Processor arrangement on a chip including data processing, memory, and interface elements

• US 9,037,807 B2
• Filed: 11/11/2010
• Issued: 05/19/2015
• Est. Priority Date: 03/05/2001
• Status: Expired due to Term

• Current Assignee: PACT XPP Schweiz AG
• Original Assignee: Pact XPP Technologies, AG
• Inventors: Vorbach, Martin
• Primary Examiner(s): NGUYEN, THAN VINH
• Application Number:
  US12/944,068
• Publication Number:
  US 20110060942A1
• Time in Patent Office: 1,650 Days
• Field of Search: 711/147, 711/148
• US Class Current: 711/148
• CPC Class Codes: G06F 11/1405 at machine instruction level
• CPC Class Codes: G06F 11/20 using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
• CPC Class Codes: G06F 12/00 Accessing, addressing or allocating within memory systems or architectures digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06
• CPC Class Codes: G06F 13/1663 Access to shared memory
• CPC Class Codes: G06F 13/40 Bus structure for computer networks G06F15/163; for optical bus networks H04B10/25
• CPC Class Codes: G06F 15/7867 with reconfigurable architecture
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: H03K 19/17756 for partial configuration or partial reconfiguration

Abstract

At the inputs and/or outputs, memories are assigned to a reconfigurable module to achieve decoupling of internal data processing and in particular decoupling of the reconfiguration cycles from the external data streams (to/from peripherals, memories, etc.).

Claims

1. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element;

wherein the bus system is adapted for dynamically establishing and releasing transmission channels between a sending one of the elements and a receiving one of the elements; and

wherein the bus system is adapted for forming at least one ring via interconnection elements that include pipeline-registers.

2. The multi-processor system according to claim 1, wherein the ring is adapted for transmitting data in two directions using separate respective interconnection elements for each direction.

3. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element;

wherein the bus system is adapted for dynamically establishing and releasing transmission channels between a sending one of the elements and a receiving one of the elements; and

wherein an identification token is assigned to transmitted data for indicating an affiliation of the transmitted data to a process.

4. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element;

wherein the bus system is adapted for dynamically establishing and releasing transmission channels between a sending one of the elements and a receiving one of the elements; and

wherein an identification token is assigned to data transferred via the interface element for indicating an affiliation of the transferred data to a process.

5. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element, and

wherein the bus system is adapted for forming at least one ring via interconnection elements that include pipeline registers.

6. The multi-processor system according to claim 5, wherein the ring is adapted for transmitting data in two directions using separate respective interconnection elements for each direction.

7. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element; and

wherein an identification token is assigned to transmitted data for indicating an affiliation of the transmitted data to a process.

8. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element; and

wherein an identification token is assigned to data transferred via the interface element for indicating an affiliation of the transferred data to a process.

9. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements, that each independently operates as cache for caching data;

at least one interface element for providing a connection to a common higher level memory; and

at least one global interconnection system adapted for interconnecting the plurality of data processing elements, the plurality of memory elements, and the at least one interface element;

wherein;

the at least one global interconnection system is accessible by each of the data processing elements and memory elements via a respective dedicated bus interface node, and permits flexible allocation and interconnection during run time;

transmitters and receivers of the data processing, memory and interface elements are adapted for being connected via the global interconnection system in an addressed manner to permit individual data transfer; and

each of the data processing elements, memory elements, and at least one interface element are separately connected to the global interconnection system for transferring data between (i) at least one of the data processing elements and at least one of the memory elements, and (ii) at least one of the memory elements and the at least one interface element; and

further wherein at least one interface element is adapted to connect to an external peripheral and the global interconnect system is adapted to transfer data to and from said external peripheral.

10. The multi-processor system according to claim 9, wherein data transfer on the global interconnection system is pipelined using registers implemented in, or assigned to, the global interconnection system.

11. The multi-processor system according to claim 9, wherein the global interconnection system programmably establishes a communication channel between a sender and a receiver.

12. The multi-processor system according to claim 9, wherein the global interconnection system programmably establishes a communication channel between (i) at least one of the data processing and memory elements and (ii) at least one of the at least one interface unit.

