{
  "id": 3607574,
  "title": "Nvidia custom 'NVHBM' promises 30% higher bandwidth, 15% lower power than commodity HBM4e — custom base die and PHY will be available to NVLink Fusion partners",
  "url": "https://urgent.news/2026/08/26/nvidia-custom-nvhbm-promises-30-higher-bandwidth-15-lower-power-than",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-08-26T22:45:22.000Z",
  "source": {
    "name": "Tom's Hardware",
    "slug": "tom-s-hardware",
    "url": "https://www.tomshardware.com/pc-components/dram/nvidia-custom-nvhbm-promises-30-percent-higher-bandwidth-15-percent-lower-power-than-commodity-hbm4e-custom-base-die-and-phy-will-be-available-to-nvlink-fusion-partners"
  },
  "original_language": "en",
  "account": "Nvidia's NVLink Fusion program provides partners with components to connect custom chips using NVLink's scaling domain. To enhance this offering, Nvidia introduces NVHBM, a custom high-bandwidth memory implementation designed for AI accelerators. NVHBM promises 30% higher bandwidth and 15% lower power consumption than standard HBM4e, offering custom silicon developers quicker time-to-market through a smaller footprint and optimized design. Instead of replacing HBM, NVHBM is a new building block exclusively available to Nvidia's custom silicon partners. By moving the memory controller into the HBM stack's base die, NVHBM frees up package space for additional compute die area, potentially increasing compute capacity by up to 30%. Additionally, NVHBM reduces interposer routing complexity, simplifying connections for advanced packaging techniques. Compared to conventional HBM4e, NVHBM delivers 15% less power consumption, which can be redirected to enhance performance-per-watt or support more functional units in custom accelerators. Although NVHBM's benefits are primarily for Nvidia's partners, its incorporation into future AI accelerator designs could lead to improved energy efficiency and performance.",
  "summary": "Nvidia has unveiled NVHBM, a custom high-bandwidth memory implementation that promises higher bandwidth and lower power for customers building chips within the NVLink Fusion partner program.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}