Enabling Hard Constraints in Differentiable Neural Network and Accelerator Co-Exploration
This work addresses a critical bottleneck in co-exploration for low-profile systems, offering an incremental improvement over existing differentiable methods.
The paper tackles the challenge of satisfying hard constraints like frame rate in differentiable neural architecture and hardware accelerator co-exploration, proposing HDX to achieve high-quality solutions that meet these constraints without compromising global design objectives.
Co-exploration of an optimal neural architecture and its hardware accelerator is an approach of rising interest which addresses the computational cost problem, especially in low-profile systems. The large co-exploration space is often handled by adopting the idea of differentiable neural architecture search. However, despite the superior search efficiency of the differentiable co-exploration, it faces a critical challenge of not being able to systematically satisfy hard constraints such as frame rate. To handle the hard constraint problem of differentiable co-exploration, we propose HDX, which searches for hard-constrained solutions without compromising the global design objectives. By manipulating the gradients in the interest of the given hard constraint, high-quality solutions satisfying the constraint can be obtained.