Efficient video encoder architecture

A method for processing video data, the method comprising:

• determining, during a first video coding stage implemented at a first frame rate, first motion vectors for a first subset of frames in a sequence of frames;
• store the first motion vectors in a memory for use in a second video coding stage;
• determining, during a second video coding stage implemented at a second frame rate that is higher than the first frame rate, second motion vectors for a second subset of frames in the sequence of frames, wherein the second subset of frames includes first frames that were processed during the first video coding stage and second frames that were not processed during the first video coding stage, wherein a first portion of the second motion vectors determined for the first frames during the second video coding stage includes the first motion vectors stored in the memory and a second portion of the second motion vectors determined for the second frames during the second video coding stage is predicted using one or more motion vectors from the first motion vectors stored in the memory, and wherein the second subset of frames includes a greater number of frames than the first subset of frames based on the second frame rate being higher than the first frame rate;
• reconstructing, during the second video coding stage at the second frame rate, the first subset of frames using the first motion vectors determined during the first video coding stage;
• and
• reconstructing, during the second video coding stage at the second frame rate, the second subset of frames using the second motion vectors determined during the second video coding stage, wherein the first video coding stage and the second video coding stage are implemented at least partly in parallel.

A non-transitory computer-readable medium having stored thereon instructions which, when executed by one or more processors, cause the one or more processors to:

• determine, during a first video coding stage implemented at a first frame rate, first motion vectors for a first subset of frames in a sequence of frames;
• output the first motion vectors for storage in a memory for use in a second video coding stage;
• determine, during a second video coding stage implemented at a second frame rate that is higher than the first frame rate, second motion vectors for a second subset of frames in the sequence of frames, wherein the second subset of frames includes first frames that were processed during the first video coding stage and second frames that were not processed during the first video coding stage, wherein a first portion of the second motion vectors determined for the first frames during the second video coding stage includes the first motion vectors stored in the memory and a second portion of the second motion vectors determined for the second frames during the second video coding stage is predicted using one or more motion vectors from the first motion vectors stored in the memory, and wherein the second subset of frames includes a greater number of frames than the first subset of frames based on the second frame rate being higher than the first frame rate;
• reconstruct, during the second video coding stage at the second frame rate, the first subset of frames using the first motion vectors determined during the first video coding stage;
• and
• reconstruct, during the second video coding stage at the second frame rate, the second subset of frames using the second motion vectors determined during the second video coding stage, wherein the first video coding stage and the second video coding stage are implemented at least partly in parallel.