Kioxia and Dell cram 10 PB into slim 2RU server
Summary
Kioxia and Dell have collaborated to create a 2RU server with nearly 10 PB of all-flash storage using 40 Kioxia LC9 245.76 TB SSDs, enabling high-density AI infrastructure.
View Cached Full Text
Cached at: 05/16/26, 06:41 PM
Similar Articles
SK hynix starts mass production of 192GB SOCAMM2 for NVIDIA AI servers
SK hynix has begun mass production of 192GB SOCAMM2 memory modules optimized for NVIDIA AI servers, offering more than double the bandwidth and 75% better power efficiency compared to traditional RDIMM, addressing memory bandwidth constraints in AI training workloads.
@jeffboudier: "We give you model choice, without infrastructure chaos" — @MichaelDell, live from #DellTechWorld Kimi K2.6, DeepSeek V…
Dell and Hugging Face announce that multiple AI models including Kimi K2.6, DeepSeek V4 Pro, GLM 5.1, MiniMax M2.7, and DeepSeek V4 Flash are now available through Dell Enterprise Hub, optimized for PowerEdge XE9780 with NVIDIA B300, simplifying model choice and infrastructure.
Meta fights soaring hardware costs by reusing old DDR4 server memory in new DDR5-only servers — custom CXL 2.0 chip marries legacy DDR4-2400 with cutting-edge DDR5-6400
Meta is reusing legacy DDR4 server memory in new DDR5-only servers by developing a custom CXL 2.0 chip that bridges the two memory types, reducing hardware costs.
@kyzoroX: Chinese factories are buying RTX 5090s in bulk, desoldering the chips, and rebuilding them as 128GB server cards. Rough…
Chinese factories are repurposing consumer RTX 5090 GPUs into 128GB server cards as a workaround to export controls on dedicated AI chips, creating a grey-market competitor to official enterprise offerings.
Are you ready for Le Chaton FAT or still wasting money on GPUs?
The author shares their storage server build optimized for local AI inference, anticipating a rumored 26T-a3b model called "Le Chaton FAT" and using high-capacity NVMe drives with ZFS for model storage.