Qualcomm(QCOM 0.00%↑) announced this morning that it has completed initial validation of its High Bandwidth Compute (HBC) memory architecture, which will be a core component of its upcoming Dragonfly data center AI accelerators. The stacked memory architecture will let Qualcomm and its manufacturing partners stack multiple layers of DRAM on a small compute die. This puts memory closer to the compute hardware that accesses it, giving Qualcomm the energy-efficiency and cost benefits of tighter hardware integration.
This latest validation brings HBC a step closer to commercial deployment, with samples expected to begin next year. Qualcomm has now confirmed that the technology works, with energy efficiency and memory bandwidth close to expectations. That lets the company begin volume production next year, while keeping customers (and investors) up to date on the progress of the Dragonfly AI250 accelerator and its successors.
