Micron Readies Client Storage for AI

· Source: Big Data & AI News - EE Times · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Internet of Things (IoT) & Connected Devices · Depth: Intermediate, quick

Summary

Micron Technology has introduced the 3610 SSD, which it claims is the industry's first PCIe Gen5 QLC SSD designed for mainstream client computing, including desktops and ultra-thin laptops. This new SSD is built with Micron's G9 NAND and targets the growing demand for local AI capabilities on end-user devices, driven by privacy and data security concerns. The 3610 SSD boasts sequential read speeds up to 11,000 MB/s, sequential write speeds up to 9,300 MB/s, 1.5 million random read IOPS, and 1.6 million random write IOPS, making it "AI-ready." Its performance is enhanced by Adaptive Write Technology (AWT), which dynamically adjusts NAND cell programming, and a DRAM-less architecture combined with HMB and DEVSLP, improving performance per watt by 43% over Gen4 TLC. Micron also noted that QLC technology offers 25% more bits per wafer, addressing current NAND supply constraints.

Key takeaway

For AI Hardware Engineers designing next-generation client devices, the Micron 3610 SSD signals a shift towards high-performance QLC NAND for local AI workloads. You should evaluate PCIe Gen5 QLC solutions for their ability to handle intensive AI model loading (e.g., 20 billion parameters in under three seconds) while maintaining power efficiency and thermal management, crucial for ultra-thin, fan-less designs.

Key insights

Local AI on client devices is driving demand for high-performance, power-efficient storage solutions like PCIe Gen5 QLC SSDs.

Principles

Method

Adaptive Write Technology (AWT) dynamically manages QLC NAND cells to optimize performance by adjusting programming to SLC, TLC, or QLC modes based on workload and drive state.

In practice

Topics

Best for: AI Hardware Engineer, Director of AI/ML, Tech Journalist

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.