1. The challenges of the new global scenario Paolo PRINETTO Politecnico di Torino (Italy) University of Illinois at Chicago, IL (USA) Paolo.Prinetto@polito.it Prinetto@uic.edu www.testgroup.polito.it Lecture 1.1

2. 2 I.1 Goal  The lecture is aimed at outlining where most of today designers’ constraints come from.  Thus it first presents the current trends in the microelectronic industries, with particular emphasis on System-on-a-Chip.  It then outlines the various phases of a product life cycle, focusing on Concurrent Engineering approaches.

3. 3 I.1 Homework  Students are warmly invited to visit:  web pages related to microelectronic trends, such as:  The International Technology Roadmap for Semiconductors home page at http://public.itrs.net

4. 4 I.1 Homework (cont’d)  web pages related to on-going standardization activities on System-on-a- Chip, such as:  VSI Alliance home page at http://www.vsi.org/  IEEE P1500 Standard for Embedded Core Test (SECT) home page at http://grouper.ieee.org/groups/1500/

5. 5 I.1 Further readings  No particular suggestion

6. Waves of electronic computing 196019701980199020002010 Time sharing Batch Mainframe Systems

7. Waves of electronic computing 1960197019801990 PC Word processing Web e-Commerce PC based Systems Time sharing Batch Mainframe Systems 20002010

8. Waves of electronic computing 196019701980199020002010 PC Word processing Web e-Commerce PC based Systems Time sharing Batch Mainframe Systems Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control

9. Two forces working in conjunction in electronics industry

10. Consumerization Miniaturization

11. 11 I.1 Consumerization of the microelectronic industry  High volumes

12. 12 I.1 Some facts...  50 millions of Pentium produced every year (1 Pentium every 0.6 s!!)  100 millions of PC and 3 millions of servers sell in ‘99 ... Pentium Pro Pentium ® Pro Processor ® ® R R

13. 13 I.1 Consumerization of the microelectronic industry  High volumes  High costs for design and production

14. 14 I.1 High costs for design 0 100 200 300 400 500 # Designers 400480286PentiumPentium 3

15. 15 I.1 High costs for production

16. 16 I.1 High costs for production $ 2+ US Billions !!!

17. ATE cost

18. The chip to be tested is inserted here !!

19. ATE cost $ 6+ US Millions The chip to be tested is inserted here !!

20. 20 I.1 Some facts: Pentium Intel  For End-of-Production test, Intel uses 300 ATEs for VLSI:  if aligned, they formed a 2 Km queue  they have, globally, 60,000 pin drivers  they consume, globally, 7.5 MW. [K. Thompson, Vice president Intel, ITC’95] Pentium Pro Pentium ® Pro Processor ® ® R R

21. 21 I.1 Consumerization of the microelectronic industry  High volumes  High costs for design and production  Lower prices of the final products

22. 22 I.1 Some facts... Since 1Q’97, the price of PCs decreases of 6% every quarter.

23. 23 I.1 Consumerization of the microelectronic industry  High volumes  High costs for design and production  Low prices of the final products  High dependability (reliability & availability)

24. Two forces working in conjunction in electronics industry Miniaturization

25. 25 I.1 Moore’s law Processor transistor counts doubles every two years  a new technology every 9 months !!  System-on-a-chip (SoC)

26. 26 I.1 Pentium 4  1.5 GHz  2 GHz in 3Q01  0.18   0.13  in 4Q01  42 M transistor

27. 27 I.1 Intel Press release “Researchers at Intel Corp. have built what they claim is the smallest and fastest CMOS transistor, measuring 30nm in size and three atom thick. The gate oxides used to build these transistors are three atomic layers thick. Intel believes this new development will allow the company within the next five to 10 years to build microprocessors containing more than 400 million transistors, running at 10GHz and operating at less than 1V. To rationalize this scale, Intel said more than 100,000 of these gates would need to be stacked to achieve the thickness of a sheet of paper and that they could compute 2 million calculations in the time it takes a bullet to travel one inch.”

28. 28 I.1 Very deep sub-micron ICs

29. 29 I.1 System - chip

30. 30 I.1 System-on-chip

31. 31 I.1 SoCs trend t

32. 32 I.1 SoCs trend Yesterday’s PCBs (System-on-a- board) t

33. 33 I.1 SoCs trend Today’s chips (System-on-a-chip) t Yesterday’s PCBs (System-on-a- board)

34. 34 I.1 SoCs trend Tomorrow’s re-usable IPs Today’s chips (System-on-a-chip) t Yesterday’s PCBs (System-on-a- board)

35. 35 I.1 What are embedded cores Pre-designed, pre-verified functional blocks, also termed IP (Intellectual Property), or macro.

36. 36 I.1 Characteristics of Core-based SoCs  Reusable cores replacing existing COTS ICs  Core representation at different hardware description level (soft, firm, hard)  Embedded cores available from diverse sources  Design efficiency achieved by ease of plug & play (a lot of standardization activities are currently going on).

