The article discusses the availability of low-cost, on-demand GPUs that are preconfigured for machine learning and AI tasks. These GPUs can be launched immediately and are designed to be user-friendly for those who need to perform machine learning tasks. The GPUs available range from the Nvidia H100 SXM with 640GB RAM and 116736 CUDA cores to the Nvidia Quadro RTX 6000 with 24GB RAM and 4608 CUDA cores. Pricing details are provided, excluding sales tax, and the final rate is usually lower.
The article also mentions an opportunity for those who have idle GPUs. It suggests that these GPUs can be rented out to those in need of compute power. This could be a profitable venture for those operating a GPU cluster or an AI company looking to monetize their idle compute power.
Key takeaways:
The service offers low-cost on-demand GPUs preconfigured for Machine Learning tasks.
Various GPU configurations are available, including models from Nvidia such as H100 SXM, H200, GeForce GTX 1080 Ti, A100 SXM, GeForce RTX 4090, GeForce RTX 3060, and Quadro RTX 6000.
The pricing for each configuration is provided, excluding sales tax, and the final rate is usually lower.
There is an opportunity to earn by renting out idle GPUs or operating a GPU cluster.
