Devices · 2h ago
Qualcomm Dragonfly AI rack targets Nvidia with 43TB LPDDR5x
Qualcomm unveiled Dragonfly, a rack-scale AI inference platform with 43TB of LPDDR5x memory, aiming to compete with Nvidia, AMD, and Huawei. The system uses proprietary HBC memory instead of HBM for cost and power efficiency. Future versions are expected to exceed 7PB/s bandwidth.
Meridian48 take
Qualcomm's HBC memory bet could disrupt the AI hardware market if it delivers on performance claims, but it faces an uphill battle against Nvidia's entrenched ecosystem.
Read the full reporting
Qualcomm targets Nvidia, AMD, Huawei with Dragonfly AI accelerator rack loaded with 43TB of LPDDR5x, future generations set to smash 7PB/s bandwidth →
TechRadar
qualcomm-dragonflyai-inference