Startup d-Matrix Will Pair Its Raptor Memory-Based XPU To Nvidia Rackscale Iron - The Next Platform

US startup d-Matrix announced the launch of its Raptor memory-based XPU chip, planned to be paired with NVIDIA's RackScale Iron platform. This collaboration represents a significant advancement in AI ... Read original →

The collaboration between d-Matrix and NVIDIA signals a pivotal shift in AI computing architecture. By integrating Raptor memory technology, the partnership enhances data processing speed and energy efficiency, particularly for large-scale AI workloads, challenging current DRAM and HBM dominance. This move accelerates the transition toward heterogeneous computing and XPU adoption. From a compliance standpoint, such integration may trigger heightened scrutiny in supply chain security amid ongoing U.S.-China tech decoupling. Competitors like AMD and Intel are likely to accelerate internal memory architecture development to reduce reliance. Over the next 12–24 months, this synergy could define new data center standards, pushing the industry toward more efficient, low-power computing units and reinforcing emerging technological moats.