libjpeg-turbo uses SIMD instructions to accelerate JPEG compression and decompression by 2-6x relative to, while maintaining full backward API and ABI compatibility with, libjpeg v6b. Our project builds upon the work of Tom Lane, the founder of the Independent JPEG Group, and attempts to further encourage adoption of the JPEG/JFIF image format (Rec. ITU-T T.81 | ISO/IEC 10918-1 and Rec. ITU-T T.871 | ISO/IEC 10918-5) by providing an open source codec, under a business-friendly license, that is as fast, easy to use, and secure as possible. libjpeg-turbo is an ISO/ITU-T reference implementation (Rec. ITU-T T.873 | ISO/IEC 10918-7) and is used by numerous web browsers, operating systems, and software applications that handle JPEG images, so it indirectly has a user base of billions of people.

Some of the features that set libjpeg-turbo apart from libjpeg include:

• The TurboJPEG API, which is much more user-friendly than the traditional libjpeg API and provides some features that would be difficult to implement with the latter, such as:
  • compressing from, decompressing to, encoding, and decoding planar YUV images
  • lossless image transformations
• SIMD-accelerated implementations of the JPEG algorithms for various popular CPU architectures
• Support for mobile platforms
• Comprehensive and continuous fuzz testing to detect potential security flaws before they are released into the wild
• A modern, flexible build system based on CMake
• Partial image decompression
• A comprehensive Java API
• Official, signed SDK packages provided for popular operating systems and CPU architectures

libjpeg-turbo is one of the few pieces of critical IT infrastructure that is sustained solely through patronage and funded development. However, this often requires a delicate balancing act: keeping the project moving forward so it can remain relevant, remain responsive to the user community, and provide a timely return on investment for funded development sponsors, while also keeping the project moving forward in a way that doesn't require the maintainer to eat labor cost or borrow against expected future funding (which sometimes doesn't arrive) in order to finish a strategically important feature or release. Developing libjpeg-turbo independently ensures that it can remain community-focused and free of any one organization’s agenda, but it also requires continuous funding, which has historically not been easy to secure.

All sponsorship money goes directly toward funding the labor necessary to maintain libjpeg-turbo, support the user community, and implement bug fixes and strategically important features.