DeepX, a promising player in the artificial intelligence (AI) semiconductor sector, has been confirmed to be participating in the development project of the Internet of Things (IoT) platform 'Greengrass,' which is a next-generation flagship project of Amazon Web Services (AWS), the world's largest cloud company. If the supply of DeepX's neural processing unit (NPU) is successful within AWS's cloud ecosystem built around the world, it is expected to serve as a stepping stone for global market entry.
According to industry sources on the 16th, AWS is strengthening its collaboration with DeepX as a core partner for the Greengrass project, which is being pursued as a strategic business to integrate existing cloud services with AI-based IoT platform services. AWS is considering a plan to incorporate DeepX's NPU into its cloud ecosystem, which is composed of Intel central processing unit (CPU) technology, as a way to accelerate data processing by implementing AI models in edge environments.
Collaboration between AWS and DeepX began in earnest in September of last year, and the partnership has advanced rapidly since then. There are plans to continue collaboration aimed at integrating DeepX's NPU into the AWS cloud in Silicon Valley in the United States. On the 7th (local time), at the Consumer Electronics Show (CES) 2025, the world's largest IT exhibition held in Las Vegas, Channa Samynathan, a senior specialist leading the IoT Greengrass project at AWS, visited DeepX's exhibition booth and discussed future cooperation with CEO Kim Nok-won.
Samynathan noted in a written interview that 'Greengrass, a tool for the AI edge era, provides a direct pathway for cloud services to reach the edge' and that 'through collaboration with DeepX, we aim to fully integrate the strengths of DeepX's NPU into AWS's Greengrass to enhance expertise in (on-device AI).'
AWS was the first in the industry to offer cloud computing technology as a service and currently has the largest customer base. However, areas such as on-device AI and IoT, which support real-time data processing at the device level without connection to data centers, remain uncharted territory for AWS.
Over the past few years, AWS has been investing in Greengrass, its edge computing project, to fully consolidate cloud and IoT. Greengrass plays a role in collecting and analyzing data on local devices and ensuring that devices on the local network can communicate securely with one another. It essentially shifts some functions of cloud services to local devices, effectively making edge computing the doorknob that connects cloud and IoT.
From AWS's perspective, as the on-device AI and IoT markets are rapidly growing, supporting edge computing has become essential to maintain its hegemony in the cloud market. AWS has outlined plans to provide messaging, data and status synchronization, security, local resource access, protocol adapters, and machine learning inference functions for devices connected to the cloud in Greengrass. In particular, it is anticipated that the NPU will play the most critical role in the inference needed for real-time data processing.
Meanwhile, DeepX is set to begin mass production of its first NPU product, DX-M1, based on 5-nanometer technology in the first half of this year. This chip, which is being produced through contract manufacturing by Samsung Electronics Foundry Division, has reportedly secured a yield rate of up to 90%. Unlike competitors' NPUs, which are produced in the 10 to 20-nanometer range, DeepX plans to take market lead by manufacturing NPUs using advanced processes.