Artificial intelligence software giant OpenAI is reportedly preparing for a radical move. After dominating the world of large language models (LLMs), it is now setting its sights on hardware. According to current information,macAccording to renowned analysts, the company plans to launch its own smartphone in 2028, which could act as an "AI-native" phone.
OpenAI apparently decided to follow the strategy Applu and Google. Instead of just developing applications for foreign platforms, they are betting on full integration. The key to success is to development of own chipsets, on which OpenAI is reportedly collaborating with market leaders, companies MediaTek a Qualcomm.
This initiative follows an earlier announcement of a collaboration with the company Broadcom, whose aim was to develop AI chips for data centers. But now the ambition Sam Altman are being moved directly into the pockets of end users. The goal is clear: hardware optimized for specific algorithms, which will allow artificial intelligence to run more efficiently directly on the device.
Architecture focused on efficiency and context
According to the analyst Ming-Chi Kua The priority of these processors will not be raw benchmark performance, but rather the specific disciplines necessary for modern AI. Key design factors include:
- Memory hierarchy management: Efficiently move data between memory and the processor for minilatency reduction.
- Optimization for smaller models: Ability to seamlessly run more efficient language models locally.
- Energy efficiency: MaxSimulation of battery life when constantly loaded with AI functions.
The phone is expected to have advanced features such as "always-on context"The device should continuously perceive the user's surroundings and digital context in order to proactively offer assistanceSimilar technology is already used today by, for example, Qualcomm Sensing Hub, but OpenAI intends to integrate it more deeply into the operating system and user experience.
Gallery
Pixel model as inspiration
OpenAI's strategy strongly resembles Google's approach with its Pixel line. Instead of chasing the highest number of CPU cores, the emphasis is on NPU (Neural Processing Unit) and software functions, which are not available elsewhere. Complex computing operations will continue to be delegated to cloudu, which makes sense given the unmatched performance of OpenAI's server farms. However, the custom chip ensures that essential interactions and private data remain securely and quickly processed right on the phone.
The journey to owning a smartphone is a long one. Final chipset specifications and supply chain confirmation are expected by the turn of the year. 2026 to 2027If all goes according to plan, mass production should begin in 2028.
The Chinese giant is set to be an important player in this ecosystem Luxshare, which should figure as an exclusive partner for system co-design and final assembly. OpenAI is gradually assembling a mosaic of partners that will allow it to enter the highly competitive mobile hardware environment with a product that is unparalleled.
