Please select your location and preferred language where available.
Event Report: NVIDIA GTC 2025 - KIOXIA SSD Solutions Empower AI Application
May 30, 2025
In March 2025, Kioxia exhibited at NVIDIA GTC 2025 (GTC 2025), the premier artificial intelligence (AI) and accelerated computing conference, and highlighted the critical role of high-performance storage in AI solutions – including its ability to scale these applications up and out.
The show took over downtown San Jose, California, with the SAP Center hosting the keynote and the McEnery Convention Center buzzing with industry leaders, researchers, and technology enthusiasts. With over 25,000 attendees, the show floor was packed from beginning to end. GTC 2025 was a non-stop exchange of ideas and innovation – and AI, and the solutions that help scale this exciting technology, was at the heart of it all.
Introducing the KIOXIA LC9 Series 122.88 TB NVMe™ Enterprise SSD
At the show Kioxia launched the high capacity KIOXIA LC9 Series 122.88 TB NVMe Enterprise SSD for AI applications which were showcased at our booth running live. From powering vector databases to enabling faster model training and inference, KIOXIA LC9 Series SSDs are designed to support AI workloads at scale. Customers and press alike took notice, engaging with our experts to understand how these SSDs can help unlock new efficiencies in AI infrastructure.
What the Media Are Saying About KIOXIA LC9 Series SSD*
ServeTheHome
“This is a hot segment of the market, and it is great to see Kioxia joining. As AI clusters get larger, the shared storage tier is usually measured in Exabytes. Teams have found that replacing hard drives with SSDs often reduces power, footprint, and TCO compared to running hybrid arrays. Moving from lower-capacity drives to the 122.88TB capacity in a 2.5″ drive form factor really highlights the advantage of flash in these systems.”
StorageReview
“Storage requirements in AI, particularly for training and inference, continue to push the boundaries of available technology. Large-scale models require extensive datasets, and high-density SSDs like the KIOXIA LC9 Series are becoming critical in managing these workloads efficiently. The ability to store vast training data on a single drive reduces complexity and improves access speeds, particularly for AI systems leveraging retrieval-augmented generation (RAG) architectures.”
- These independent third-party media articles were originally published in English. In case of any discrepancies in translation, please refer to the original English version.
Tackling AI’s Growing Data Demands
At Kioxia’s booth there was a live demo of the KIOXIA All-in-Storage ANNS with Product Quantization (KIOXIA AiSAQ) technology – Featuring KIOXIA CD8P Series PCIe® 5.0 NVMe Data Center SSDs, showcasing the performance of storage solutions for AI. Attendees experienced how KIOXIA AiSAQ technology improves data processing efficiency and supports AI workloads.
Kioxia presented at an expo hall theater session to discuss how the demand Generative AI systems place on compute, memory, and storage resources is raising the cost of DRAM – and what Kioxia is doing about it.
In the presentation, we provided insights into new approaches to enhance efficiency in AI workloads and how to improve storage performance while reducing reliance on DRAM.
Additionally, titled “Improve Vector DB Performance While Reducing DRAM Use in AI Systems,” we detailed how Kioxia has developed a method to scale vector databases while reducing DRAM dependency and costs. By leveraging flash memory and open-source software, Kioxia provides a more efficient and cost-effective approach to managing the increasing data demands of AI workloads.
In case you missed it, watch it at NVIDIA On-Demand.
As AI continues to evolve, Kioxia is committed to pioneering storage innovations that empower AI workloads with the speed, efficiency, and scalability they require. You can learn more about all the solutions Kioxia showcased at GTC 2025 from our press release.
- Definition of capacity: Kioxia Corporation defines a megabyte (MB) as 1,000,000 bytes, a gigabyte (GB) as 1,000,000,000 bytes and a terabyte (TB) as 1,000,000,000,000 bytes. A computer operating system, however, reports storage capacity using powers of 2 for the definition of 1GB = 2^30 bytes = 1,073,741,824 bytes and 1TB = 2^40 bytes = 1,099,511,627,776 bytes and therefore shows less storage capacity. Available storage capacity (including examples of various media files) will vary based on file size, formatting, settings, software and operating system, and/or pre-installed software applications, or media content. Actual formatted capacity may vary.
- NVMe is a registered or unregistered mark of NVM Express, Inc. in the United States and other countries.
- PCIe is a registered trademark of PCI-SIG.
- Other company names, product names, and service names may be trademarks of third-party companies.
Share
Data storage and memory technologies are essential for high performance generative AI systems. Kioxia has a long-term commitment to advancing AI through industry contributions of research and development in flash memory, SSDs, and associated technologies.
KIOXIA solid state drive (SSD) portfolio is built with BiCS FLASH™ 3D flash memory, delivering optimized SSD products and solutions for client PCs, enterprise servers and storage, and cloud data centers.
In proportion to the explosion in demand for IoT devices and AI, data centers require the various technologies that can accomplish both higher performance and better energy efficiency at the same time. Kioxia is addressing the development of optical SSD through contributing Japan national project “The technologies development for the next generation green data centers”.