PulseAugur / Brief
EN
LIVE 15:54:52

Brief

last 24h
[2/2] 221 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. TOOL · 36氪 (36Kr) 中文(ZH) ·

    Samsung Electronics Completes Development of 900-Layer Ultra-High Stacking 3D NAND Flash Prototype, Report Says

    Samsung Electronics has successfully developed a prototype for a 900-layer 3D NAND flash memory, a significant advancement in memory technology. This prototype was achieved by stacking two 450-layer 3D NAND chips using Cell-on-Row (COR) technology, and its functionality has been confirmed. This development is a key step towards higher density and more efficient storage solutions. AI

    IMPACT Enables denser and more efficient storage, crucial for AI model training and deployment.

  2. SIGNIFICANT · Tom's Hardware English(EN) · · [2 sources]

    Nvidia's memory costs soar 485%, latest AI systems now cost $7.8 million to build — memory now comprises 25% of the total cost, Rubin GPUs a mere $50,000 apiece

    Nvidia's next-generation AI systems, particularly those utilizing the Vera Rubin VR200 NVL72 configuration, are projected to cost hyperscalers approximately $7.8 million each. A significant driver of this cost increase is the memory components, which now constitute about 25% of the total system price, amounting to roughly $2 million per unit. This surge in memory expense is attributed to a threefold increase in LPDDR5X memory capacity and the addition of substantial 3D NAND storage, alongside onboard HBM4 memory on the Rubin GPUs. AI

    Nvidia's memory costs soar 485%, latest AI systems now cost $7.8 million to build — memory now comprises 25% of the total cost, Rubin GPUs a mere $50,000 apiece

    IMPACT Confirms escalating hardware costs as a major constraint for AI infrastructure scaling.