MCMComm: Hardware-Software Co-Optimization for End-to-End Communication in Multi-Chip-Modules
By Ritik Raj, Shengjie Lin, William Won and Tushar Krishna
Georgia Institute of Technology, GA, USA
Increasing AI computing demands and slowing transistor scaling have led to the advent of Multi-Chip-Module (MCMs) based accelerators. MCMs enable cost-effective scalability, higher yield, and modular reuse by partitioning large chips into smaller chiplets. However, MCMs come at an increased communication cost, which requires critical analysis and optimization. This paper makes three main contributions: (i) an end-to-end, off-chip congestion-aware and packaging-adaptive analytical framework for detailed analysis, (ii) hardware software co-optimization incorporating diagonal links, on-chip redistribution, and non-uniform workload partitioning to optimize the framework, and (iii) using metaheuristics (genetic algorithms, GA) and mixed integer quadratic programming (MIQP) to solve the optimized framework. Experimental results demonstrate significant performance improvements for CNNs and Vision Transformers, showcasing up to 1.58x and 2.7x EdP (Energy delay Product) improvement using GA and MIQP, respectively.
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 Technical Papers
- System-Technology Co-Optimization for Dense Edge Architectures using 3D Integration and Non-Volatile Memory
- Co-Optimization of Power Delivery Network Design for 3-D Heterogeneous Integration of RRAM-Based Compute In-Memory Accelerators
- Advancing Trustworthiness in System-in-Package: A Novel Root-of-Trust Hardware Security Module for Heterogeneous Integration
- Occamy: A 432-Core 28.1 DP-GFLOP/s/W 83% FPU Utilization Dual-Chiplet, Dual-HBM2E RISC-V-based Accelerator for Stencil and Sparse Linear Algebra Computations with 8-to-64-bit Floating-Point Support in 12nm FinFET
Latest Technical Papers
- Spatiotemporal thermal characterization for 3D stacked chiplet systems based on transient thermal simulation
- Interconnect-Aware Logic Resynthesis for Multi-Die FPGAs
- Scope: A Scalable Merged Pipeline Framework for Multi-Chip-Module NN Accelerators
- Scaling Routers with In-Package Optics and High-Bandwidth Memories
- TDPNavigator-Placer: Thermal- and Wirelength-Aware Chiplet Placement in 2.5D Systems Through Multi-Agent Reinforcement Learning