
NVIDIA NVLink Fusion Expands With NVHBM Custom High-Bandwidth Memory
NVIDIA expanded NVLink Fusion with NVHBM, a custom high-bandwidth memory technology that integrates the memory controller into the HBM base die. This allows for higher memory performance and efficiency in semi-custom AI infrastructure.
Why it matters
This technology enables companies to develop more powerful custom AI chips faster by reducing engineering complexity and improving hardware efficiency.
The details
NVHBM frees up to 25% more area on the XPU compute die. Annapurna Labs will support NVLink Fusion starting with Trainium4 chips, allowing Amazon chips and NVIDIA GPUs to use a common rack-scale architecture. NVLink Fusion also allows partners to connect custom XPUs and CPUs to NVIDIA's rack-scale platform.
What's next
Annapurna Labs will work with NVIDIA on NVHBM technology and the NVLink scale-up architecture to enhance AI workload performance.
Get the weekly recap
The stories like this one, picked and explained — once a week, straight to your inbox.