Overview

DMCN is a Decentralized Mobile Computing Network, a DePIN (decentralized physical infrastructure network) designed to compute for the field of AI.

Our mission: is to connect every mobile devices, make computing power never idle. Join DMCN, Connect to Compute, Connect to Earn.

The foundation of DMCN is rooted in a simple idea: We have already witnessed a time where the demand for computing power soars (thx to AIGC) does not meet its supply, while a great amount of idling computing power remain unexploited in our everyday lives. A natural resonance comes as can we somehow connect them, like a lock and its only key. Thus, a new world has opened before us. AIGC has defined the computing power of the new era, with Nvidia's silicon deluge being unstoppable; Blockchain technology is in its ascendant phase, its triangle of efficiency, security, and decentralization as solid as a fortress. And the yet-to-be-tapped idle computing power in billions of mobile devices around the globe, will be DMCN's gift to the new world.

Contemporary machine learning models often utilize parallel and distributed computing techniques. It is essential to utilize the capabilities of multiple cores across various systems to enhance performance or scale up to larger datasets and models. The training and inference phases are not merely tasks executed on a single device but typically require a collaborative network of GPUs that operate in conjunction to achieve optimal results.

Nevertheless, traditional cloud service providers offer only about 40% of the capacity required to meet the demand from AI/ML companies, leading to significant challenges in accessing distributed computing resources. Key issues include:

Scarce Resources: Securing hardware through traditional cloud services such as AWS or Azure can take weeks, and sought-after GPU models are frequently out of stock.

Steep Costs: Acquiring high-quality GPUs can be prohibitively expensive, with projects often incurring monthly expenses in the hundreds of thousands of dollars for training and inferencing tasks.

DMCN addresses this issue by harnessing the substantial idle computing power available in mobile devices, which are abundant and inexpensive to connect. This computing power is integrated into a Decentralized Physical Infrastructure Network (DePIN), providing engineers with extensive, on-demand computing resources. This system is designed to be accessible, customizable, cost-effective, and straightforward to deploy, ensuring a professional and efficient computing solution.