According to the forecast of ABI research, a global technology market consulting company, the revenue of edge AI chipset market will reach 12.2 billion US dollars, surpassing the cloud AI chipset market. The cloud AI chipset market will reach 11.9 billion US dollars in 2025.
Currently, the cloud is at the heart of AI because most AI training and reasoning workloads occur in public and private clouds, ABI research said in a statement.
Traditionally, centralizing these workloads in the cloud can bring the benefits of flexibility and scalability. However, the industry has witnessed a paradigm shift in AI.
Driven by the need for privacy, network security, and low latency, AI training and reasoning workloads are being performed on gateways, devices, and sensors.
Recent advances in key areas, including connectivity to cloud computing, new AI learning architectures and high-performance computing chipsets, have played a key role in this shift.
As companies begin to look for AI solutions in the areas of image and object recognition, autonomous material processing, predictive maintenance, and human-computer interface of terminal devices, they need to solve problems such as data privacy, power efficiency, low latency and strong device computing performance, edgeai will be the solution. By integrating AI chipsets designed to perform high-speed reasoning and quantify joint or collaborative learning models, edge AI extends task automation and enhancement capabilities to device and sensor levels across departments. By 2025, it will grow and surpass the cloud AI chipset market. , said jet Su, chief analyst at ABI research.
This is a great opportunity for chipset vendors with a variety of product portfolios to explore and build on, rather than relying solely on specific niche markets.
Intel is a good example. In 2019, the chipset group witnessed the strong growth of Mobileye, its ADAS chipset subsidiary, and surpassed NVIDIA as the leading AI supplier.
Intel is expected to continue to put forward higher requirements for its cloud AI chipsets and generate strong demand for its mobile eye, movidius and FPGA product solutions.
In the consumer market, covid-19 destroys the demand for many smart devices, especially smartphones, smart homes and wearable devices, which affects the deployment of AI accelerators for these devices.
Still, ABI research expects the market to rebound in 2022, Su added. It is worth noting that the impact on the chipset supply chain is relatively small as manufacturing plants in Singapore and Taiwan were still operating during the outbreak.
Key link technologies (such as 5G, Wi-Fi6) and automatic solutions (such as self driving cars) suppliers have minimal impact on their product roadmap. They will continue to test and deploy, predicting a rapid rebound in demand for edge AI chipsets after 2022.
Edge AI will use many other emerging technologies to pave the way for new business opportunities in the consumer and business sectors, Su said