This repository provides an API for extracting GPU information from Nvidia, AMD, and Intel GPUs in JSON format. The data is collected from wiki sources and updated weekly using GitHub Actions. The API contains detailed information on various GPU models, including their specifications, performance metrics, and other relevant details.
You can access the API at the following path: https://raw.githubusercontent.com/voidful/gpu-info-api/gpu-data/gpu.json
{
"AD104-250": {
"Model": "GeForce RTX 4070",
"Launch": "2023-04-13 00:00:00",
"Code name": "AD104-250",
"Fab (nm)": 4.0,
"Die size (mm2)": 294.5,
"Bus interface": "PCIe 4.0 x16",
"Core clock (MHz)": "1920",
"Core config": "5888:184:64:184:46 (46)(4)",
"Memory Bandwidth (GB/s)": 504.0,
"Memory Bus type": "GDDR6X",
"Memory Bus width (bit)": "192",
"Vendor": "NVIDIA",
"Fillrate Pixel (GP/s)": 158.4,
"Fillrate Texture (GT/s)": 455.4,
"TDP (Watts)": 200.0,
"Release Price (USD)": 599.0,
"SM count": "46",
"Process": "TSMC N4",
"Transistors (billion)": 35.8,
"L Cache (MB)": 36,
"Memory Size (GB)": 12,
"Clock speeds Memory (MT/s)": 21000,
"Release price (USD) Founders Edition": "$599",
"Clock speeds Boost core clock (MHz)": 2475,
"Single-precision TFLOPS": "22.6",
"Double-precision TFLOPS": "0.353",
"Half-precision TFLOPS": "22.6",
"Pixel/unified shader count": 5888.0,
"GPU Type": "Desktop"
},
"AD104-400": {
"Model": "GeForce RTX 4080",
"Code name": "AD104-400",
"Fab (nm)": 4.0,
"Die size (mm2)": 294.5,
"Bus interface": "PCIe 4.0 x16",
"Core clock (MHz)": "2310",
"Core config": "7680:240:80:240:60 (60)(5)",
"Memory Bandwidth (GB/s)": 504.0,
"Memory Bus type": "GDDR6X",
"Memory Bus width (bit)": "192",
"Vendor": "NVIDIA",
"Fillrate Pixel (GP/s)": 208.8,
"Fillrate Texture (GT/s)": 626.4,
"TDP (Watts)": 285.0,
"Release Price (USD)": 899.0,
"SM count": "60",
"Process": "TSMC N4",
"Transistors (billion)": 35.8,
"L Cache (MB)": 48,
"Memory Size (GB)": 12,
"Clock speeds Memory (MT/s)": 21000,
"Clock speeds Boost core clock (MHz)": 2610,
"Single-precision TFLOPS": "35.5",
"Double-precision TFLOPS": "0.554",
"Half-precision TFLOPS": "35.5",
"Processing power (TFLOPS) Tensor compute (FP16) (2: sparse)": "142 (284) 160 (321)",
"Ray-tracing Performance (TFLOPS)": 92.7,
"Pixel/unified shader count": 7680.0,
"GPU Type": "Desktop"
}
}This repository is inspired by and borrows from the following project: https://github.com/owensgroup/gpustats
To use the API, you can simply make an HTTP request to the API path mentioned above. The API will return the GPU information in JSON format, which you can then parse and use in your application.
Here's an example of the JSON data returned by the API for two GPU models:
{
"AD104-250": {
"Model": "GeForce RTX 4070",
"Launch": "2023-04-13 00:00:00",
...
"GPU Type": "Desktop"
},
"AD104-400": {
"Model": "GeForce RTX 4080",
...
"GPU Type": "Desktop"
}
}You can then extract specific information about a GPU model using its key, such as "AD104-250" or "AD104-400".
If you'd like to contribute to this project or have any suggestions, feel free to open an issue or submit a pull request. We appreciate any feedback and assistance in improving the quality and accuracy of the GPU information provided by this API.