13. The multi-processor system according to claim 9, wherein the global interconnection system is adapted for being dynamically allocated and interconnected during operation.

14. The multi-processor system according to claim 10, wherein the chip is attached to a light transmitting substrate.

15. The multi-processor system according to claim 14, wherein data is transmitted by light via the light transmitting substrate.

16. The multi-processor system according to claim 14, wherein data is transmitted between two or more of any of the data processing and memory elements by light via the light transmitting substrate.

17. The multi-processor system according to claim 14, wherein data is transmitted between (i) at least one of the data processing and memory elements and (ii) the at least one interface by light via the light transmitting substrate.

18. The multi-processor system according to claim 10, wherein the chip comprises a light transmitting substrate.

19. The multi-processor system according to claim 18, wherein at least some of the data of the global interconnection is transferred by light via the light transmitting substrate.

20. The multi-processor system according to claim 10, wherein the data processing elements are separate from the memory elements, each being adapted for being separately addressed on the global interconnection system.

21. The multi-processor system according to claim 9, wherein the data transfer is via the global interconnection system and managed by control signals transmitted via the global interconnection system.

22. The multi-processor system according to claim 9, wherein:

an address is transferred from one of the plurality of data processing elements to the plurality of memory elements; and

the transferred address corresponds to a memory location of one of the plurality of memory elements which transfers data based on the correspondence to the transferred address.

23. The multi-processor system according to claim 9, wherein an address is transferred via the global interconnection system and from one of the plurality of data processing elements to one of the plurality of memory elements, which responsively transfers data conditional upon its inclusion of a memory location that corresponds to the transferred address.

24. The multi-processor system according to claim 9, wherein:

an address is transferred from one of the plurality of data processing elements to a first one of the plurality of memory elements, the address not corresponding to a memory location included in the first memory element;

the first memory element forwards the address to a second one of the plurality of memory elements based on the lack of correspondence of its included memory locations to the address; and

responsive to receipt of the address, the second memory element transfers data conditional upon its inclusion of a memory location that corresponds to the address.

25. The multi-processor system according to claim 9, wherein the multi-processor system supports preloading of data into the memory elements prior to processing of the data.

26. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements, that each independently operates as cache for caching data;

at least one interface element for providing a connection to a common higher level memory; and

at least one global interconnection system adapted for interconnecting the plurality of data processing elements, the plurality of memory elements, and the at least one interface element;

wherein;

the at least one global interconnection system is accessible by each of the data processing elements and memory elements via a respective dedicated bus interface node, and permits

flexible allocation and interconnection during run time;

transmitters and receivers of the elements are adapted for being connected via the global interconnection in an addressed manner to permit individual data transfer; and

each of the data processing elements, memory elements, and at least one interface element are connected to the global interconnection system for transferring data between (i) at least one of the data processing elements and at least one of the memory elements, and (ii) at least one of the memory elements and the at least one interface element; and

wherein an identification token is assigned to transmitted data for indicating its affiliation of the transmitted data to a process.

27. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements, that each independently operates as cache for caching data;

at least one interface element for providing a connection to a common higher level memory; and

at least one global interconnection system adapted for interconnecting the plurality of data processing elements, the plurality of memory elements, and the at least one interface element;

wherein;

the at least one global interconnection system is accessible by each of the data processing elements and memory elements via a respective dedicated bus interface node, and permits

flexible allocation and interconnection during run time;

transmitters and receivers of the elements are adapted for being connected via the global interconnection in an addressed manner to permit individual data transfer; and

each of the data processing elements, memory elements, and at least one interface element are connected to the global interconnection system for transferring data between (i) at least one of the data processing elements and at least one of the memory elements, and (ii) at least one of the memory elements and the at least one interface element; and

wherein an identification token is assigned to data transferred via the interface element for indicating an affiliation of the transferred data to a process.

28. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element;

wherein the bus system is adapted for dynamically establishing and releasing transmission channels between a sending one of the elements and a receiving one of the elements;

wherein the data transmission is via the bus system and managed by control signals transmitted via the bus;

wherein the bus system is adapted for transmitting data in at least two directions and comprises separate and independent structures for each of the data transmission directions; and

wherein the bus system is adapted for forming at least one ring via interconnection elements that include pipeline-registers.

29. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element;

wherein the bus system is adapted for transmitting data in at least two directions and comprises separate and independent structures for each of the data transmission directions; and

wherein the bus system is adapted for forming at least one ring via interconnection elements that include pipeline-registers.

30. A multi-processor system on a chip, comprising:

a plurality of data processing elements that each includes at least one arithmetic-logic unit (ALU) and a plurality of registers adapted for storing data;

a plurality of memory elements that each independently operates as a cache for caching data; and

at least one interface element for providing a connection to a common higher level memory;

wherein each of the data processing elements, each of the memory elements, and each of the at least one interface element are interconnected via a bus system for transferring data at least between (i) at least one of the data processing elements and at least one of the memory elements and (ii) at least one of the memory elements and the at least one interface element; and

wherein each of the data processing elements and each of the memory elements are separately connected to the bus system; and

further wherein at least one interface element is adapted to connect to an external peripheral and the bus system is adapted to transfer data to and from said external peripheral.

31. The multi-processor system according to claim 30, wherein the bus system is pipelined.

32. The multi-processor system according to claim 30, wherein the bus system is adapted for simultaneously transferring data by a plurality of connects.

33. The multi-processor system according to claim 30, wherein the bus system programmably establishes a communication channel between a sender and a receiver.

34. The multi-processor system according to claim 30, wherein the bus system programmably establishes a communication channel between (i) at least one of the data processing and memory elements and (ii) at least one of the at least one interface unit.

35. The multi-processor system according to claim 30, wherein the bus system is adapted for being dynamically allocated and interconnected during operation.

36. The multi-processor system according to claim 30, wherein the chip is attached to a light transmitting substrate.

37. The multi-processor system according to claim 36, wherein data is transmitted by light via the light transmitting substrate.

38. The multi-processor system according to claim 37, wherein

data is transmitted between two or more of any of the data processing and memory elements by light via the light transmitting substrate.

39. The multi-processor system according to claim 37, wherein data is transmitted between

(i) at least one of the data processing and memory elements and

(ii) the at least one interface by light via the light transmitting substrate.

40. The multi-processor system according to claim 30, wherein the chip comprises a light transmitting substrate.

41. The multi-processor system according to claim 40, wherein at least some of the data of the bus is transferred by light via the light transmitting substrate.

42. The multi-processor system according to claim 30, wherein the bus system is adapted for dynamically establishing and releasing transmission channels between a sending one of the elements and a receiving one of the elements.

43. The multi-processor system according to claim 42, wherein the bus system is pipelined.

44. The multi-processor system according to claim 42, wherein the bus system is adapted for simultaneously transferring data by a plurality of connects.

45. The multi-processor system according to claim 42, wherein the bus system programmably establishes a communication channel between a sender and a receiver.

46. The multi-processor system according to claim 42, wherein the bus system programmably establishes a communication channel between (i) at least one of the data processing and memory elements and (ii) at least one of the at least one interface unit.

47. The multi-processor system according to claim 42, wherein the bus system is adapted for being dynamically allocated and interconnected during operation.

48. The multi-processor system according to claim 42, wherein the chip is attached to a light transmitting substrate.

49. The multi-processor system according to claim 48, wherein data is transmitted by light via the light transmitting substrate.

50. The multi-processor system according to claim 49, wherein data is transmitted between two or more of any of the data processing and memory elements by light via the light transmitting substrate.

51. The multi-processor system according to claim 49, wherein data is transmitted between

(i) the data processing and memory elements and

(ii) the at least one interface by light via the light transmitting substrate.

52. The multi-processor system according to claim 42, wherein the chip comprises a light transmitting substrate.

53. The multi-processor system according to claim 52 wherein at least some of the data of the bus is transferred by light via the light transmitting substrate.

54. The multi-processor system according to claim 42, wherein the data processing elements are separate from the memory elements, each being adapted for being separately addressed on the bus system.

