
Kioxia has started shipping samples of its KIOXIA NX1 series, a next-generation data center SSD that combines PCIe 5.0 performance with direct liquid cooling support—the company's first product to offer that cooling capability. The NX1 delivers sequential write performance up to around 38% faster than the prior XD8 series and random write performance up to around 20% faster, enabling better thermal management and efficiency for large AI servers in hyperscale data centers.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Kioxia has developed the KIOXIA NX1 series, a PCIe 5.0 NVMe SSD in E1.S form factor for data centers, and has begun sample shipments to some hyperscale customers. The product is the company's first to support direct liquid cooling (DLC) and uses a next-generation custom controller.
Why it matters
The NX1 series delivers higher performance and power efficiency for GPU-equipped servers and hyperscale data centers. Compared with the prior XD8 series, sequential write performance is up to around 38% faster and random write performance is up to around 20% faster — improvements that help cloud service providers and hyperscale operators improve operational efficiency and infrastructure performance as AI infrastructure scales.
What to watch
The product will be displayed at FMS: The Future of Memory and Storage, held in Santa Clara, California from August 4–6, 2026.
Kioxia Corporation has developed the KIOXIA NX1 series, a new family of NVMe SSDs designed specifically for high-performance data center environments. The product uses the E1.S form factor and supports the PCIe 5.0 interface standard. Notably, the NX1 is Kioxia's first SSD to support direct liquid cooling (DLC), a feature that addresses thermal challenges in densely packed, high-performance server environments. The product incorporates a next-generation custom controller and is engineered to deliver both high performance and power efficiency in GPU-equipped servers and hyperscale data centers. The company has already begun shipping samples to selected hyperscale customers. The NX1 succeeds Kioxia's XD series and is positioned to unlock improvements in operational efficiency and infrastructure performance for cloud service providers and hyperscale operators. When compared with the prior XD8 series, the NX1 delivers significantly faster write speeds: sequential write performance improves by up to around 38%, and random write performance improves by up to around 20%. These gains reflect the benefits of leveraging PCIe 5.0's higher bandwidth. The addition of DLC support is particularly relevant given current industry trends. As AI infrastructure expands in scale and complexity, storage systems must now support high-density, high-performance AI servers. The NX1's coolplate-based direct liquid cooling configuration allows efficient heat dissipation in these demanding environments, preventing thermal bottlenecks that could otherwise limit performance or reliability. Kioxia plans to exhibit the KIOXIA NX1 series at FMS: The Future of Memory and Storage, a major industry conference scheduled for August 4–6, 2026 in Santa Clara, California.
Kioxia's introduction of the NX1 series reflects the industry's shift toward higher-density, higher-power AI infrastructure. The company positions this product as a successor to its XD series, leveraging PCIe 5.0 to address the performance needs of cloud service providers and hyperscale data center operators. The addition of direct liquid cooling support is significant because it signals recognition that modern high-performance servers—particularly those equipped with GPUs—generate heat that traditional passive cooling or air cooling cannot adequately manage. By offering DLC integration from the storage layer, Kioxia enables more efficient thermal designs across the entire system. The performance gains (38% improvement in sequential write, 20% in random write versus the XD8) are substantial enough to materially reduce latency-sensitive workloads and improve overall throughput in large-scale deployments, directly supporting the operational efficiency gains that hyperscalers seek.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
No comments yet. Be the first to share your thoughts!
Log in to join the discussion





Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime