Compression Enabled MRAM Memory Chiplet Subsystems for LLM Inference Accelerators
By Andy Green (Numen) and Nilesh Shah (Zeropoint Technologies)
This session unveils an AI-specific chiplet alternative to GPU-based inference that can deliver HBM-like bandwidth at 30-50% lower power.
Related Chiplet
- FlexGen Multi-Die Smart Network-on-Chip (NoC) IP
- Ncore Multi-Die Interconnect IP
- Integrated voltage regulator (IVR) chiplet
- High-performance connectivity chiplets
- eFPGA Chiplet
Related Videos
- Chiplet Quilting for the Age of Inference
- Paving the way for the Chiplet Ecosystem
- Standards for Chiplet Design with 3DIC Packaging (Part 1)
- Standards for Chiplet Design with 3DIC Packaging - Part 2
Latest Videos
- Advanced Packaging: Enabling AI at Scale
- Accelerating Chiplet Design with OCP Standards Compliant Chiplet Framework and Automation
- Chiplet Identity and Roots of Trust in FCSA Systems: An Architecture-Agnostic Security View
- From Transistors to AI: The Chiplet Revolution with Ramune Nagisetty
- Scaling AI with Chiplets & CPO