PARALLEL DATA PROCESSING APPARATUS

• US 20080052492A1
• Filed: 06/29/2007
• Published: 02/28/2008
• Est. Priority Date: 04/09/1999
• Status: Active Grant

• Current Assignee: Rambus Inc.
• Original Assignee: Rambus Inc.
• Inventors: Stuttard, Dave, O'Dea, Eamon, McConnell, Ray, Rhoades, John, Cameron, Ken, Winser, Paul, Faulds, Gordon, David, Russell, Greer, Trey, Atkin, Phil, Day, Tim, Williams, Dave
• Granted Patent: US 8,762,691 B2
• Time in Patent Office: Days
• US Class Current: 712/22
• CPC Class Codes: G06F 15/8007 single instruction multiple data [SIMD] multiprocessors
• CPC Class Codes: G06F 15/8015 One dimensional arrays, e.g. rings, linear arrays, buses
• CPC Class Codes: G06F 9/02 using wired connections, e.g. plugboards
• CPC Class Codes: G06F 9/3001 Arithmetic instructions
• CPC Class Codes: G06F 9/3004 to perform operations on memory
• CPC Class Codes: G06F 9/30087 Synchronisation or serialisation instructions
• CPC Class Codes: G06F 9/30101 Special purpose registers
• CPC Class Codes: G06F 9/3836 Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
• CPC Class Codes: G06F 9/3838 Dependency mechanisms, e.g. register scoreboarding
• CPC Class Codes: G06F 9/3851 from multiple instruction streams, e.g. multistreaming
• CPC Class Codes: G06F 9/3885 using a plurality of independent parallel functional units
• CPC Class Codes: G06F 9/3887 controlled by a single instruction for multiple data lanes [SIMD]
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining

Abstract

A data processing apparatus includes a plurality of processing elements arranged in a single instruction multiple data array. The apparatus includes an instruction controller operable to receive instructions from a plurality of instructions streams, and to transfer instructions from those instructions streams to the processing elements in the array, such that the data processing apparatus is operable to process a plurality of processing threads substantially in parallel with one another. A data transfer controller is provided which is operable to control transfer of data between the internal memory units associated with the processing elements, and memory external to the array.

Claims

1. A data processing apparatus comprising:

a SIMD (single instruction multiple data) array of processing elements, each of which includes an internal memory unit, the processing elements being divided into a plurality of processing blocks, which processing blocks are operable to process respective groups of data items;

an instruction controller operable to receive instructions from a plurality of instructions streams, and to transfer instructions from those instructions streams to the processing elements in the array, such that the data processing apparatus is operable to process a plurality of processing threads substantially in parallel with one another; and

a data transfer controller operable to control transfer of data between the internal memory units associated with the processing elements, and memory external to the array.

2. A data processing apparatus as claimed in claim 1, wherein the instruction controller is operable to supply instructions from at least one of the processing threads to the data transfer controller.

3. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array.

4. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to optimize data transfer operations to remove multiple accesses to an external memory location when a plurality of processing elements require access to that external memory location.

5. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to consolidate data transfer operations where external memory location information from more than one processing element relates to the same data packet location.

6. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to consolidate data transfer operations where external memory location information from more than one processing element relates to the same data packet location, each data packet location storing an amount of data greater than that required by a processing element.

7. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to compare at least a part of such external memory location information supplied by one of the processing elements with at least respective parts of external memory location information supplied by other processing elements in the array, the data transfer controller being operable to optimize data transfer operations by removing multiple accesses to an external memory location when a plurality of processing elements supply external memory location information relating to that external memory location.

8. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer.

9. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of values starting from a base address, the values being spaced apart from one another by a predetermined step size.

10. A data processing apparatus as claimed in claim 9, wherein the base address is provided by a source external to the apparatus.

11. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of consecutive values starting from a base address.

12. A data processing apparatus as claimed in claim 11, wherein the base address is provided by a source external to the apparatus.

13. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of values which are spaced apart from one another by a predetermined amount, the predetermined amount being independent of the size of data to be transferred.

14. A data processing apparatus as claimed in claim 13, wherein the base address is provided by a source external to the apparatus.

15. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, wherein the data transfer controller is operable to supply respective external memory location values to the processing elements following data transfer.

16. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer by a co-processor external to the data processing apparatus.

17. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer by a co-processor external to the data processing apparatus, and is dependent upon data to be transferred.

18. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable asynchronously with the SIMD array.

19. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to synchronise data transfer operations with processing threads.

