Memory module threading with staggered data transfers

A memory, comprising:

• at least one memory module for coupling to a memory controller via a primary data bus;
• wherein the at least one memory module includes
• a substrate;
• buffer circuitry disposed on the substrate, the buffer circuitry including a primary interface for coupling to a primary data bus, the primary data bus to route data of a primary data transfer continuous throughput;
• plural memory device groups coupled to the buffer circuitry via respective secondary data bus paths, each of the plural memory device groups having a continuous throughput smaller than the primary data transfer continuous throughput, each of the plural memory device groups responsive to a memory request to transfer data groups collectively forming a data thread, the plural memory device groups isolated from the primary data bus by the buffer circuitry;
• and
• wherein the buffer circuitry includes steering logic to perform staggered first transfers of first data groups from a first data thread across the primary data bus, the staggered first transfer occurring before all of the first data groups are aggregated, the steering logic to perform staggered second transfers of second data groups from a second data thread across the primary data bus, the staggered second transfers occurring before all of the second data groups are aggregated.

A method of operating a memory module having a substrate, buffer circuitry disposed on the substrate for coupling to a primary data bus, the primary data bus to route data of a primary data transfer continuous throughput, the method comprising:

• receiving a memory request to transfer data groups from plural memory device groups, the data groups collectively forming a data thread:
• performing staggered first transfers of first data groups from a first data thread across the primary data bus, the staggered first transfers occurring before all of the first data groups are aggregated;
• and
• performing staggered second transfers of second data groups from a second data thread across the primary data bus, the staggered second transfers occurring before all of the second data groups are aggregated.