Two-Layer Crystal Square for Transistors on a Single Chip

Jan Zizka et al. (Eds) : CCSEIT, MoWiN, IT, AIAP, ICBB - 2015
pp. 01–08, 2015. © CS & IT-CSCP 2015 DOI : 10.5121/csit.2015.51101
TWO-LAYER CRYSTAL SQUARE FOR
TRANSISTORS ON A SINGLE CHIP
Haissam El-Aawar
Associate Professor, Computer Science/Information Technology Departments
Lebanese International University – LIU
Bekaa-Lebanon
haisso@yahoo.com, haissam.aawar@liu.edu.lb
ABSTRACT
The number of transistors on a chip plays the main role in increasing the speed and performance of a
microprocessor; more transistors, more speed. Increasing the number of transistors will be limited due to
the design complexity and density of transistors. This article aims to introduce a new approach to
increasing the number of transistors on a chip. The basic idea is to construct two-layer crystal square for
transistors; this allows to increase the number of transistors two additional times (four times as many) if
the number of transistors incorporated in a one layer of crystal square will approximately double every 24
months according to Moore’s Law without changing rapidly the design complexity and density in a crystal
square and without changing the size of a chip (length and width), in this case the height of a chip must be
changed for the two layers.
KEYWORDS
Crystal square, Number of transistors, Feature size, Die size, Design complexity, Density of transistors.
1. INTRODUCTION
Different types of a microprocessor, is also known as a central processing unit (CPU), such as
i80x86, Pentium (Intel), PowerPC (IBM), SPARC (Sun), Motorola and many other types perform
the same thing in approximately the same way [1].
Moore’s Law states that the number of transistors on a chip would double every two years [2, 3].
The designers of CPU are interested to increase the speed and performance of CPU with each
new CPU-generation according to the Moore’s law by doubling the number of transistors
according to the design rule by reducing the size of the transistors, which allows to fit more
transistors on a chip without changing rapidly the size of the chip; in general the size of the chip
increases slowly over time. Increasing the number of transistors is one of the main factors used to
increase the speed and performance of the CPU; more transistors, more clock cycles [4, 5].
2. PHYSICAL LIMITATION OF INCREASING THE NUMBER OF
THRANSISTORS
Increasing the number of transistors will be limited due to the following limitation [6, 7, 8, 9]:

2 Computer Science & Information Technology (CS & IT)
2.1 DESIGN RULE, DENSITY AND DESIGN COMPLEXITY [10,11,12,13]
Since the wires and transistors are drawn photographically on the chips, the size of the transistors
and the width of the wires are determined by the pixel size of the imaging process. The number of
transistors that can be fitted on a given chip size is determined by the size of the transistors. As
the feature size on the chip goes down, the number of transistors rises, which means the smaller
the transistors, the more transistors can be fitted on the chip. Feature sizes have decreased from
10 micron in 1971 to 0.022 micron in 2012; the semiconductor manufacturing processes have
decreased from 10 µm in 1971 to 22 nm in 2012 [4, 13, 14] (see Table 1).
Since the surface area of a transistor determines the transistor count per square millimeter of
silicon, the transistors density increases quadratically with a linear decrease in feature size [1].
The increase in transistor performance is more complicated. As the feature sizes shrink, devices
shrink quadratically in the horizontal and vertical dimensions. A reduction in operating voltage to
maintain correct operation and reliability of the transistor is required in the vertical dimension
shrink. This combination of scaling factors leads to a complex interrelationship between the
transistor performance and the process feature size.
Due to the shrinking of the pixel size and the increasing of the density, the design complexity
increases. If the pixel size shrinks double and the density increases double every two years
according to Moore’s Law, the physical limitation will definitely appear in few years, which
means that it will be very difficult to apply Moore’s Law in future. Some studies have shown that
physical limitations could be reached by 2018 [6] or 2020-2022 [15, 16, 17, 18].
2.2 NUMBER OF TRANSISTORS AND DIE SIZE [5, 11, 19]
The die size is a size of a chip; the chip size depends on the chip manufacturing processes. As
mentioned above, the smaller feature size, the larger number of transistors on a chip, this allows
increasing the die size slowly over time. Owing to increasing of the number of transistors the chip
size should be increased. The number of transistors of Intel microprocessors has increased from
2300 in 1971 to 3.1 billion in 2012 (1.35 million times during 40 years) and the die size has
increased from 12 mm2 in 1971 to 544 mm2 in 2012 (45.34 times during 40 years). The number
of transistors of MIPS microprocessors has increased from 110,000 (R2000) in 1986 to 7.15
million (R12000) in 1998 (65 times during 12 years) and the die size has increased from 80 mm2
in 1986 to 229.22 mm2 in 1998 (2.86 times during 12 years). The number of transistors of IBM
microprocessors has increased from 1.5 million (Power3) in 1998 to 2.1 billion (Power7+) in
2012 (1400 times during 14 years) and the die size has increased from 270 mm2 in 1998 to 567
mm2 in 2012 (2.1 times during 14 years) (see Table 1).
2.3 PIXEL SIZE AND POWER CONSUMPTION [1,20]
The transistor’s switching speed is determined by the pixel size, the smaller transistors switch
faster, so they require less power to be switched. Small pixel size allows the battery life of
portable devices to be saved for a long time. The energy consumption required for switching
transistors is known as dynamic power consumption, which can be increased due to charging and
discharging the capacitive output load. Increasing the amount of power requires more hot
generation, which can cause an interference of transistors with each other.

