NVIDIA NVLink Fusion Expands With NVHBM Custom High-Bandwidth Memory
NVIDIA expands NVLink Fusion with NVHBM, a new high-bandwidth memory technology for AI infrastructure.
Intelligence analysis by Qwen 2.5 (3B)
NVIDIA introduces NVHBM, a custom high-bandwidth memory technology for AI infrastructure, expanding NVLink Fusion.
NVHBM is a new memory technology that lets AI chips use more memory and do it more efficiently, saving power and space.
Analysis
{"heading_1":"NVHBM Architecture","subheading_1":"NVHBM Architecture Overview","paragraph_1":"NVLink Fusion enables partners to connect custom XPUs and CPUs to NVIDIA's rack-scale platform, offering a proven technology stack for scale-up and scale-out networking.","paragraph_2":"NVLink Fusion allows hyperscalers and AI-native companies to focus on XPU innovation while using a stable technology stack for infrastructure deployment.","subheading_2":"NVHBM Performance and Efficiency","subheading_3":"NVHBM Implementation and Partners","subheading_4":"NVLink Fusion Benefits"}
Key points
- NVHBM integrates a custom memory controller into HBM base die
- NVHBM delivers up to 30% greater memory bandwidth and 15% lower power consumption
- NVHBM frees up to 25% more area on XPU compute die compared to standard HBM4E
- NVHBM is a standard implementation available from multiple memory providers
- NVHBM will be validated and offered by leading memory partners
NVHBM could lead to faster and more efficient AI systems, reducing costs and improving performance.
If NVHBM integration is not successful, it could delay the deployment of custom AI chips and increase costs.
