Global / Semiconductors
d-Matrix to use NVIDIA NVLink Fusion for next-generation Raptor XPUs
An inference chipmaker joins NVIDIA's ecosystem by adopting the platform's networking and rack architecture for chip deployment.
d-Matrix, an AI inference chipmaker, announced it will integrate NVIDIA NVLink Fusion to connect its Raptor XPUs to NVIDIA's AI infrastructure. The move places d-Matrix within NVIDIA's broader hardware ecosystem for rack-scale AI deployment.
Published · significance 52 of 100 (medium) · 1 source
What happened
d-Matrix announced adoption of NVIDIA NVLink Fusion to connect its next-generation Raptor XPUs to NVIDIA's AI platform. The integration includes NVIDIA's MGX rack architecture and Spectrum-X networking for scale-out connectivity. d-Matrix joins a growing list of ecosystem partners building on NVIDIA's infrastructure.
Why it matters
This signals continued consolidation around NVIDIA's hardware ecosystem for AI inference workloads. Chipmakers dependent on NVIDIA's interconnect and platform standards have limited ability to diversify their go-to-market strategy, strengthening NVIDIA's control over the inference hardware layer. The move reflects market dynamics where specialised AI chipmakers must integrate with dominant platform providers rather than compete independently.
What changes
d-Matrix's Raptor XPUs will be deployable at rack scale through NVIDIA's standardised architecture, simplifying customer deployment. However, d-Matrix's product roadmap is now tied to NVIDIA's technology choices and roadmap updates.
Involved
Sources
- d-Matrix Adopts NVIDIA NVLink Fusion for Rack-Scale XPU Deployment — NVIDIA Newsroom
Written by AI from the reports above; scored by a published formula. How we work. Found a mistake? Email lockedinshreyash@gmail.com. Up To Date summarises and links to original reporting; it never reproduces articles.
Related coverage
- ASML's next-generation High NA chipmaking tools sold out through 2027 — Semiconductor supply for AI chips tightens as manufacturers commit to pricier equipment. (2026-09-14)
- Jensen Huang forecasts Nvidia will grow 70 per cent next year — Huang insists the gains reflect broad demand, not circular dealing between Nvidia and its customers. (2026-09-10)
- China's Huawei sets 2027 launch for new AI chips as it targets Nvidia (2026-09-17)
- d-Matrix to use Nvidia chip-linking technology in AI inference servers — Inference specialist gains access to Nvidia's interconnect standard as training-to-deployment workload shift accelerates. (2026-09-11)
- Justice Department investigates Nvidia's $17 billion licensing agreement with Groq — A formal information request signals antitrust scrutiny of chip technology consolidation in AI infrastructure. (2026-09-10)