Nvidia's Llama Nemotron models were developed quickly, with researchers sacrificing their compute power for the project.According to Jonathan Cohen, Vice President of Applied Research at Nvidia, GPUs are considered currency for AI researchers.The development of Llama Nemotron models took one to two months, thanks to the company-wide culture of prioritizing major projects.The project required sacrifices in computing power and personnel, showcasing teamwork and leadership at Nvidia.