37. 37 I.1 SoC market growth 19989.0 199911.0 200014.5 200119.0 200224.0 200332.0 Worldwide revenue (B of $) [Dataquest]

38. 38 I.1 Example: A mixed signal modem (board implementation) RAM ROM Micro- controller ROM DSP CoDec Transmission medium Data terminal

39. 39 I.1 Transmission medium Data terminal DSP core RAM ROM Interface logic CoDec Example: A mixed signal model (SoC implementation)

40. Target Applications 196019701980199020002010 PC Word processing Web e-Commerce Time sharing Batch Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control

41. Target Applications 196019701980199020002010 PC Word processing Web e-Commerce Time sharing Batch Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control ATM switches, Bridges, Modems, Routers, …

42. Target Applications 196019701980199020002010 PC Word processing Web e-Commerce Time sharing Batch Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control ATM switches, Bridges, Modems, Routers, … Cellular phones, PDAs, Pagers, Organizers, …

43. Target Applications 196019701980199020002010 PC Word processing Web e-Commerce Time sharing Batch Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control ATM switches, Bridges, Modems, Routers, … Cellular phones, PDAs, Pagers, Organizers, … Digital Cameras, Games, …

44. Target Applications 196019701980199020002010 PC Word processing Web e-Commerce Time sharing Batch Embedded Systems Telecommunication & Network Portable Consumer Products Multimedia Embedded Control ATM switches, Bridges, Modems, Routers, … Cellular phones, PDAs, Pagers, Organizers, … Digital Cameras, Games, … Automotive, Smart cards, Printers, …

45. 45 I.1 What are the consequences ? What are the consequences ?

46. 46 I.1 Just few reminds...  Product life cycle  Time to Market  Time to Volume  Time to Money  Market Window

47. 47 I.1 t The product life-cycle

48. 48 I.1 t The product life-cycle User’s requirements

49. 49 I.1 t The product life-cycle User’s requirements Design

50. 50 I.1 t The product life-cycle User’s requirements Production Design

51. 51 I.1 t The product life-cycle User’s requirements Production Design In-field operation

52. 52 I.1 t The product life-cycle User’s requirements Production Design In-field operation Re-cycling

53. 53 I.1 t Time-To-Market ( TTM ) User’s requirements Design Production

54. Time to Market TTM Time to Volume TTV Time to Money TT$ Revenues Costs Profits

55. 55 I.1 Smaller market windows Which are the What are the consequences ?

56. 56 I.1 Smaller market windows 1980 5 y

57. 57 I.1 Smaller market windows 1980 5 y 1 y 1990

58. 58 I.1 Smaller market windows 1980 5 y 1 y 1990 < 6 m 2000

59. 59 I.1 Dramatic shrink of products life-cycle

60. 60 I.1 What should we do? What should we do?

61. 61 I.1 To survive, you must be able to meet the trends.

62. 62 I.1 The real challenge: To meet user quality requirements for your new products

63. 63 I.1 The set of properties and characteristics of a product (a service) capable of guaranteeing customer’s satisfaction. Quality ISO 8402

64. 64 I.1 The quality of any product or service is what the customer says it is.

65. 65 I.1 Reduce your Time-To-Market

66. 66 I.1 The impact of delay on profits Let’s consider a product characterized by:  market grow rate : 20%  annual price reduction rate : 12%  product life-cycle: 5 y [McKinsey & Co]

67. 67 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00%

68. 68 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% Increase of component cost: 50%

69. 69 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% -3.5% Increase of component cost: 50%

70. 70 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% -3.5% Increase of component cost: 50% Increase of component cost: 90%

71. 71 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% -3.5% -22% Increase of component cost: 50% Increase of component cost: 90%

72. 72 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% -3.5% -22% Increase of component cost: 50% Increase of component cost: 90% 6 months of delay in TTM

73. 73 I.1 Profit loss -50.00% -40.00% -30.00% -20.00% -10.00% 0.00% -3.5% -22% -33% Increase of component cost: 50% Increase of component cost: 90% 6 months of delay in TTM

74. 74 I.1 How can we reduce our TTM ? How can we reduce our TTM ?

75. 75 I.1 Set up Distributed Cooperative Working Environments to support Concurrent & Distributed Engineering

76. 76 I.1 [ Institute for Defense Analyses, USA ] Concurrent Engineering is a systematic approach to the integrated, concurrent design of products and their related processes, including manufacture and support.

77. 77 I.1 [ Institute for Defense Analyses, USA ] Concurrent Engineering is intended to cause the developers, from the outset, to consider all the elements of the product life cycle from conception through disposal, including quality, cost, schedule, and user requirements. [IEEE Spectrum, July 1991]

78. 78 I.1 Redo until right Concurrent Engineering

79. 79 I.1 Redo until right First time right Concurrent Engineering

80. 80 I.1 Redo until right First time right Concurrent Engineering Teamwork

81. 81 I.1 Redo until right First time right Concurrent Engineering Teamwork Teamwork (TigerTeam)

82. 82 I.1 TigerTeam People from many departments collaborate over the life of a product to ensure that it reflects customers’ needs and desires.

83. 83 I.1 Concurrent Engineering Manufacturability Diagnosability Reliability Integrability Testability Verifiability Re-cyclability Design for X-ability

84. 84 I.1 Some examples...  Boeing Corporation Airplane Group: Boeing 777 (440 passengers) developed in 1.5 years less than 767:  2200 terminals  8 mainframes IBM 3090-600J  connected Seattle, Renton, Everett (DC), Wichita (KA), Philadelphia (PA) & partners like Mitsubishi, Kawasaki, Fuji

85. 85 I.1 Some examples...  FIAT: 300 people to design and set up the PUNTO production line  IBM, Apple, Motorola: PowerPC 601: 12 months from agreement signature to the first chip.

86. 86 I.1 Reduce your design time

87. 87 I.1 Improve designers productivity

88. 88 I.1 Exploit chip complexity

89. 89 I.1 System-on-chip

90. 90 I.1 Our next product must be better, smaller, cheaper, faster, consume less power, & be ready for yesterday !!