Computer Science & Information Technology (CS & IT) 3
Table 1. Evolution of Microprocessors: 1991 to 2012
Manufacturer Processor Date of
introduction
Number of
transistors
Process Area [mm2
]
Intel
Intel4004 1971 2,300 10 µm 12
Intel8008 1972 3,500 10 µm 14
Intel8080 1974 4,400 6 µm 20
Intel8085 1976 6,500 3 µm 20
Intel8086 1978 29,000 3 µm 33
Intel80286 1982 134,000 1.5 µm 44
Intel80386 1985 275,000 1.5 µm 104
Intel80486 1989 1,180,235 1 µm 173
Pentium 1993 3,100,000 0.8 µm 294
Pentium Pro 1995 5,500,000 0.5 µm 307
Pentium II 1997 7,500,000 0.35 µm 195
Pentium III 1999 9,500,000 0.25 µm 128
Pentium 4 2000 42,00,000 180 nm 217
Itanium 2 McKinely 2002 220,000,000 180 nm 421
Core 2 Duo 2006 291,000,000 65 nm 143
Core i7 (Quad) 2008 731,000,000 45 nm 263
Six-Core Core i7 2010 1,170,000,000 32 nm 240
Six-Core Core i7/8-
Core Xeon E5
2011 2,270,000,000 32 nm 434
8-Core Itanium
Poulson
2012 3.100,000,000 32 nm 544
MIPS
R2000 1986 110,000 2.0 µm 80
R3000 1988 150,000 1.2 µm 56
R4000 1991 1,200,000 0.8 µm 213
R10000 1994 2,600,000 0.5 µm 299
R10000 1996 6,800,000 0.35 µm 299
R12000 1998 7,1500,000 0.25 µm 229
IBM
POWER3 1998 15,000,000 0.35 µm 270
POWER4 2001 174,000,000 180 nm 412
POWER4+ 2002 184,000,000 130 nm 267
POWER5 2004 276,000,000 130 nm 389
POWER5+ 2005 276,000,000 90 nm 243
POWER6+ 2009 790,000,000 65 nm 341
POWER7 2010 1.200,000,000 45 nm 567
POWER7+ 2012 2.100,000,000 32 nm 567

