Reconfigurable elements

• US 8,686,549 B2
• Filed: 09/30/2009
• Issued: 04/01/2014
• Est. Priority Date: 09/03/2001
• Status: Active Grant

• Current Assignee: PACT XPP Schweiz AG
• Original Assignee: Pact XPP Technologies, AG
• Inventors: Vorbach, Martin
• Primary Examiner(s): NGUYEN, CUONG QUANG
• Application Number:
  US12/571,173
• Publication Number:
  US 20100095088A1
• Time in Patent Office: 1,644 Days
• Field of Search: 257678-697, 257777-796, 257/E23.085
• US Class Current: 257/686
• CPC Class Codes: G06F 15/17375 One dimensional, e.g. linear array, ring
• CPC Class Codes: G06F 15/7842 on one IC chip (single chip microcontrollers)
• CPC Class Codes: G06F 9/30036 Instructions to perform operations on packed data, e.g. vector, tile or matrix operations

Abstract

A cell element field for data processing having function cells for execution of algebraic and/or logic functions and memory cells for receiving, storing and/or outputting information is described. A control connection may lead from the function cells to the memory cells.

Claims

1. A data processing chip comprising:

a plurality of dice, a first of the dice comprising a plurality of processor units and at least one second of the dice comprising at least one data storage memory device, the first and second dice being arranged such that, for each of at least a subset of said processor units provided in a first layer, a dedicated memory storage device is provided in a second layer.

2. The processing chip according to claim 1, wherein the processor units are fully programmable.

3. The processing chip according to claim 1, wherein the processor units are reduced instruction set computer (RISC) processors.

4. The processing chip according to claim 1, wherein the first and at least one second dice are connected by through-substrate vias.

5. The processing chip according to claim 1, wherein the first and at least one second dice are connected by vias.

6. The processing chip according to claim 1, wherein the first and second dice are connected by wire bonds.

7. The processing chip according to claim 1, wherein the first and second dice are connected with flip-chip technology.

8. The processing chip according to claim 1, wherein the first and second dice are connected with bumps.

9. The processing chip according to any one of claims 1, 4, 5, 6, 7, and 8, wherein said at least one data storage memory device is a dynamic random access memory (DRAM) memory element.

10. The processing chip according to one of claims 1, 4, 5, 6, 7, and 8, wherein said at least one data storage memory device includes a cache.

11. The processing chip according to claim 1, wherein some of the processor units are defective and switched off.

12. The processing chip according to claim 1, wherein the processor units include redundant processor units.

13. The processing chip according to claim 1, wherein the at least one data storage memory device has redundant structures.

14. The processing chip according to claim 1, wherein the processor units have an additional physical layer for providing an interconnection arrangement.

15. A system comprising:

a plurality of chips in a stacked arrangement, at least one of the chips having a power dissipation higher than a power dissipation of at least one other of the chips, wherein said chips are stacked such that said at least one of the chips is arranged at an outer layer of the stack arrangement.

16. The system according to claim 15, wherein the plurality of chips include processor units that are fully programmable.

17. The system according to claim 15, wherein the plurality of chips include reduced instruction set computer (RISC) processors.

18. The system according to claim 15, wherein a first of the plurality of chips and a second of the plurality of chips are connected by through-substrate vias.

19. The system according to claim 15, wherein a first of the plurality of chips and a second of the plurality of chips are connected by vias.

20. The system according to claim 15, wherein a first of the plurality of chips and a second of the plurality of chips are connected by wire bonds.

21. The system according to claim 15, wherein a first of the plurality of chips and a second of the plurality of chips are connected with flip-chip technology.

22. The system according to claim 15, wherein a first of the plurality of chips and a second of the plurality of chips are connected with bumps.

23. The system according to one of claims 15, 18, 19, 20, 21, and 22, wherein said plurality of chips include dynamic random access memory (DRAM) memory elements.

24. The system according to one of claims 15, 18, 19, 20, 21, and 22, wherein said plurality of chips include cache memory elements.

25. The system according to claim 15, wherein the plurality of chips include processor units of which some (a) are defective and (b) are switched off.

