➀ KIOXIA AiSAQ is a research project that transitions AI data storage from DRAM to SSDs, enhancing performance by offloading compressed vectors; ➁ It significantly reduces DRAM usage, improving AI workload efficiency; ➂ The technology will be showcased at CES 2025, highlighting its potential impact on large datasets and AI models.

➀ Kioxia demonstrated an optical interface SSD at FMS, developed with funding from NEDO to reduce data center power consumption by up to 40%. ➁ The optical interface allows for data transfer over long distances, enabling storage to be kept in separate rooms with minimal cooling requirements. ➂ The demonstration showed a slight loss in IOPS performance but significant latency advantages over traditional copper network links.

➀ Kioxia demonstrated a new RAID offload methodology for enterprise SSDs at FMS 2024, aiming to address performance scaling issues in RAID arrays. ➁ The proposed solution utilizes PCIe direct memory access and the SSD controller's memory buffer (CMB) to avoid data movement to the CPU. ➂ The PoC showed significant reductions in CPU and system DRAM utilization, and Kioxia plans to contribute this feature to the NVM Express working group for potential standardization.

➀ Kioxia showcased its BiCS 8 3D NAND technology at FMS 2024, featuring 218 layers and a new CMOS Bonded Array (CBA) process. ➁ The CBA process allows independent parallel processing of CMOS wafer and cell array wafer, improving reliability and performance. ➂ Despite fewer layers compared to competitors, Kioxia's lateral scaling enables competitive bit density and operating speeds.

1. Kioxia introduces 1 Tb and 2 Tb 3D QLC NAND devices with a 3600 MT/s interface speed. 2. The new NAND devices feature 238 active layers and CMOS directly Bonded to Array (CBA) design. 3. Kioxia's 2 Tb 3D QLC NAND offers 70% higher write power efficiency and is suitable for AI and large storage applications.