HyperAI built a leaderboard that compares GPU performance metrics specifically for machine learning and AI workloads.

URL: https://hyper.ai/gpu-leaderboard

It shows up-to-date hardware specifications and benchmark figures (e.g., FP16/FP32/FP64 performance, memory bandwidth) for a variety of modern GPUs. The goal is to help researchers, engineers, and practitioners understand how different GPUs stack up for ML/AI use cases.

Key features * 29+ GPUs with detailed hardware specs * Side-by-side performance comparison across FP types * Focused on AI/ML workloads rather than gaming

Would love feedback on the layout and utility — especially from folks who do GPU benchmarking or care about hardware trade-offs in AI workloads.