20. A data processing apparatus as claimed in claim 1, wherein the data transfer controller is operable to synchronise data transfer operations with processing threads using semaphores.

21. A data processing apparatus as claimed in claim 1, wherein data transfer operations have respective priority levels.

22. A data processing apparatus as claimed in claim 1, wherein data transfer operations have respective priority levels, and wherein a data transfer operation having a high priority level pre-empts a data transfer operation having a lower priority level.

23. A data processing apparatus as claimed in claim 21, wherein completion of a data transfer operation having a high priority level causes a data transfer operation having a lower priority level to restart.

24. A data processing apparatus as claimed in claim 1, wherein data transfer operations have respective priority levels, and wherein the data transfer controller includes a first-in-first-out (FIFO) buffer for storing data transfer information for each priority level.

25. A data processing apparatus as claimed in claim 1, wherein data transfer operations have respective priority levels, and wherein the data transfer controller includes a first-in-first-out (FIFO) buffer for storing data transfer information for each priority level, and wherein the data transfer controller is operable to supply data transfer status information to the instruction controller.

26. A data processing apparatus comprising:

a SIMD (single instruction multiple data) array of processing elements, each of which includes an internal memory unit;

an instruction controller operable to receive instructions from a plurality of instructions streams, and to transfer instructions from those instructions streams to the processing elements in the array, such that the data processing apparatus is operable to process a plurality of processing threads substantially in parallel with one another; and

a data transfer controller operable to control transfer of data between the internal memory units associated with the processing elements, and memory external to the array.

27. A data processing apparatus as claimed in claim 26, wherein the instruction controller is operable to supply instructions from at least one of the processing threads to the data transfer controller.

28. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array.

29. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to optimize data transfer operations to remove multiple accesses to an external memory location when a plurality of processing elements require access to that external memory location.

30. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to consolidate data transfer operations where external memory location information from more than one processing element relates to the same data packet location, each data packet location storing an amount of data greater than that required by a processing element.

31. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location information supplied by each of the processing elements in the array, and wherein the data transfer controller is operable to compare at least a part of such external memory location information supplied by one of the processing elements with at least respective parts of external memory location information supplied by other processing elements in the array, the data transfer controller being operable to optimize data transfer operations by removing multiple accesses to an external memory location when a plurality of processing elements supply external memory location information relating to that external memory location.

32. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer.

33. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of values starting from a base address, the values being spaced apart from one another by a predetermined step size.

34. A data processing apparatus as claimed in claim 33, wherein the base address is provided by a source external to the apparatus.

35. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of consecutive values starting from a base address.

36. A data processing apparatus as claimed in claim 33, wherein the base address is provided by a source external to the apparatus.

37. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, such external memory location values forming a series of values which are spaced apart from one another by a predetermined amount, the predetermined amount being independent of the size of data to be transferred.

38. A data processing apparatus as claimed in claim 37, wherein the base address is provided by a source external to the apparatus.

39. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer, wherein the data transfer controller is operable to supply respective external memory location values to the processing elements following data transfer.

40. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer by a co-processor external to the data processing apparatus.

41. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to control access to memory external to the array using external memory location values, and wherein an external memory location value is calculated for each processing element requiring data transfer by a co-processor external to the data processing apparatus, and is dependent upon data to be transferred.

42. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable asynchronously with the SIMD array.

43. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to synchronise data transfer operations with processing threads.

44. A data processing apparatus as claimed in claim 26, wherein the data transfer controller is operable to synchronise data transfer operations with processing threads using semaphores.

45. A data processing apparatus as claimed in claim 26, wherein data transfer operations have respective priority levels.

46. A data processing apparatus as claimed in claim 26, wherein data transfer operations have respective priority levels, and wherein a data transfer operation having a high priority level pre-empts a data transfer operation having a lower priority level.

47. A data processing apparatus as claimed in claim 45, wherein completion of a data transfer operation having a high priority level causes a data transfer operation having a lower priority level to restart.

48. A data processing apparatus as claimed in claim 26, wherein data transfer operations have respective priority levels, and wherein the data transfer controller includes a first-in-first-out (FIFO) buffer for storing data transfer information for each priority level.

49. A data processing apparatus as claimed in claim 26, wherein data transfer operations have respective priority levels, and wherein the data transfer controller includes a first-in-first-out (FIFO) buffer for storing data transfer information for each priority level, and wherein the data transfer controller is operable to supply data transfer status information to the instruction controller.