Multi chiplet gpus boost speed and cut energy use drastically
Divide and conquer: Scalable performance and energy in MCM GPUs
Hardware Architecture
Summary
GPUs, which help computers run fast graphics and calculations, usually rely on one big chip. The authors study a design where many smaller chips work together on one package. They found that spreading the work across multiple chips with the right connection pattern can make the GPU run much faster and use less energy. This approach, called disaggregation, could be key for future better and more efficient GPUs.
What this means in practice
- •For gpu hardware designers: Design multi-chip GPUs with optimized chiplet counts and interconnects to significantly improve compute speed and reduce power consumption.
- •For data center infrastructure teams: Deploy GPUs using disaggregated multi-chip modules to achieve higher energy efficiency and better performance per watt in large-scale computing tasks.
Authors
Mario Ibáñez Bolado, Borja Pérez Pavón, Jose Luis Bosque Orero, Julio Ramón Beivide
Abstract
Multi-chip-module (MCM) GPUs offer a promising path to scale compute capability beyond monolithic designs by integrating multiple chiplets on a common package. However, the impact of disaggregation on performance scalability and energy consumption remains underexplored. The design space grows rapidly across dimensions such as SMs per chiplet, chiplet count, and interconnection network. The inter-chiplet network is particularly critical, as it determines whether additional compute resources translate into performance gains. This limited understanding leaves industry and research without clear guidance on the performance and energy trade-offs of MCM GPU scaling. In this work, we investigate whether distributing compute and memory capability across multiple chiplets offers a more scalable alternative to concentrating resources. We quantify their effects on performance, energy, and efficiency and examine how inter-chiplet topology influences scalability at different system sizes. Our results demonstrate that a 16 chiplet Torus configuration with 256 SMs delivers a remarkable $2.40\times$ performance improvement over a state-of-the-art MCM architecture with the same compute capability, while simultaneously reducing energy consumption by $4.45\times$. These substantial gains provide evidence that disaggregation is a first-order architectural factor and will be critical to unlocking the performance and energy-efficiency potential of next-generation GPUs.