Programmable multi-tasking memory management system

• US 6,526,462 B1
• Filed: 11/19/1999
• Issued: 02/25/2003
• Est. Priority Date: 11/19/1999
• Status: Expired due to Fees

• Current Assignee: Realchip, Inc.
• Original Assignee: Realchip, Inc.
• Inventors: Elabd, Hammam
• Primary Examiner(s): Myers, Paul R.
• Application Number:
  US09/443,934
• Time in Patent Office: 1,194 Days
• Field of Search: 710/242, 710/243, 710/308, 710/309, 710/119-125, 710/22, 710/27
• US Class Current: 710/242
• CPC Class Codes: G06F 9/5016 the resource being the memory

Abstract

A highly intelligent programmable multi-tasking memory management system manages memory requests associated with a system on chip (SOC) device. The memory management system includes a routing controller or central processing unit (RCPU) that is used for routing/switching stream data between communication cores and digital signal processors with minimum reliance and demand on a main or virtual central processing unit (VCPU) residing on a system bus. Tasks are partitioned between the VCPU and the RCPU within the SOC architecture for communication applications. The VCPU performs system/application tasks while the RCPU simultaneously performs multiple memory routing/switching tasks and multiple concurrent memory access connections. The memory management system also enables other processors and communication cores to update their internal data once new data is written in the memory system. In addition, a method and system are provided for performing predictive protocol fetch for multiple DSPs on the SOC to increase data processing throughput.

Claims

1. A method for performing multiple tasks simultaneously in a System On a Chip (SOC) device, comprising:

processing system and application tasks using a first processor, the first processor being connected to a system bus and being adapted to receive system and applications requests from a plurality of communication cores; and

processing memory management tasks using a memory management system having a second processor while processing system and application tasks are being performed, the memory management system being connected to the system bus and to the plurality of communications cores through dedicated memory buses,

wherein the second processor is adapted to perform at least one of memory routing and switching tasks among the plurality of communication cores.

2. The method according to claim 1, wherein processing the memory management tasks is software programmable.

3. The method according to claim 1, wherein the second processor is adapted to perform at least one of memory routing and switching tasks for multiple memory requests simultaneously.

4. The method according to claim 1 further comprising transmitting updated memory data from the second processor to the first processor.

5. The method according to claim 1, wherein processing memory management tasks includes:

inputting memory requests from the plurality of communication cores to a request buffer;

outputting the memory requests from the request buffer to a foreground request schedule and a background request schedule;

queuing the memory requests in the foreground request schedule followed by queuing the memory requests in the background request schedule; and

processing simultaneously a predetermined number of memory requests.

6. The method according to claim 5, wherein the predetermined number of memory requests corresponds to the number of memory controllers in the memory management system.

7. The method according to claim 5, wherein the foreground requests correspond to requests residing in an SRAM buffer.

8. A memory management system for routing or switching stream data among a plurality of peripheral masters on a System on a Chip (SOC) device, the system comprising:

a multiplexer for selecting a predetermined number of peripheral memory requests from the plurality of peripheral masters;

a multi tasking arbiter and load balancing system coupled to the multiplexer and adapted to provide an arbitration, load balancing and scheduling scheme to the multiplexer;

a bus bridge for receiving a system request from a system master; and

a routing CPU coupled to the multiplexer and the bus bridge for processing both the peripheral memory requests and the system request.

9. A system according to claim 8, wherein the arbitration, load balancing and scheduling scheme comprises one of a round robin, psuedo-random, priority, and fixed bandwidth assignment per master.

10. A System On a Chip (SOC) device comprising:

a virtual master for processing system and application tasks requested by a plurality of peripheral masters, the plurality of peripheral masters adapted to transmit memory requests associated with one or more memory banks; and

a memory management system having a routing master for receiving and processing the memory requests from the plurality of peripheral masters.

11. The SOC device according to claim 10, wherein the memory management system is software programmable.

12. The SOC device according to claim 11, wherein the memory management system is programmed to receive and process memory requests from a first predetermined number of peripheral masters.

13. The SOC device according to claim 12, wherein the memory management system is reprogrammed to receive and process memory requests from a second predetermined number of peripheral masters different from the first predetermined number of peripheral masters.

14. The SOC device according to claim 10, wherein the memory management system receives the memory requests from a plurality of dedicated memory buses.

15. The SOC device according to claim 14, wherein the plurality of dedicated memory buses connect the plurality of peripheral masters to-the routing master.

16. The SOC device according to claim 14, wherein the plurality of dedicated memory buses comprise non-arbitrated buses.

17. The SOC device according to claim 14, wherein the plurality of dedicated memory buses are used to decode memory addresses.