We’ve released CompreFace with support for the latest generation GPU!
In our 1.0.1 minor version, we updated the CUDA library, which is used for running CompreFace on GPU, to version 11.2. If you don’t use GPU for facial recognition with CompreFace, you can continue using the 1.0.0 version, as this is the only change.
With the new CUDA version, CompreFace now supports the latest generation of Nvidia GPUs. However, CUDA dropped support of some of the older GPUs. If you use GPU, please read this article carefully to make sure yours will work.
How to Determine if My GPU is Supported by CompreFace
CompreFace uses the CUDA library to run computations on GPU. This means that all GPUs supported by CUDA are supported by CompreFace as well.
According to Nvidia terminology, each GPU has Compute Capability, often called the “SM version.” This version number identifies the computational capabilities supported by the GPU hardware and is used by applications, like CompreFace, at runtime to determine which hardware features and/or instructions are available on the present GPU. CUDA uses those capabilities, this is why Compute Capability is the main criteria to determine if your GPU is supported by CUDA.
What GPUs are Supported by CUDA 11.2 (CompreFace 1.0.1)
CUDA 11.2 supports all GPUs with Compute Capability 3.5 to 8.6. This means that it works with some GPUs from Kepler architecture and all GPUs through Ampere architecture.
Here is the list of oldest GPUs supported by CUDA 11.2 from Kepler architecture:
- Desktop/Laptop: GeForce GTX Titan Z, GeForce GTX Titan Black, GeForce GTX Titan, GeForce GTX 780 Ti, GeForce GTX 780, GeForce GT 640 (GDDR5), GeForce GT 630 v2, GeForce GT 730, GeForce GT 720, GeForce GT 710, GeForce GT 740M (64-bit, DDR3), GeForce GT 920M
- Professional: Quadro K6000, Quadro K5200
- Server: Tesla K40, Tesla K20x, Tesla K20
No older GPUs will work with CompreFace 1.0.1.
Here is the list of the latest GPUs supported by CUDA 11.2 from Ampere architecture:
- Desktop/Laptop: GeForce RTX 3090 Ti, RTX 3090, RTX 3080 Ti, RTX 3080 12GB, RTX 3080, RTX 3070 Ti, RTX 3070, RTX 3060 Ti, RTX 3060, RTX 3050, RTX 3050 Ti(mobile), RTX 3050(mobile), RTX 2050(mobile), MX570
- Professional: RTX A6000, RTX A5500, RTX A5000, RTX A4500, RTX A4000, RTX A2000, RTX A5000(mobile), RTX A4000(mobile), RTX A3000(mobile), RTX A2000(mobile)
- Server: A40, A16, A10, A2
What GPUs were Supported by CUDA 10.1 (CompreFace 1.0.0)
We do realize that even old generation GPUs have some advantages in facial recognition speed. In this case, we suggest using our previous CompreFace version 1.0.0, which was based on CUDA 10.0. CUDA 10.0 supports all GPUs with Compute Capability from 3.0 to 7.5, so it works with all GPUs from Kepler architecture and all GPUs through Turing architecture.
Here is the list of oldest GPUs supported by CUDA 10.0 from Kepler architecture:
- Desktop/Laptop: GeForce GTX 770, GeForce GTX 760, GeForce GT 740, GeForce GTX 690, GeForce GTX 680, GeForce GTX 670, GeForce GTX 660 Ti, GeForce GTX 660, GeForce GTX 650 Ti BOOST, GeForce GTX 650 Ti, GeForce GTX 650, GeForce GTX 880M, GeForce GTX 870M, GeForce GTX 780M, GeForce GTX 770M, GeForce GTX 765M, GeForce GTX 760M, GeForce GTX 680MX, GeForce GTX 680M, GeForce GTX 675MX, GeForce GTX 670MX, GeForce GTX 660M, GeForce GT 750M, GeForce GT 650M, GeForce GT 745M, GeForce GT 645M, GeForce GT 740M, GeForce GT 730M, GeForce GT 640M, GeForce GT 640M LE, GeForce GT 735M, GeForce GT 730M
- Professional: Quadro K5000, Quadro K4200, Quadro K4000, Quadro K2000, Quadro K2000D, Quadro K600, Quadro K420, Quadro K500M, Quadro K510M, Quadro K610M, Quadro K1000M, Quadro K2000M, Quadro K1100M, Quadro K2100M, Quadro K3000M, Quadro K3100M, Quadro K4000M, Quadro K5000M, Quadro K4100M, Quadro K5100M, NVS 510, Quadro 410
- Server: Tesla K10, GRID K340, GRID K520, GRID K2
No older GPUs will work with CompreFace 1.0.0.
Here is the list of the latest GPUs supported by CUDA 10.0 from Turing architecture:
- Desktop/Laptop: NVIDIA TITAN RTX, GeForce RTX 2080 Ti, RTX 2080 Super, RTX 2080, RTX 2070 Super, RTX 2070, RTX 2060 Super, RTX 2060 12GB, RTX 2060, GeForce GTX 1660 Ti, GTX 1660 Super, GTX 1660, GTX 1650 Super, GTX 1650, MX550, MX450
- Professional: Quadro RTX 8000, Quadro RTX 6000, Quadro RTX 5000, Quadro RTX 4000, T1000, T600, T400, T1200(mobile), T600(mobile), T500(mobile), Quadro T2000(mobile), Quadro T1000(mobile)
- Server: Tesla T4
We hope these minor updates will enhance your experience with CompreFace. If you have any questions along the way, feel free to contact our team.
Join us! Exadel provides cutting-edge products and services. Check out our open positions here.
Visit this page to find out more about CompreFace
Don’t miss the latest articles and news about CompreFace