Transistor count

https://en.wikipedia.org/wiki/Transistor_count

 

Transistor count

From Wikipedia, the free encyclopedia

Jump to navigationJump to search

Plot of MOS transistor counts for microprocessors against dates of in­tro­duction. The curve shows counts doubling every two years, per Moore's law.

The transistor count is the number of transistors on an integrated circuit (IC). It typically refers to the number of MOSFETs (metal-oxide-semiconductor field-effect transistors, or MOS transistors) on an IC chip, as all modern ICs use MOSFETs. It is the most common measure of IC complexity (although the majority of transistors in modern microprocessors are contained in the cache memories, which consist mostly of the same memory cell circuits replicated many times). The rate at which MOS transistor counts have increased generally follows Moore's law, which observed that the transistor count doubles approximately every two years.

As of 2019, the largest transistor count in a commercially available microprocessor is 39.54 billion MOSFETs, in AMD's Zen 2 based Epyc Rome, which is a 3D integrated circuit (with eight dies in a single package) fabricated using TSMC's 7 nm FinFET semiconductor manufacturing process.[1][2] As of 2018, the highest transistor count in a graphics processing unit (GPU) is Nvidia's GV100 Volta with 21.1 billion MOSFETs, manufactured using TSMC's 12 nm FinFET process.[3] As of 2019, the highest transistor count in any IC chip is Samsung's 1 TB eUFS (3D-stacked) V-NAND flash memory chip, with 2 trillion floating-gate MOSFETs (4 bits per transistor).[4] As of 2019, the highest transistor count in a non-memory chip is a deep learning engine called the Wafer Scale Engine by Cerebras, using a special design to route around any non-functional core on the device; it has 1.2 trillion MOSFETs, manufactured using TSMC's 16 nm FinFET process.[5][6][7][8]

In terms of computer systems that consist of numerous integrated circuits, the supercomputer with the highest transistor count as of 2016 is the Chinese-designed Sunway TaihuLight, which has for all CPUs/nodes (1012 for the 10 million cores and for RAM 1015 for the 1.3 million GB) combined "about 400 trillion transistors in the processing part of the hardware" and "the DRAM includes about 12 quadrillion transistors, and that's about 97 percent of all the transistors."[9] To compare, the smallest computer, as of 2018 dwarfed by a grain of sand, has on the order of 100,000 transistors, and the one, fully programmable, with the fewest transistors ever has 130 transistors or fewer.

In terms of the total number of transistors in existence, it has been estimated that a total of 13 sextillion (1.3×1022) MOSFETs have been manufactured worldwide between 1960 and 2018, accounting for at least 99.9% of all transistors. This makes the MOSFET the most widely manufactured device in history.[10]

 

Contents

• 1Transistor count
  • 1.1Microprocessors
  • 1.2GPUs
  • 1.3FPGA
  • 1.4Memory
  • 1.5Transistor computers
  • 1.6Logic functions
  • 1.7Parallel systems
  • 1.8Other devices
• 2Transistor density
  • 2.1MOSFET nodes
• 3See also
• 4Notes
• 5References
• 6External links

Transistor count

 

Part of an IBM 7070 card cage populated with Standard Modular System cards

Among the earliest products to use transistors were portable transistor radios, introduced in 1954, which typically used 4 to 8 transistors, often advertising the number on the radio's case. However, early junction transistors were relatively bulky devices that were difficult to manufacture on a mass-production basis, limiting the transistor counts and restricting their usage to a number of specialised applications.[11]

The MOSFET (MOS transistor), invented by Mohamed Atalla and Dawon Kahng at Bell Labs in 1959,[12] was the first truly compact transistor that could be miniaturised and mass-produced for a wide range of uses.[11] The MOSFET made it possible to build high-density integrated circuits (ICs),[13] enabling Moore's law[14][15] and very large-scale integration.[16] Atalla first proposed the concept of the MOS integrated circuit (MOS IC) chip in 1960, followed by Kahng in 1961, both noting that the MOSFET's ease of fabrication made it useful for integrated circuits.[11][17] The earliest experimental MOS IC to be demonstrated was a 16-transistor chip built by Fred Heiman and Steven Hofstein at RCA Laboratories in 1962.[15] Further large-scale integration was made possible with an improvement in MOSFET semiconductor device fabrication, the CMOS process, developed by Chih-Tang Sah and Frank Wanlass at Fairchild Semiconductor in 1963.[18]

Microprocessors[edit]

See also: Microprocessor chronology and Microcontroller

This subsection needs additional citations for verification. Relevant discussion may be found on the talk page. Please help improve this article by adding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "Transistor count" – news · newspapers · books · scholar · JSTOR (December 2019) (Learn how and when to remove this template message)

A microprocessor incorporates the functions of a computer's central processing unit on a single integrated circuit. It is a multi-purpose, programmable device that accepts digital data as input, processes it according to instructions stored in its memory, and provides results as output.

