Qualcomm's First HBC Chip Passes Hardware Validation; Moxin AI Closes Nearly 1 Billion Yuan Series C+

Two AI-hardware developments surfaced on November 4. Qualcomm said its first-generation high-bandwidth computing (HBC) chip has completed hardware validation, successfully testing the full memory read-write path and near-memory compute, a technique that moves compute next to memory to cut data movement and lift bandwidth and energy efficiency while lowering AI-inference power and cost; a data-center product has taped out, with the Dragonfly AI250 using HBC expected to sample in the first half of 2027. Separately, Moxin Artificial Intelligence announced a Series C+ round of nearly 1 billion yuan, backed by Renai Capital, the Haitang Fund, Dongheng Petroleum and others, with proceeds going to a full-stack upgrade of its sparse dedicated inference chips, next-generation chip R&D, and a sparse-computing industry-academia-research alliance. The moves highlight intensifying investment in domestic inference silicon as Chinese models scale and the cost of serving massive token volumes becomes the central bottleneck.

Source: Interface News, Nov 4, 2026 | 2026-11-04