3. TWO-LAYER CRYSTAL SQUARE
This article suggests as a solution for avoiding the physical limitations mentioned above and for
applying Moore’s Law more years a new approach of constructing a crystal square contained
transistors on a chip. This approach is called two-layer crystal square of transistors.
Assume a microprocessor (let’s say X) has the following specifications: date of introduction –
2015, one-layer crystal square of transistors, transistor count (number of transistors) – 3 billion,
pixel size (feature size) – 0.038 micron, die size (area) – 600 mm2 (see Figure 1).
Figure 1. Crystal Square of Transistors
A crystal square of transistors can be constructed using two layers that are equal in size and
transistor count, this allows to increase the number of transistors twice (Moore’s Law) without
touching the feature size and die size (see Figure 2).
Figure 2. Two-Layer Crystal Square of Transistors
The two-layer crystal square of transistors shown in Figure 2 contains double transistor count
without changing feature size and die size. Double number of transistors allows increasing a
microprocessor’s speed approximately twice. In this case the voltage of transistors must be
decreased in order to reduce the dynamic power and energy.
Assume the second generation of the microprocessor (X1) has the following specifications: date
of introduction – 2017, Two-layer crystal square, and each layer with the following
characteristics: number of transistors – 3 billion (6 billion both), pixel size – 0.038 micron, die
size – 600 mm2 (see Figure 3).

Figure 3. Second Generation
Applying this approach would double transistor count and speed of the microprocessors every
two years with slowly increasing the density and area, and decreasing feature size and voltage, so
the physical limitations will be reduced and Moore’s Law will be continued for many years (may
be till 2037 or more).
Assume the third generation of the microprocessor (X2) has the following specifications: date of
introduction – 2019, Two-layer crystal square, and each layer with the following characteristics:
transistor count – 6 billion (12 billion both), pixel size – 0.035 micron, die size – 688 mm2 (see
Figure 4).
Figure 4. Third Generation
The third generation doubles the transistor count with small decreasing of the feature size (from
0.038 micron to 0.035 micron) and small increasing of the area (from 600 mm2 to 688 mm2), of
course, with more density, which allows to fit more transistors in small area.
All the assuming characteristics of the microprocessors (X, X1, X2) and others (XN-1 and XN)
are shown in Table 2. This table shows some of the assuming microprocessors from 2015 till the
year 2037. The number of transistors will have doubled every two years. The feature size will
have increased 3 nm (from 38 to 35) every two years (from the year 2019 – after applying a two-
layer crystal technology in the year 2017), which means that the feature size would be reached 0.8
nm by the year 2037, which increases the design complexity and causes reaching the physical
limitation. The area will also have increased about 12-15% every two years (from the year 2019);
the small increasing of area can be applied if and only if the density of transistors increases with
each generation. The increasing of the density would also be reached the physical limitation by
the year 2037, as more density needs more power, which requires more hot generation.

Table 2. Assuming Evolution of Microprocessors: 2015 to 2037
Processor Date of
introduction
Number of
transistors
(billion)
Feature
size (nm)
Area [mm2
]
X 2015 3 38 600
X1 (Two-layer crystal) 2017 6 38 600 (each layer)
-- -- -- -- ---
XN-1 (Two-layer crystal) 2035 3072 11 ---
XN (Two-layer crystal) 2037 6144 0.8 ---
As shown in the table below, several measures of microprocessors technology, such as transistor
count, is improving at exponential rate based on Moore’s Law (see Figure 5).
Figure 5. Number of transistors related to Moore’s Law

