AIC to Showcase AI Storage Platforms for Scalable Inference at NVIDIA GTC 2026

SAN JOSE, Calif., March 16, 2026 — AIC, a global leader in enterprise storage and server solutions, will exhibit at NVIDIA GTC 2026, taking place March 16-19 at the San Jose McEnery Convention Center. At booth #140, AIC will present its latest AI storage platforms designed to support CMX-aligned architectures for large-scale inference, enabling shared NVMe tiers that extend GPU memory for long-context and agent-based workloads.

AI inference requires high-density, low-latency storage to support KV cache, vector databases, and continuous data streams. AIC’s NVMe and DPU-enabled platforms provide the flash capacity and bandwidth needed to keep GPUs supplied with data across rack-scale deployments. In parallel, AIC is showcasing system designs that support GPU-initiated storage access models, where data is delivered directly from NVMe to accelerated compute to reduce CPU overhead and improve I/O efficiency for data-intensive workloads.

Featured Demonstrations at AIC’s Booth #140:

• F2026-G5 JBOF with BlueField integration
• A high-density, DPU-accelerated NVMe platform designed to support CMX-aligned shared flash tiers, enabling scalable NVMe-oF architectures for large-scale AI inference deployments.
• CMX, 2U 24-bay Bluefield DPU Solution. 
• A compact, high-performance NVMe platform populated with Solidigm PCIe Gen5 enterprise NVMe SSDs and powered by the NVIDIA BlueField DPU to accelerate KV cache access, enabling higher-performance and more efficient AI inference. The platform is optimized for high-capacity flash configurations that support KV cache expansion, vector database workloads, and CMX-aligned inference environments.
• 3U SCADA-Optimized System
• A 3U system configured to support GPU-driven storage access models, enabling accelerated data delivery for data-intensive AI workloads such as vector search and graph processing.
• VAST Data CERES Platform
• High-throughput NVMe storage supporting AI data pipelines, retrieval-augmented generation, and real-time analytics environments. Utilizing VAST’s Disaggregated Shared Everything (DASE) architecture.

In addition to the 2U 24-bay NVMe Storage System populated with Solidigm enterprise SSDs on display at AIC’s booth, an AIC F2032 platform will be featured in Micron Technology’s booth (#1407) as part of their SSD showcase. These collaborations reflect AIC’s continued work with leading flash vendors to deliver scalable NVMe platforms for next-generation AI infrastructure.

“Designed for the AI era, Solidigm’s eSSDs are uniquely positioned to power the next wave of AI infrastructure by delivering the performance, density, and efficiency required for emerging inference-centric architectures. The AIC F2026 platform with Solidigm eSSDs acts as a foundational context memory tier that helps operators extend GPU cache beyond traditional limits, accelerate throughput, and unlock scalable, long-context AI inference without compromising power or footprint,” said Greg Matson, SVP, Head of Products and Marketing, Solidigm.

“AI infrastructure is evolving rapidly, and storage architecture must evolve with it,” said Michael Liang, President and CEO of AIC. “Working closely with NVIDIA and our ecosystem partners, AIC is developing NVMe platforms aligned with CMX architectures and DPU-accelerated data paths to support emerging workloads such as KV cache expansion and GPU-driven AI pipelines. As AI systems scale toward multi-agent and large-context workloads, scalable and disaggregated storage becomes a critical part of the architecture. Our focus is ensuring the storage systems needed for the next generation of AI are being built today."

Event Details:

NVIDIA GTC

March 16-19, 2026

AIC Booth #140

Attendees are invited to visit AIC’s booth to explore the showcased platforms and speak with AIC representatives about building scalable NVMe infrastructure for modern AI inference environments.