Please select your location and preferred language where available.
Event Report: NVIDIA GTC 2026
KIOXIA Accelerates AI Innovation
June 26, 2026
In March 2026, Kioxia returned to NVIDIA GTC in downtown San Jose, showcasing how our memory and SSD solutions are accelerating AI innovation. GTC 2026 remains one of the most important stages for developers, researchers, and business leaders pushing AI forward. This year, Kioxia put a spotlight on flash storage as a core enabler of scalable, efficient AI infrastructure. High-performance flash storage helps organizations get more from their AI investments - moving data faster, improving access, and shortening the path to insight.
NVIDIA GTC 2026 took over downtown San Jose, with activity across the city. The main event at the San Jose Convention Center and a packed keynote at the SAP Center underscored the momentum behind AI.
Introducing the KIOXIA GP Series SSDs
GTC 2026 also marked the launch of our new KIOXIA GP Series SSDs for AI GPU-initiated workloads. These super high IOPS SSDs deliver high performance and low-latency memory expansion for the NVIDIA Storage-Next™ Architecture.
Designed for AI and high-performance computing, the KIOXIA GP Series expands GPU-accessible memory capacity - helping accelerate data access where it matters most. Evaluation samples are expected to be available to select customers by the end of 2026.
What the Media Are Saying About KIOXIA GP Series SSD*
ServeTheHome
“At NVIDIA GTC 2026, KIOXIA made one of the more interesting storage announcements of the show. The company revealed two new SSD products targeting the rapidly growing AI inference market, and the positioning here is meaningfully different from standard data center NVMe™ fare. These are drives designed explicitly for GPU-initiated access, a use case that will only become more important as AI model complexity outpaces local HBM capacities.”
Tom's Hardware
“The new drive is part of Kioxia and Nvidia's move to bring incredibly fast storage to AI GPUs. The GP series comes with Kioxia's XL-FLASH™, which is designed to achieve over 10 million IOPS, a figure that is around three to four times greater than traditional datacenter SSDs.”
TweakTown
“KIOXIA announces a brand-new type of SSD, and it's a game-changer for AI.”
- 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.
On the GTC 2026 Show Floor: AI-Ready Storage in Action
At the expo the Kioxia booth stayed full – often standing room only. Demos ran continuously, conversations moved fast, and interest in storage’s role in AI was unmistakable. Kioxia’s Innovation Lab team kept things moving, guiding attendees through hands-on demos and real-world use cases. Demonstrations at the booth included:
Super High IOPS SSD Emulator for AI-Ready Storage
A standout demo delivered over 100 million IOPS, giving attendees a look at what ultra-high-performance storage can enable in demanding environments like NVIDIA SCADA™.
High-Capacity QLC Storage for Large Datasets
KIOXIA LC9 Series enterprise NVMe SSDs - up to 245.76 TB - demonstrated fast access to massive datasets, supporting both AI training and inference at scale.
Optimizing Memory Footprint for Vector Databases
KIOXIA AiSAQ™ technology showed how to reduce DRAM requirements without sacrificing performance. Paired with KIOXIA CM9 Series enterprise NVMe SSDs, it offers a more scalable path for managing growing AI datasets.
KIOXIA AiSAQ™ Software Download
KIOXIA AiSAQ™ software is open-source. Please follow the link to download.
GitHub:
Looking Ahead
GTC 2026 made one shift clear: AI is moving from experimentation to production - across data centers, physical AI, and agentic AI systems. Kioxia is committed to delivering memory and SSD solutions that help customers move accelerate their AI innovation, scale smarter, and unlock what’s next.
- Product image may represent a design model.
- 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.
- NVIDIA, Storage-Next and SCADA are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries.
- NVMe is a registered or unregistered mark of NVM Express, Inc. in the United States and other countries.
- Other company names, product names, and service names may be trademarks of third-party companies.