AI Chips for Edge Applications 2026-2036: Technology, Market, and Forecasts
The global AI chip market for edge devices is expected to exceed USD 80 billion by 2036, with the largest applications by market size being automotive and AI smartphones. Artificial intelligence (AI) is already showing great transformative potential across a variety of applications, from detecting fraud in high-frequency trading to using generative AI as a huge time-saver for written documentation and creative prompts. While the use of semiconductor chips with neural network architectures (these architectures are particularly good at handling machine learning workloads, and machine learning is an integral part of AI capabilities) is becoming more prevalent within data centers, data centers are at the edge where significant opportunities for AI deployment exist. For end users, the benefits of being able to deliver more functionality to edge devices and, in certain applications, completely outsourcing human time to intelligent systems are significant. AI is already in the flagship smartphones of some of the world’s top designers, and will be introduced in devices ranging from cars to humanoid robots.
IDTechEx has published a market report that provides unique and independent insights into the global edge AI chip technology landscape and the corresponding market. The report includes a comprehensive analysis of players involved in AI chip design for edge devices, as well as a detailed assessment of technological innovations and market trends. Market analysis and forecasts focus on total revenue with detailed forecasts segmented by geography (China, Europe, US, Rest of the World) and applications (Automotive, Humanoid Robots, AI Smartphones, AI Laptops, Edge Sensors for Predictive Maintenance).
This report presents data analysis and insights from leading companies and is based on IDTechEx’s expertise in the semiconductor, computing and electronics sectors.
This study provides valuable insights into:
- Businesses that need AI-enabled hardware.
- Companies that design and manufacture AI chips and/or AI-enabled embedded systems.
- Companies that supply components used in AI-enabled embedded systems.
- Companies investing in AI and semiconductor design, manufacturing, and packaging.
- Companies that develop devices that may require AI capabilities.
Computing can be segmented by where in the network the computation occurs: in the cloud or at the edge of the network. This report focuses on specialized chips deployed at the edge for AI and machine learning applications.
Artificial intelligence at the edge
The differences between edge and cloud computing environments are not straightforward because each environment has unique requirements and capabilities. An edge computing environment is one where computations are performed on devices at the edge of the network (and therefore closer to the user), typically the same devices where the data is created. This is in contrast to cloud or datacenter computing, which is located at the center of the network. These edge devices include cars, cameras, laptops, mobile phones, self-driving cars, and more. Computations are performed close to the user, at the edge of the network where the data is located. Therefore, given this definition of edge computing, edge AI is the deployment of AI applications at the edge of the network. The advantage of running AI applications on edge devices is that there is no need to send or receive data between the cloud and edge devices to perform computations. Edge devices running AI algorithms can therefore make quick decisions without requiring internet or cloud connectivity. Given that many edge devices operate in power cells, the AI chips used in such edge devices must consume less power than in data centers to be able to operate effectively in these edge devices. As a result, simpler algorithms are typically introduced that do not require as much power.
Growth of AI at the edge
Despite being predicted to exceed USD 80 billion by 2036, significant growth in the edge AI market over the next decade will be challenging. This is due to the saturation and stop-start nature of certain markets where AI architectures have already been adopted in existing chipsets, requiring rigorous testing before mass deployment. For example, the smartphone market has already begun to become saturated. Smartphones continue to become premium (premium smartphones account for an increasing share of total smartphone sales each year), and as premium smartphone sales increase, so do AI revenues. These smartphones incorporate AI co-processing into their chipsets, which itself is expected to start to become saturated over the next decade.
IDTechEx predicts that consumer electronics (AI smartphones and AI PCs) and automobiles (self-driving and intelligent cockpit features) will be the largest markets for edge AI chips. Source: IDTechEx report “AI Chips for Edge Applications 2026-2036: Technologies, Markets, and Forecasts.”
Edge AI for cars
Higher automation, defined by the Society of Automotive Engineers (SAE) from Level 0 (no automation) to 5 (full automation), is a megatrend in the automotive sector. Robotaxis continue to expand into new cities around the world, and private cars with self-driving capabilities will also increase. In 2026, the transition from SAE Level 2+ (Hands Off, Eyes Off) to Level 3 (Conditional Eyes Off) will shift responsibility for the vehicle from the driver to the OEM in some scenarios. Therefore, these vehicles will have enhanced edge AI capabilities to ensure reliable, consistent, and safe operation. Otherwise, OEMs may face legal issues. Additionally, intelligent cockpits will require more AI computing that can be integrated into separate chips or combined with autonomous driving and ADAS (Advanced Driver Assistance Systems) on a single chip.
Edge AI for humanoids
As of 2026, humanoid robots are starting to gain even more attention, especially in the automotive manufacturing industry, where we are beginning to see their expansion and implementation. While adoption will begin in the automotive industry, IDTechEx expects to see adoption in more open and challenging environments such as patrol, surveillance, and homes over the next decade.
In parallel with the growth of the overall humanoid robot market, IDTechEx predicts that the AI computing required per robot will increase significantly as manufacturing sites assign more difficult tasks away from the typical picking, placing, and other logistics tasks that current humanoid robots are deployed with.
Edge AI for consumer electronics
As of January 2026, all major smartphone OEMs have AI-enabled features in their flagship smartphones, from photo generative editing to personalized content creation. IDTechEx predicts that smartphone AI chips will dominate the overall edge AI chip market, with AI chips becoming standard in flagship phones and even more common in mid-range phones. As the cost of cutting-edge hardware on the smallest process nodes continues to rise, mid-range phones will gradually eat into the market share of lower-end phones as manufacturers push higher-range phones to maintain profits.
IDTechEx defines an AI PC as a PC that has a dedicated AI chip as part of its system-on-chip (SoC) and has a performance of more than 40 TOPS (tera operations per second). In 2025, this was an emerging market, with less than 10% of new PC sales fitting this definition. IDTechEx predicts that the majority of new PC sales will be AI PCs by the early 2030s, as major manufacturers such as Lenovo and Apple will capture a larger share of AI PC sales.
Edge AI for sensors
Edge AI for predictive maintenance is quickly becoming a hot topic for startups as well as large sensor suppliers such as Bosch. By running machine learning methods locally on sensors, systems can predict maintenance and repairs before they are actually needed, resulting in significant uptime increases and potential cost savings. The AI computing required is typically much lower than, for example, for self-driving cars or AI PCs, so they are cheaper. As the number of smart factories is expected to increase over the next decade, more MEMS and IMU sensors will incorporate AI capabilities at the edge.
