Intel ends development and support for its Deep Link technology

Table of Contents
After just a couple of generations of Intel releasing dedicated graphics cards, one of the core technologies that gave it an advantage over some other combinations is being discontinued. As discovered by Haze inside the Deep Link Github issue tracker, it’s been put forward by an Intel dev that it is no longer actively being maintained to support newer cards and features.
That is seemingly further confirmed with the support page for Deep Link, where it specifies it only supports 11th Gen and newer CPUs and just the Intel Arc Alchemist graphics cards. The related products section only lists out the A range of Intel GPUs.
That means if you have a newer card like the Arc B580, which itself is a rather impressive option in the current market, you won’t be able to benefit from the technology, even from whatever it’s stuck with.
What was Deep Link?
Deep Link listed four main features that it allowed your system to utilize. These include Hyper Encode, Dynamic Power Share, Stream Assist, and Hyper Compute. So, it could be just one way Intel was trying to promote its new GPUs. With the Battlemage generation working and recent layoffs, it might not require these anymore.
But if you were interested, Hyper Encode allowed accelerating encoding on multiple media engines, combining the processing power of both components, and worked on certain apps. These were HandBrake, Intel oneAPI, Blackmagic DaVinci Resolve, and Magix Video, working with the formats AVC, HEVC, and VP9 to work with the best quality video.
Power Share, on the other hand, worked on power distribution. It splits how much each component gets, depending on your use case at the time, to get you the best performance. Similarly, Hyper Compute worked by combining compute engines and AI accelerators of the two to work together instead of relying on just one of the components to do all the work.
While Stream Assist worked to offload streaming tasks to integrated graphics on the processor, leaving the GPU to do all the work on the game rendering. That led to keeping your performance intact for both your stream and game, but was limited to OBS, Xspli, and AVerMedia CamEngine, and OBS is where the issue first cropped up for the user to realize it didn’t function.