Time:2026-04-17Reading:558Second
Due to the majority of DRAM capacity being allocated to high profit products such as HBM to support the development of the artificial intelligence industry, the global mobile industry is facing a long-term memory shortage. As a giant in the field of mobile chips, Qualcomm is actively seeking new ways to address this challenge.
Recently, the news about Qualcomm's collaboration with Changxin Storage and Zhaoyi Innovation to develop customized memory and chips has attracted widespread attention. Analyst Guo Mingchi revealed the deep details of this cooperation, revealing Qualcomm's new trends in the layout of end-to-end AI.
It is reported that Qualcomm is collaborating with Zhaoyi Innovation to develop a discrete NPU specifically designed for smartphones. The target customer group of this product is very clear, which is Chinese smartphone brands, aiming to enhance the AI processing capability of flagship phones.
This new NPU is expected to be officially shipped by the end of 2026 or early 2027. In terms of product positioning, it will mainly be applied to high-end flagship models priced above 4000 yuan, becoming a key hardware to enhance the core competitiveness of mobile phones.
In terms of technical specifications, this NPU can provide about 40TOPS of powerful computing power and is equipped with 4GB of customized 3D memory. It is worth noting that this customized memory is manufactured by Changxin Storage and adopts advanced packaging stacking technology.
By applying TSV and hybrid bonding technology, this solution can provide higher memory bandwidth than the current mainstream LPDDR5X. This architecture advantage can effectively solve the data transmission bottleneck in AI computing, making the operation of end side large models more efficient and power-saving.
However, despite technological breakthroughs, the market expectations for the project have been somewhat lowered. Guo Mingchi stated that due to the significant increase in memory prices, the overall cost of NPU has been pushed up, and the expected shipment of cooperative products is lower than originally envisioned.
In addition, the specific application scenarios and business models of end-to-end AI are still in the exploratory stage and have not yet shown an extremely clear profit path. These uncertain factors pose significant challenges to the deep collaboration between Qualcomm and the supply chain in the process of promotion.