55. The multi-processor system according to claim 42, wherein the data transmission is via the bus system and managed by control signals transmitted via the bus.

56. The multi-processor system according to claim 55, wherein the bus system is adapted for transmitting data in at least two directions and comprises separate and independent structures for each of the data transmission directions.

57. The multi-processor system according to claim 42, wherein:

an address is transferred from one of the plurality of data processing elements to the plurality of memory elements; and

the transferred address corresponds to a memory location of one of the plurality of memory elements which transfers data based on the correspondence to the transferred address.

58. The multi-processor system according to claim 42, wherein an address is transferred via the bus system and from one of the plurality of data processing elements to one of the plurality of memory elements, which responsively transfers data conditional upon its inclusion of a memory location that corresponds to the transferred address.

59. The multi-processor system according to claim 42, wherein:

an address is transferred from one of the data processing elements to a first one of the plurality of memory elements, there being a lack of correspondence of the address to each of a plurality of memory locations included in the first memory element;

the first memory element forwards the address to a second one of the plurality of memory elements based on the lack of correspondence of its included memory locations to the address; and

responsive to receipt of the address, the second memory element transfers data conditional upon its inclusion of a memory location that corresponds to the address.

60. The multi-processor system according to claim 34, wherein pipeline stages are located between the memory elements.

61. The multi-processor system according to claim 34, wherein separated bus systems are used for transmitting addresses and write data and for returning read data.

62. The multi-processor system according to claim 30, wherein the multiprocessor system is adapted for an address to be sent from one of the plurality of data processing elements and to be received by more than one of the plurality of memory elements, and for selection of one of the more than one of the plurality of memory elements for data transmission based on a match of an address that includes data within an address range of the one of the more than one of the plurality of memory elements.

63. The multi-processor system according to claim 30, wherein the multiprocessor system is adapted for an address to be sent from one of the plurality of data processing elements and to be received by more than one of the plurality of memory elements, and for selection of one of the more than one of the plurality of memory elements for data transmission based on a match of an address within an address range of the one of the more than one of the plurality of memory elements.

64. The multi-processor system according to claim 30, wherein the interconnection via the bus system for the transferring of data between the data processing elements and the memory elements is defined by a transmitted address.

65. The multi-processor system according to claim 9 or 30, wherein the memory elements are connected to the data processing elements via switches.

66. The multi-processor system according to claim 9 or 30, wherein the memory elements are connected to the data processing elements via multiplexors.

67. The multi-processor system according to claim 9 or 30, wherein the memory elements are connected to the data processing elements via routers.

68. The multi-processor system according to claim 30, wherein the bus system runs a multi-layered protocol.

69. The multi-processor system according to claim 30, wherein the data processing elements are separate from the memory elements, each being adapted for being separately addressed on the bus system.

70. The multi-processor system according to claim 30, wherein the data transmission is via the bus system and managed by control signals transmitted via the bus.

71. The multi-processor system according to any one of claims 30 and 70, wherein the bus system is adapted for transmitting data in at least two directions and comprises separate and independent structures for each of the data transmission directions.

72. The multi-processor system according to claim 30, wherein:

an address is transferred from one of the plurality of data processing elements to the plurality of memory elements; and

the transferred address corresponds to a memory location of one of the plurality of memory elements which transfers data based on the correspondence to the transferred address.

73. The multi-processor system according to claim 30, wherein an address is transferred via the bus system and from one of the plurality of data processing elements to one of the plurality of memory elements, which responsively transfers data conditional upon its inclusion of a memory location that corresponds to the transferred address.

74. The multi-processor system according to claim 30, wherein:

an address is transferred from one of the data processing elements to a first one of the plurality of memory elements, the address not corresponding to a memory location included in the first memory element;

the first memory element forwards the address to a second one of the plurality of memory elements based on the lack of correspondence of its included memory locations to the address; and

responsive to receipt of the address, the second memory element transfers data conditional upon its inclusion of a memory location that corresponds to the address.

75. The multi-processor system according to claim 30, wherein the processor arrangement supports preloading of data into the memory elements prior to processing of the data.