The Snapdragon X80 5G Modem-RF System is Qualcomm's latest seventh-generation 5G solution.
Qualcomm Snapdragon X80 5g Modem, FastConnect 7900 mobile Connectivity System Unveiled At MWC 2024
The Snapdragon X80 5g Modem-RF System is Qualcomm's latest seventh-generation 5g solution.
The Snapdragon X80 5g Modem-RF System is Qualcomm's latest seventh-generation 5g solution. With a Peak Download Speed of 10Gbps and a Peak Upload Speed of 3.5Gbps, it sets a new standard for high-speed connectivity. Highlighted features of the Snapdragon X80 comprise the incorporation of a specialized 5g AI Processor, compatibility with NB-NTN satellite communications, implementation of a 6-antenna architecture tailored for smartphones, and the introduction of AI-driven mmWave range extension capabilities.
Additionally, it showcases AI advancements to bolster data speeds, reduce latency, enhance the quality of service, expand coverage, improve location accuracy, and optimize power efficiency. Operators are likely to gain from its exceptional performance and adaptability, leveraging features such as 6x carrier aggregation and advanced spectrum utilization.
Devices equipped with the Snapdragon X80 are scheduled for launch in the second half of 2024, suggesting its integration into the Snapdragon 8 Gen 4 SoC.
The San Diego, California-based chip maker also provided a sneak peek of Large language and Vision Assistant (LLaVA), a multimodal model (LMM) boasting over 7 billion parameters, according to a report by business Standard. This advanced model, showcased on an Android smartphone, can process various types of data inputs, including text and images. Qualcomm highlighted that the LMM operates on-device, offering improved privacy, reliability, personalization, and cost-effectiveness.
In addition to showcasing LLaVa, Qualcomm demonstrated Low-Rank Adaptation (LoRA) executing Stable Diffusion on an Android smartphone. LoRA, according to Qualcomm, empowers users to generate high-quality custom images data-aligned with their personal or artistic tastes. As a streamlined iteration of AI models with fewer trainable parameters, LoRA promises efficiency, scalability, and customization for on-device generative AI applications.