The article discusses how a new storage tier can enhance long context inference for AI agents while significantly reducing cache storage costs. It also highlights the specifics of the storage system, potential market leaders, and associated risks.
The article discusses advancements in semiconductor technology showcased at IEDM 2025, including interconnects beyond copper, the introduction of 1,000 CFETs, and next-generation NAND from SK Hynix. It also highlights the potential of 2D materials in future applications.
The article discusses the critical role of cryogenic etching processes in the development of next-generation 3D NAND technology. It emphasizes the importance of performance and repeatability in these processes for advancing NAND capabilities.
Greenliant is sampling its high endurance NVMe NANDrive™ EX Series SSDs for various industrial, aerospace, and mission-critical applications. These SSDs utilize Greenliant's EnduroSLC™ Technology, highlighting their suitability for demanding environments.
The article discusses several key topics in the semiconductor industry, including the development of an open NAND interface, advancements in fab digital twins, and improvements in data center efficiency. Additionally, it touches on emerging trends in image processing and the upcoming 6G technology.
The article discusses the challenges posed by deep vertical holes and re-entrant features in the production of next-generation 3D NAND. It highlights the limitations of current metrology methods in addressing these complexities.
The article introduces KVNAND, a novel architecture that enables large language model inference on edge devices without the need for DRAM by utilizing compute-enabled 3D NAND flash for both model weights and key-value cache. This innovation addresses the challenges of low arithmetic intensity and high bandwidth demands associated with traditional methods.
November NAND average selling prices (ASPs) surged by 20% to over 60% across all product categories and densities, according to TrendForce. The 1Tb TLC saw the most significant shortage, indicating strong demand pressures in the market.
The article discusses recent advancements in semiconductor research, highlighting innovations such as hydrogel NAND gates, long-distance remote epitaxy, and PAM-8 receivers. These developments suggest ongoing efforts to enhance semiconductor performance and capabilities.
Large-capacity microSD cards are experiencing significant demand in Japan due to a memory shortage that is also affecting high-capacity hard drives. Consumers are increasingly purchasing microSD cards ranging from 512 GB to 2 TB, driven by the advancements in the microSD Express standard.
The price of 4TB SSDs has significantly decreased, with some 8TB models also becoming more affordable. Retailers are actively promoting these deals during the Black Friday season.
The WD Elements 14TB external hard drive is currently priced at $169.99, marking a 55% discount from its regular price, making it an attractive option for consumers this Black Friday.
NAND flash prices are expected to increase by double digits in Q1 2026 as manufacturers reportedly raise production turns. This price hike indicates a shift in the market dynamics for NAND flash memory.
The article discusses solutions to the endurance, error correction code (ECC), and performance challenges faced by QLC NAND flash SSDs in hyperscale environments. It highlights the integration of a PCIe Gen5 controller with hardware-accelerated LDPC and PerformaShape QoS as a promising approach.
The article discusses the increasing importance of NAND flash memory in AI datacenters and the advancements in QLC SSD technology. It also highlights the challenges facing the industry and identifies key market players shaping this landscape.
The article examines the flow of data within AI infrastructure, highlighting the transition from HBM to SSD/HDD. It also discusses the workings of NAND flash and the challenges posed by high-capacity QLC NAND flash SSDs for AI applications.
The article discusses how the growing demand for artificial intelligence is significantly impacting the memory market, particularly for high-bandwidth memory (HBM), DRAM, NAND, and DDR4. This surge in demand is leading to a transformative period in the memory sector, often referred to as a supercycle.
The article discusses the potential of High Bandwidth Flash (HBF) as a new memory tier for artificial intelligence workloads, highlighting its limitations compared to faster memory types like HBM and GDDR. It also explores various use cases for this high-capacity memory despite its slower performance.