4. CONCLUSION
The problem of applying Moore’s law in microprocessor technology as much as possible is still
topical research field although it has been studied by the research community for many decades.
The main objective of this article is to find as much as possible a complete and logical solution
for avoiding physical limitation in manufacturing of microprocessors technology and applying
Moore’s Law for a long time.
As mentioned above, some studies have shown that physical limitations could be reached by 2018
or 2022. Applying the new approach in microprocessor technology will delay the physical
limitation for few more years, because it doubles the transistor count every two years based on
Moore’s Law, which approximately double the microprocessor’s speed, with small decreasing of
the feature size and small increasing of the area.
ACKNOWLEDGMENT
The author would like to thank the president of Lebanese International University HE Abdel
Rahim Mourad and the LIU Bekaa campus administration for their continuous encouragement of
research activities at the university.
REFERENCES
[1] John L. Hennessy and David A. Patterson, “Computer Architecture, A Quantitative Approach”, 5th
ed., pp. 17-26, 2011.
[2] Gorden E. Moore, “cramming more Components onto Integrated Circuits”, Electronics, Vol. 38, No.
8, April 19, 1965.
[3] Jane Laudon, Kenneth Laudon, “Essentials of Management Information Systems”, Chapter 4: IT
Infrastructure: Hardware and Software, 10th ed., 2012.
[4] Steve Gilheany, “Evolution of Intel Microprocessors: 1971 to 2007”.
[5] Wolfgang Arden, “Future roadblocks and solutions in silicon technology as outlined by the ITRS
roadmap” in Mterials Science in Semiconductor Processing, vol. 5 issue 4-5 August – October, 2002,
pp. 313-319.
[6] Jan M. Rabaey, “Design at the end of Silicon Roadmap”, Keynotes Address III, University of
California, Berkelev, IEEE, ASP-DAC 2005.
[7] Damon Poeter, “Intel’s Gelsinger Sees Clear Path to 10nm Chips”, June 30, 2008.
[8] Hasan S., Humaria, Asghar M., “Limitation of Silicon Based Computation abd Future Prospects” in
Proceedings of Second International Conference on Communication Software and Networks, 2010.
ICCSN’10, pp. 599-561.
[9] Robert W. Keyes, “Physical limits of silicon transistors and circuits”, September 2005.
[10] F. Morals, L. Torres, M. Robert, D. Auvergne, “Estimation of layout densities for CMOS digital
circuits”, Proceeding International Workshop on Power and Timing Modeling Optimization
Simulation (PATMOS’98), pp. 61-70, November 1998, Lyngby, Danemark.
[11] Ulrich Sigmund, Marc Steinhaus, and Theo Ungerer, “On Performance, Transistor Count and Chip
Space Assessment of Multimedia-enhanced Simultaneous Multithreaded Processors”, Workshop on
Multi-Threaded Execution, Architecture and Compilation (MTEAC-4), Monterrey, Ca., Dec., 10,
2000.
[12] Singh B.P., Singh Renu, “Advanced Microprocessors and Microcontrollers”, New Age International,
Jan 1, 2008, 592 pages.
[13] Intel, “Transistors to Transformations, Form Sand to Circuits-How Intel Makes Chips”.
[14] “A History of Microprocessors Transistor Count 1971 to 2013”.

[15] Ahmad, Khaled; Schuegraf, Klaus, “Transistor Wars: Rival architecture face off in a bid to keep
Moore’s Law alive”, IEEE Spectrum: 50, November 2011.
[16] Brooke Crothers, “End of Moore’s Law: it’s not just about physics”, August 28, 2013.
[17] Robert Colwell, “The Chip Design Game at the End of Moore’s Law”, Hot Chips, August 2013.
[18] Joel Hruska, “Intel’s former chief architect: Moore’s law will be dead within a decade”, August 30,
2013.
[19] Yale N. Patt, Sanjay J. Patel, Marius Evers, Daniel H. Friendly, Jared Stark, “One Billion Transistors,
One Uniprocessor, One Chip”, IEEE Computer Vol. 30 Issue 9, pp. 51-57, 1997.
[20] P. Zdebel, “Law Power/Low Voltage CMOS Technologies, A Comparative Analysis”,
Microelectronics Engineering, Vol. 39, Elsevier, Dec. 1997, pp. 123-137.
AUTHOR
Haissam El-Aawar is an Associate Professor in the Department of Computer Science and
Information Technology at the Lebanese International University where he has been a
faculty member since 2009. Haissam completed his Ph.D. and M.Sc. degrees at the State
University "Lviv Polytechnic" in Ukraine. His research interests lie in the area of Artificial
Intelligence, theory of complexity, microprocessors evaluation, CISC- and RISC-
architectures, robotics control, mobility control and wireless communication.

Two-Layer Crystal Square for Transistors on a Single Chip

More Related Content

Similar to Two-Layer Crystal Square for Transistors on a Single Chip (20)

Recently uploaded (20)

Two-Layer Crystal Square for Transistors on a Single Chip