Showcasing High-Efficiency SSDs and 128GB R-DIMM Memory to Help Enterprises Build Flexible AI Data Infrastructure
TAIPEI, TAIWAN – March 26, 2025 – TRUSTA, the enterprise storage brand of memory leader ADATA, is making its debut at the China Flash Market Summit 2026 (CFMS 2026) with a strong presence. Guided by its vision of “Together, Empowering Tomorrow,” TRUSTA is focused on helping enterprise users bring AI into real-world deployment. Centered on the theme of “Flexible On-Prem AI Computing Solutions”, TRUSTA is showcasing its integrated hardware-software portfolio, including PCIe Gen5 and Gen4 enterprise SSDs, 128GB high-capacity DDR5 R-DIMM memory, and the AI Scaler Toolkit software platform. Together, these solutions enable enterprises to accelerate on-prem AI inference and computing deployment with greater architectural flexibility and cost efficiency.
AI Scaler Toolkit: Orchestrating On-Prem AI
As demand for high-end GPU and HBM resources continues to grow amid limited supply and rising costs, TRUSTA AI Scaler Toolkit helps optimize AI inference workloads by offloading portions of large language model (LLM) inference tasks and dynamically balancing workloads across GPUs, DRAM, and SSDs. This enables more efficient resource allocation and improved inference performance.
At the same time, AI Scaler Toolkit creates strong synergy with today’s increasingly popular on-prem AI Agents for automated workflows. While AI Scaler Toolkit manages compute orchestration and resource scheduling, AI Agents enable smoother local automation execution. Together, they allow small and medium-sized enterprises to build a complete on-prem AI environment with better cost efficiency.
Enterprise Storage Solutions Tailored for Diverse Scenarios
At CFMS 2026, TRUSTA is presenting a full lineup of enterprise storage solutions, led by its flagship T7P5 PCIe Gen5 SSD, available in multiple form factors including U.2, E1.S, and E3.S to meet the compatibility needs of various server architectures. Delivering sequential read/write speeds of up to 13,500/10,300 MB/s, the T7P5 also achieves 447 MB/s per watt in transfer efficiency, offering 1.7 times higher power efficiency than comparable products in its class. With support for up to 3 DWPD (Drive Writes Per Day), it ensures the endurance and reliability required for intensive, high-frequency AI workloads, helping enterprise customers accelerate edge AI adoption with lower power consumption and better overall cost efficiency.
In addition, TRUSTA is showcasing mainstream enterprise storage solutions including the TD7P40 PCIe Gen4 SSD and TD5S30 SATA SSD, balancing performance and broad deployment versatility. On the memory side, TRUSTA is also introducing DDR5 R-DIMM modules with capacities of up to 128GB, featuring high capacity, strong stability, and ECC error correction support to meet the demanding requirements of AI computing, virtualization, and data center workloads. With a broad range of specifications and flexible scalability, these solutions are well suited for a variety of small- and medium-sized compute nodes and private deployment environments.
Beyond performance, the rapid growth of AI applications is also driving tighter supply of NAND Flash and memory resources, making supply stability an increasingly critical factor in enterprise AI infrastructure planning. Through deep collaboration with key global supply chain partners, TRUSTA ensures stable supply capabilities, helping enterprises reduce deployment risks and improve operational certainty when building AI servers and data centers.
Exhibition Information
l Event: China Flash Market Summit 2026 (CFMS 2026)
l Date: March 27, 2026
l Venue: JW Marriott Hotel Shenzhen Bao'an
l Booth: 2F Exhibition Area, Booth L11

O-Sense




