Breaking the Memory Wall: How d-Matrix Is Redefining AI Inference with Chiplets
By Maurizio Di Paolo Emilio, embedded.com | April 23, 2025
As AI workloads push the limits of performance, power efficiency, and memory bandwidth, chiplets are rapidly emerging as the architectural solution of choice. In this interview, Sree Ganesan, Vice President of Product at d-Matrix, dives deep into how its pioneering chiplet-based platform is revolutionizing AI inference. From solving the memory wall with Digital In-Memory Computing (DIMC) to enabling seamless multi-chiplet communication via custom interconnects, d-Matrix reveals how its innovations are unlocking 10x faster token generation, 3x better energy efficiency, and a scalable roadmap for generative AI.
To read the full article, click here
Related Chiplet
- DPIQ Tx PICs
- IMDD Tx PICs
- Near-Packaged Optics (NPO) Chiplet Solution
- High Performance Droplet
- Interconnect Chiplet
Related News
- EdgeCortix Awarded New 3 Billion Yen NEDO Project to Develop Advanced Energy-Efficient AI Chiplet for Edge Inference and Learning
- D-Matrix Targets Fast LLM Inference for ‘Real World Scenarios’
- Powering AI with Chiplet Innovation: MSquare's ML100 IO Die Live at COMPUTEX 2025
- Baya Systems to Share Insights on Chiplet Integration, AI Scalability at DAC 2025
Latest News
- Xanadu and EV Group partner to build industrial-scale photonic quantum hardware
- AI Optical Interconnect Boom Drives U.S. Firms to Expand Southeast Asia Outsourcing, Opening the Door for Cross-Industry Entrants
- GlobalFoundries accelerates adoption of co-packaged optics for advanced AI data centers with SCALE optical module solution
- TSMC SoIC roadmap targets 2029 chip stacking
- Applied Materials Broadens Advanced Packaging Portfolio with Acquisition of NEXX