26. The system according to claim 15, wherein the plurality of chips include redundant processor units.

27. The system according to claim 15, wherein the plurality of chips include data storage memories having redundant structures.

28. The system according to claim 15, wherein the plurality of chips include:

a first layer including a processor;

a second layer including a memory device; and

a third layer for providing an interconnection arrangement.

29. A multi-processor device comprising

a plurality of processor cores;

a plurality of memory elements; and

an interconnection network for interconnecting the memory elements and processor cores;

wherein;

the memory elements are operable as cache;

a given memory element can be connected to a processor core selectable and changeable at runtime via the interconnection network; and

the memory elements have a port for connecting to higher level memory so that data can be moved, via said port, into external memory out of said memory elements and from said external memory into said memory elements.

30. The multi-processor device according to claim 29, wherein the plurality of processor cores are separated from the plurality of memory elements.

31. The multi-processor device according to claim 29, wherein the plurality of processor cores are arranged in a field and at least some of the processor cores have access to a data transmission network for transferring data over long distances within the field.

32. The multi-processor device according to claim 29, wherein the multi-processor device is adapted such that a plurality of combinations of memory elements of the plurality of memory elements and processor cores of the plurality of processor cores can be simultaneously formed at run time.

33. The multi-processor device according to claim 29, wherein at least some of the memory elements have multiple ports for connecting to multiple busses of the interconnection network.

34. The multi-processor device according to claim 33, wherein at least some of the memory elements are simultaneously accessible at runtime by multiple processor cores of the plurality of processor cores via multiple busses of the interconnection network.

35. The multi-processor device according to claim 29, wherein the plurality of processor cores are separated from the plurality of memory elements.

36. A multi-core processor device comprising:

multiple processor cores; and

an interconnection for communication of data;

wherein;

a plurality of said multiple processor cores is implemented as a first physical layer on a die;

the interconnection structure is implemented as a separate physical layer on a second die; and

the first and second dice are arranged in a stacked manner.

37. The multi-core processor device according to claim 36, wherein an interconnect for vertical interconnection is provided.

38. The multi-core processor device according to claim 36, wherein the first and second dice are connected by through-substrate vias.

39. A processor integrated device, comprising:

at least two dies that are stacked;

an interconnect structure; and

an arrangement of programmable data processing units interconnected by the interconnect structure;

wherein;

at least some of the programmable data processing units include Arithmetic Logic Units (ALUs);

the interconnect structure includes switches;

the programmable data processing units are implemented on at least a first one of the at least two dies; and

at least parts of the interconnect structure are implemented on at least a second one of the at least two dies.

40. The processor integrated device according to claim 39, further comprising at least one dedicated interface unit that interfaces between the programmable data processing units and an external device.

41. The processor integrated device according to claim 40, further comprising at least one dedicated Peripheral Component Interconnect (PCI) unit.

42. The processor integrated device according to claim 40, further comprising at least one dedicated Ethernet unit.

43. The processor integrated device according to claim 39, wherein at least some of the ALUs are capable of splitting operands into a respective set of data, data of which is processable independently of each other and splitting a result into a respective set of data, data of which is processable independently of each other.

44. The processor integrated device according to claim 43, wherein the interconnect structure is capable of transmitting each data part of each of the sets of data separately.

45. The processor integrated device according to claim 39, wherein at least some of the ALUs are capable of single instruction multiple data (SIMD) processing.

46. The processor integrated device according to claim 45, wherein the interconnect structure is capable of transmitting each data part of each of the sets of data separately.

47. The processor integrated device according to claim 39, wherein at least some switches of the interconnect structure are implemented on the second die.

48. The processor integrated device according to claim 39, wherein at least some of the dies are interconnected by vias.

49. The processor integrated device according to claim 48, wherein the vias extend through a silicon layer.

50. The processor integrated device according to claim 39, further comprising at least one third die that includes memory units coupled to at least some of the programmable data processing units.

51. The processor integrated device according to claim 39, further comprising at least one third die that includes memory units to which respective ones of programmable data processing units have exclusive and direct connections.

52. The processor integrated device according to claim 39, further comprising at least one third die that includes memory units, wherein one of the first and third dies is stacked on top of the other of the first and third dies.