Researchers from multiple universities have tackled the challenge of designing large-scale DNN chiplet accelerators. These accelerators aim to optimize monetary cost (MC), performance, and energy efficiency. The complexity stems from various parameters, such as network-on-chip (NoC) communication, core positions, and different DNN attributes. This exploration of design space is crucial to find effective solutions for DNN accelerators.
Currently, existing DNN accelerators need help in achieving an optimal balance between MC, performance, and energy efficiency. The architecture and mapping co-exploration framework, Gemini, was introduced to address this need. Gemini employs a novel encoding method for low-power (LP) spatial mapping schemes to explore hidden optimization opportunities effectively. The framework uses a dynamic programming-based graph partition algorithm and a Simulated-Annealing-based (SA-based) approach for optimization.
Within Gemini, the mapping component uses the SA algorithm with five operators tailored to efficiently explore the LP spatial mapping space. These operators contribute to enhanced performance and energy efficiency by dynamically optimizing data transmission, intra-core dataflow, and D2D link communication. The evaluation involves assessing MC, energy consumption, and delay through an Evaluator module.
In terms of architecture, Gemini provides a highly configurable hardware template for precise evaluations of performance, energy, and MC. The framework’s experiments have demonstrated that its explored architecture and mapping scheme outperforms existing state-of-the-art (SOTA) designs like Simba with Tangram mapping. Gemini achieves significant improvements with only a marginal increase in MC, showcasing its effectiveness in co-exploring the architecture and mapping space.
In conclusion, the Gemini framework offers a comprehensive solution to the intricate challenges of designing DNN chiplet accelerators. The experiments not only validate Gemini’s effectiveness but also highlight the potential benefits of chiplet technology in architecture design. Overall, Gemini is a valuable tool for researchers and practitioners aiming to design high-performance and energy-efficient DNN accelerators.
Check out the Paper. All credit for this research goes to the researchers of this project. Also, don’t forget to follow on Twitter. Join our 35k+ ML SubReddit, 41k+ Facebook Community, Discord Channel, and LinkedIn Group. If you like our work, you will love our newsletter.