For delivering an audio and/or video content using adaptive streaming from a cache server to a client device, the audio and/or video content being segmented into data segments available in various representations with respective audio and/or video qualities, the representations being time-aligned per segment, the segments being further divided into chunks, a method comprises: obtaining a round-trip time value applicable between the cache server and the client device; computing a minimum bulk transfer size from the maximum average bitrate of the various representations of the audio and/or video content and from the obtained round-trip time value; computing a minimum bulk duration, with respect to the audio and/or video content, for the representation to be delivered to the client device; and delivering the audio and/or video content by burst transfers in the form of bulks of aggregated successive chunks for each segment in accordance with the computed minimum bulk duration.