Intel's Ben Widawsky published the set of patches enabling renderbuffer decompression for the i965 driver plus the necessary GBM modifications. With these patches there is the potential for massive bandwidth savings. Results shared by Widawsky on a Skylake GT4 GPU show the compression dropping the read bandwidth from 603 MiB/s to 259 MiB/s and the write bandwidth dropping from 615 MiB/s to 337 MiB/s, when using a modified version of kmscube for testing.
Ben further explained this renderbuffer decompression, "this feature reduces memory bandwidth by allowing the GPU to work with losslessly compressed data and having that compression scheme understood by the display engine for decompression. The display engine will decompress on the fly and scanout the image...The hardware achieves this savings by maintaining an auxiliary buffer containing 'opaque' compression information. It's opaque in the sense that the low level compression scheme is not needed, but, knowledge of the overall layout of the compressed data is required. The auxiliary buffer is created by the driver on behalf of the client when requested. That buffer needs to be passed along wherever the main image's buffer goes."
But before Intel Linux users get too excited about massive memory bandwidth savings, software needs to be updated to make use of the renderbuffer decompression. The modesetting DDX driver will need to be updated along with clients like Wayland's Weston compositor plus there is protocol work left for EGL, Wayland, DRI, etc. Nevertheless, the bandwidth savings and Intel's large OTC staff should be able to push this through to mainline fruition in the months ahead. More details via this patch series.