GpuPerHour

GpuPerHour is a service that helps users find and rent graphics processing units (GPUs) for their computing needs. It presents a wide selection of GPU options from various providers, making it easier to compare prices and specifications.

Benefits

GpuPerHour offers a large variety of GPU choices, with a total of 870 options available. Users can sort these options by price per GPU, starting from the lowest cost. This allows for efficient comparison of different providers, GPU models, memory (VRAM), host computer details, location, and hourly rental rates. The service aims to help users find the most cost-effective GPU solution for their specific tasks.

Use Cases

This service is useful for individuals and businesses that require powerful computing resources for tasks like artificial intelligence, machine learning, data analysis, or complex simulations. For example, users can find options such as the NVIDIA H100 SXM with 80GB VRAM from DigitalOcean, starting at $2.99 per hour. Alternatively, they can find more budget-friendly options like NVIDIA Tesla V100 GPUs with 16GB VRAM from VERDA, starting at just $0.05 per hour in Helsinki, Finland. The platform also lists various NVIDIA GeForce and RTX models from providers like Vast.ai and Hyperstack, offering different configurations and price points to suit diverse project requirements and budgets.

Pricing

Pricing varies significantly depending on the GPU model, provider, and location. Examples include:

• NVIDIA H100 SXM (80GB VRAM) from DigitalOcean starts at $2.99/GPU/hr.
• NVIDIA Tesla V100 (16GB VRAM) from VERDA starts at $0.05/GPU/hr.
• NVIDIA RTX A4000 (16GB VRAM) from Vast.ai is available for $0.11/GPU/hr.
• NVIDIA GeForce RTX 2080 Ti (11GB VRAM) from Vast.ai is listed at $0.13/GPU/hr.
• NVIDIA GeForce RTX 3060 (12GB VRAM) from Vast.ai is available for $0.14/GPU/hr.
• NVIDIA RTX A4000 (16GB VRAM) from Hyperstack starts at $0.15/GPU/hr.

Multiple GPU setups are also available, with total hourly costs listed alongside the per GPU price.