Mirendil taps AI Hypercomputer TPUs and GPUs for pre- and post-training applications

Mirendil, a frontier AI lab, partnered with Google Cloud to utilize their AI Hypercomputer, leveraging TPUs and NVIDIA GPUs for pre- and post-training applications. This will accelerate and democratize AI research and development. The partnership enables Mirendil to access critical compute quickly and match workloads to the best-suited architecture. This collaboration will help scientists and engineers run the research loop faster and at a greater scale. Mirendil is already live with a TPU v5P cluster, with NVIDIA AI accelerated computing systems coming online soon.