Samsung UFS 5.0 hits 10.8GB/s, doubling speeds to accelerate on-device AI in mobile devices

5 Sources

Share

Samsung has developed the industry's first UFS 5.0 storage solution, achieving unprecedented data transfer speeds of 10.8GB/s—more than double its predecessor. The breakthrough addresses the growing demand for faster on-device AI processing as generative AI shifts from cloud to local hardware. With 40% improved power efficiency and a 16.7% smaller form factor, the technology targets flagship smartphones, XR headsets, and AI wearables set for mass production in Q4 2025.

News article

Samsung UFS 5.0 Sets New Benchmark for Mobile Storage

Samsung Electronics has announced the development of Samsung UFS 5.0, the industry's first next-generation storage standard designed specifically to handle the intensive demands of on-device AI applications. The UFS 5.0 storage solution achieves a data transfer speed of up to 10.8 gigabytes per second (GB/s) for sequential read speeds and up to 9.5GB/s for write speeds—performance metrics that represent more than double the capabilities of the previous UFS 4.1 standard

1

. This significant leap in bandwidth addresses a critical bottleneck as generative AI processing rapidly migrates from cloud networks to local hardware, fundamentally transforming storage from a passive data repository into core infrastructure for AI computation.

The UFS 5.0 storage chip integrates the latest embedded memory interface standard from JEDEC and is built on Samsung's ninth-generation vertical NAND technology

5

. According to Jangseok Choi, head of Memory Product Planning at Samsung Electronics, "In the era of on-device AI, storage devices are evolving into a key driver defining AI experiences"

4

. The enhanced performance is expected to deliver significantly reduced latency and faster response times when running large language models in on-device AI environments, making AI model processing feel noticeably snappier for users

2

.

Improved Power Efficiency Extends Battery Life

Beyond raw speed, the UFS 5.0 memory solution delivers improved power efficiency of more than 40% compared to Samsung's UFS 4.1 solution

1

. Samsung achieved this through implementing clock gating and multi-voltage technologies—innovations that considerably reduce the power required to transfer the same amount of data. Clock gating works by turning off signals to idle parts of the chip, while multi-voltage technology lowers energy consumption and heat by supplying only the necessary power to each circuit

5

. These enhancements drastically lower overall power consumption and extend battery life in next-generation mobile devices, a crucial advantage as faster on-device AI tasks demand more energy throughout the day

2

.

Compact Design Enables Broader Device Applications

Samsung has engineered the UFS 5.0 into an ultra-compact package measuring just 7.5mm x 13mm x 0.9mm, making it 16.7% smaller than its predecessor

1

. This reduced physical footprint significantly boosts design flexibility and internal space utilization for device manufacturers across a wide range of applications, including flagship smartphones, wearable devices, and XR headsets

4

. The smaller form factor becomes particularly important as manufacturers pack increasingly sophisticated AI capabilities into compact form factors.

Mass Production and Industry Adoption Timeline

Samsung will begin mass production of its UFS 5.0 in the fourth quarter of this year, offering capacities up to 1TB

1

. Industry sources indicate that Qualcomm Snapdragon 8 Elite Gen 6 Pro and Snapdragon 8 Elite Gen 6 chips—identified by codenames SM8975 and SM8950—will support the new storage standard

3

. MediaTek's Dimensity 9600 Pro and Samsung's in-house Exynos 2700 are also likely to incorporate UFS 5.0 support

3

.

The development signals a broader industry shift as next-generation on-device AI applications require substantially more local data processing capacity. As generative AI becomes more capable of running directly on smartphones and other devices without pinging servers, the volume of data that must be processed locally continues to surge

2

. Samsung plans to scale up supply to meet the growth of next-generation device markets, expanding beyond flagship smartphones to XR headsets and AI wearables

1

. With this bandwidth unlocking much faster processing of larger context windows, the technology allows on-device AI models to truly mature

3

, positioning storage as a defining factor in how users experience AI on their everyday gadgets.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved