We have created high performance CUDA JPEG library for rapid development of applications capable to compress or decompress images with extremely high throughput. Our solution is based on NVIDIA CUDA technology and offers ultra fast JPEG compression and decompression on the GPU due to parallel implementation of Baseline JPEG algorithm.

CUDA JPEG library gives opportunity to speed up image compression without deep understanding of compression algorithm and GPU programming. We offer solution which can do that for you really fast. Probably, much faster that you have ever experienced.

CUDA JPEG library is worth using in high performance applications because jpeg compression speed up on GPU could be really useful if you need to compress significant number of images.

JPEG library featuresjpeg library

  • Baseline JPEG compression and decompression for grayscale (8-bit) and color (24-bit) images
  • 12-bit JPEG compression on CUDA
  • Extremely fast lossy image encoding and decoding with variable compression ratio
  • Data input: 8/24-bit images from CPU/GPU memory or HDD/SSD
  • Data output: final compressed JPEG image (in CPU/GPU memory or HDD/SSD)
  • Standard set of computations for parallel implementation of Baseline JPEG: Color Transform, 2D DCT, Quantization, Zig-zag, AC/DC, DPCM, RLE, Huffman coding, Byte stuffing, JFIF formatting
  • Maximum input image size is 16000×16000 or even more (optional)
  • Optimized for the latest NVIDIA GPUs (Kepler, Maxwell, Pascal)
  • Compatibility with FFmpeg to read/write MJPEG streams
  • Windows-7/8/10 and Linux Ubuntu/CentOS compatibility
  • Sample applications for MS Visual Studio 2017

Demo Version of CUDA JPEG Compression and Decompression

Demo version of CUDA JPEG software is available by request. Now this is a part of our Image Processing SDK and it works with the following parameters: input image, output image, device id, sampling format, quality value, subsampling mode, restart interval. With demo software (MSVS 2017 project with DLL) user can test codec's performance on available GPU prior to purchasing and evaluate speed up in comparison with existing image compression libraries on CPU. Demo version can work with 8/24-bit and 12/36-bit images. The application is fully functional and can be used for evaluation purposes only.

Applications for CUDA JPEG library

  • Batch raw image compression to JPEG
  • Realtime image processing (high quality demosaicing + jpeg compression) from high speed and high resolution cameras
  • Batch image resize and recompression (decoding + crop + resize + sharp + encoding)
  • Photo compression software
  • Machine vision and scientific scope: fast image compression of high speed camera output
  • Full image processing on GPU for video cameras
  • GPU RAW Processor
  • Fast VCR software for XIMEA cameras (realtime raw image processing with CUDA JPEG encoding and integrated camera control) - done


For any further information concerning CUDA JPEG library or free trial please contact us via email or contact form.

Contact Form

This form collects your name and email. Check out our Privacy Policy on how we protect and manage your personal data.