Oppo said that it was the first smartphone firm to deploy the what is known as Mixture of Experts (MoE) AI architecture in a smartphone. This step reportedly improves processing efficiency and establishes the groundwork for future technological and mobile hardware integration.

According to the Chinese manufacturer, device-based MoE is a "crucial breakthrough" as it reduces compute and data transport usage and greatly improves processing efficiency by activating specialized sub-models to execute particular tasks.

According to laboratory experiments, on-device MoE architecture speeds up AI operations by about 40%, which lowers resource requirements and boosts energy efficiency.

As numerous operations are managed "locally" on the device, Oppo asserted, this contributes to improved privacy, extended battery life, as well as faster AI replies.

In order to overcome the difficulties of utilizing enormous AI models that require a significant amount of processing power and affect performance, particularly on devices that have limited hardware capabilities, the firm partnered with top chipset vendors to achieve the task.

According to Oppo, MoE speeds up the industry's adoption of AI by reducing the processing costs of AI, enabling more devices—from flagship to low-cost devices—to carry out sophisticated AI activities.

Oppo pointed to over 5,860 patent filings in the industry to illustrate its continued commitment to improving AI technology and broadening its accessibility for consumers.

In reality, the competition amongst smartphone makers to introduce competitive on-device AI functions is escalating.

Earlier this year, notable Oppo rival Honor made a strong push in the field, and chipmaker Qualcomm has also shown why on-device AI is better than cloud-based solutions.