Source: Nvidia Not only that, but Nvidia also intentionally handicapped the H200 and B200 systems with FP8 despite using FP4 on the GB200. All of these systems could benefit from a memory perspective with using FP4, which would push all of these curves to the right with regards to the inte...
NVIDIA GPU GPU Memory On-Demand Price/hr NVIDIA L40S 48GB $2.75 NVIDIA H100 80 GB $8.25 For example, fine-tuning Mistral 7B on 1500 samples on a single NVIDIA L40S takes ~10 minutes and costs ~$0.45. We’re just getting started We are excited to collaborate with NVIDIA to ...