Thunder Compute (YC S24)

Thunder Compute is a cool cloud space where developers can run their work on cloud GPUs with just one line of code. Carl and Brian started it. They met at Georgia Tech and have been close friends for over six years. Their goal is to make setting up cloud platforms for projects simple and easy.

Key Features

Thunder Compute offers a flexible cloud space that can change to fit any project with just one command. Some cool features include:

• On Demand CPU Only Instances: All instances on Thunder Compute are on demand CPU only instances, like those found on AWS, GCP, or Azure.
• Virtual GPU over TCP Technology: Running commands like $ nvidia-smi on Thunder Compute works as expected with a physical GPU, while lspci shows no connected GPUs.
• Performance Optimization: Even though TCP is slower than PCIe, Thunder Compute is optimized to minimize performance impact, making the real world slowdown often unnoticeable for common data science tasks.

Benefits

Thunder Compute makes setting up cloud platforms easy. It balances scalability, cost, and simplicity. It allows developers to focus on their workloads without worrying about complex setups.

Use Cases

Here are some examples of how Thunder Compute works:

• Using your local filesystem?

$ tnr start sync

• Scaling to 50 A100 GPUs?

$ tnr device A100 n 50

• Running a Jupyter Notebook?

$ notebook

• Developing without a GPU?

$ tnr device cpu

All of this is available at industry leading prices and can be set up in minutes with a simple pip install tnr command.

Cost Price

Thunder Compute offers its services at industry leading prices, making it an affordable option for developers.

Funding

[No funding details provided in the article.]