace, and vendors are approaching PQC integration in unique ways.
Strategic Priorities for Semiconductor Companies
- Transition to platform-level security offerings spanning hardware, firmware, and software.
- Embed native security features (e.g., RoT, cryptography) across all chipsets.
- Invest in PQC readiness and crypto-agility.
- Provide integration, customization, and certification support to meet OEM requirements.
Semiconductors Size up Physical AI, with NVIDIA Dominating the Robotics Development Pipeline
Robotics is the next major growth lever for semiconductors, especially as automotive business wanes. According to ABI Research, revenue from robotics semiconductors will increase from US$1.47 billion in 2026 to US$4.9 billion by 2035. Processing holds the lion’s share of the market (learn more in our Charts & Data).
At CES 2026, industry leaders such as Qualcomm, AMD, and Intel introduced specialized chipsets tailored to Physical AI and advanced robotics. These SoCs enable robots to interact with their surroundings via AI. For example, Qualcomm unveiled its robotics-specific processor Dragonwing IQ10 Series at CES 2026. It enables mobile robots to manipulate their environment and interact with humans. This is powered by Video-Language-Action (VLA) models, Video Language Models (VLMs), and other AI models running within Qualcomm’s robotics platform. The result is a robotics-first chipset that lets humanoids complete tasks as complex as sorting fruit onto plates of the same color.
ABI Research Principal Analyst George Chowdhury noted similar demos and announcements at CES 2026 from Ambarella, AMD, and NVIDIA. He also stated that “Other silicon vendors, including NXP, Microchip, and STMicroelectronics, have a renewed interest in the market and are discovering the best ways to meet the demands of a potential boom.” Chowdhury explains that Big Tech, hyperscalers, and automotive vendors will also want to board the robotics hype train. The commercial opportunities are vast, but scaling Physical AI across the enterprise space will require further innovation cycles.
The most influential company in scaling Physical AI is undoubtedly NVIDIA. The company’s GTC 2026 event made it impossible to ignore. Texas Instruments (TI), NXP, STMicroelectronics, and other legacy semiconductor companies are increasingly dependent on the NVIDIA ecosystem for their robotics-enabling solutions.
Notable integrations announced at GTC 2026 include:
- TI is integrating Millimeter Wave (mmWave) radar, motor-controlled Integrated Circuits (ICs), and power systems directly into Jetson Thor and Holoscan.
- NXP is adopting Holoscan as well for better sensor fusion. Its i.MX and S32-series controllers now slot into NVIDIA-defined architecture.
- STMicroelectronics is testing and virtually commissioning robotics components with NVIDIA’s Isaac Sim.
While NVIDIA is a dominant force within the general semiconductor robotics market, the company now faces major competition from vendors developing SoCs for humanoids. For instance, NXP MPUs are used in roughly 20% of all humanoids we track. And about one in four humanoids in circulation leverage Intel products. NVIDIA is seeing similar pressures in the drone market, driven by companies like Qualcomm and Ambarella having close ties with DJI.
The currently‑fractured robotics hardware ecosystem is being pulled into a vertically-integrated pipeline where NVIDIA owns simulation, training, inference, and now increasing portions of sensor fusion, networking, and motion‑control orchestration. This leaves competitors (semiconductors) less room to differentiate on anything other than component‑level performance.
George Chowdhury, Principal Analyst ABI Research
Strategic Priorities for Semiconductor Companies
- Develop robotics-optimized SoCs with strong AI, sensing, and real-time processing capabilities. ABI Research forecasts processing applications to account for ~45% of global robotics semiconductor revenue through 2035, and 13% to 20% for sensing.
- Align with dominant ecosystems (e.g., NVIDIA) while differentiating at the component level.
- Focus on power efficiency and edge autonomy for scalable robotics deployments.
- Build partnerships with robotics OEMs, hyperscalers, and system integrators.
Preparing for a New Era of Industry Structural Transformation
The race toward a US$1 trillion semiconductor industry is well underway in 2026, driven by exponential growth in AI, data centers, digital security, connectivity, and robotics. Memory shortages are also shaking things up in a big way. As consumer device markets take a back seat in semiconductor allocation, vendors are rethinking their Go-to-Market (GTM) strategies.
Boardroom conversations must increasingly revolve around chipset innovations that:
- Meet new memory allocation realities.
- Bridge the gap between Wi-Fi generations.
- Support updated silicon-level security requirements.
- Differentiate at the robotics component level.
There is a lot to take in right now as a semiconductor vendor. Product teams must prioritize novel designs and features, and marketing teams must craft original narratives. All the while, the C-Suite is tasked with keeping its core initiatives in sync with the erratic rhythm of the industry.
ABI Research is a trusted market intelligence partner for several semiconductor industry leaders. Our approach includes granular forecasting data across various chipset segments, strategic consulting, competitive benchmarking, and content marketing services. Learn more about how semiconductor vendors leverage our insights to meet their expansion goals by reviewing our semiconductor solutions.
Go deeper with our semiconductor insights from notable technology conferences in 2026:
- Key Takeaways from CES 2026: Signals, Gaps, and Opportunities Across Emerging Tech Markets
- MWC Barcelona 2026: Will AI Finally Crack the Connectivity Monetization Code?
- NVIDIA GTC 2026: Extending AI Leadership
Aisling Dawson