The development of MOS integrated circuit technology in the 1960s led to the development of the first microprocessors.[19] The 20-bit MP944, developed by Garrett AiResearch for the U.S. Navy's F-14 Tomcat fighter in 1970, is considered by its designer Ray Holt to be the first microprocessor.[20] It was a multi-chip microprocessor, fabricated on six MOS chips. However, it was classified by the Navy until 1998. The 4-bit Intel 4004, released in 1971, was the first single-chip microprocessor. It was made possible with an improvement in MOSFET design, MOS silicon-gate technology (SGT), developed in 1968 at Fairchild Semiconductor by Federico Faggin, who went on to use MOS SGT technology to develop the 4004 with Marcian Hoff, Stanley Mazor and Masatoshi Shima at Intel.[19]

All chips over e.g. a million transistors have lots of memory, usually cache memories in level 1 and 2 or more levels, accounting for most transistors on microprocessors in modern times, where large caches have become the norm. The level 1 caches of the Pentium Pro die accounted for over 14% of its transistors, while the much larger L2 cache was on a separate die, but on-package, so it's not included in the transistor count. Later chips included more levels, L2 or even L3 on-chip. The last DEC Alpha chip made has 90% of it for cache.[21]

While Intel's i960CA small cache of 1 KB, at about 50,000 transistors, isn't a big part of the chip, it alone would have been very large in early microprocessors. In the ARM 3 chip, with 4 KB, the cache was over 63% of the chip, and in the Intel 80486 its larger cache is only over a third of it because the rest of the chip is more complex. So cache memories are the largest factor, except for in early chips with smaller caches or even earlier chips with no cache at all. Then the inherent complexity, e.g. number of instructions, is the dominant factor, more than e.g. the memory the registers of the chip represent.

Processor	MOS transistor count	Date of introduction	Designer	MOS process	Area
MP944 (20-bit, 6-chip)	?	1970[20][a]	Garrett AiResearch	?	?
Intel 4004 (4-bit, 16-pin)	2,250	1971	Intel	10,000 nm	12 mm2
Intel 8008 (8-bit, 18-pin)	3,500	1972	Intel	10,000 nm	14 mm2
NEC μCOM-4 (4-bit, 42-pin)	2,500[22][23]	1973	NEC	7,500 nm[24]	?
Toshiba TLCS-12 (12-bit)	11,000+[25]	1973	Toshiba	6,000 nm	32 mm2
Intel 4040 (4-bit, 16-pin)	3,000	1974	Intel	10,000 nm	12 mm2
Motorola 6800 (8-bit, 40-pin)	4,100	1974	Motorola	6,000 nm	16 mm2
Intel 8080 (8-bit, 40-pin)	6,000	1974	Intel	6,000 nm	20 mm2
TMS 1000 (4-bit, 28-pin)	8,000	1974[26]	Texas Instruments	8,000 nm	11 mm2
MOS Technology 6502 (8-bit, 40-pin)	4,528[b][27]	1975	MOS Technology	8,000 nm	21 mm2
Intersil IM6100 (12-bit, 40-pin; clone of PDP-8)	4,000	1975	Intersil
CDP 1801 (8-bit, 2-chip, 40-pin)	5,000	1975	RCA
RCA 1802 (8-bit, 40-pin)	5,000	1976	RCA	5,000 nm	27 mm2
Zilog Z80 (8-bit, 4-bit ALU, 40-pin)	8,500[c]	1976	Zilog	4,000 nm	18 mm2
Intel 8085 (8-bit, 40-pin)	6,500	1976	Intel	3,000 nm	20 mm2
TMS9900 (16-bit)	8,000	1976	Texas Instruments
Motorola MC14500B (1-bit, 16-pin)	?	1977	Motorola	?	?
Bellmac-8 (8-bit)	7,000	1977	Bell Labs	5,000 nm
Motorola 6809 (8-bit with some 16-bit features, 40-pin)	9,000	1978	Motorola	5,000 nm	21 mm2
Intel 8086 (16-bit, 40-pin)	29,000	1978	Intel	3,000 nm	33 mm2
Zilog Z8000 (16-bit)	17,500[28]	1979	Zilog
Intel 8088 (16-bit, 8-bit data bus)	29,000	1979	Intel	3,000 nm	33 mm2
Motorola 68000 (16/32-bit, 32-bit registers, 16-bit ALU)	68,000[29]	1979	Motorola	3,500 nm	44 mm2
Intel 8051 (8-bit, 40-pin)	50,000	1980	Intel
WDC 65C02	11,500[30]	1981	WDC	3,000 nm	6 mm2
ROMP (32-bit)	45,000	1981	IBM	2,000 nm
Intel 80186 (16-bit, 68-pin)	55,000	1982	Intel	3,000 nm	60 mm2
Intel 80286 (16-bit, 68-pin)	134,000	1982	Intel	1,500 nm	49 mm2
WDC 65C816 (8/16-bit)	22,000[31]	1983	WDC	3,000 nm[32]	9 mm2
NEC V20	63,000	1984	NEC
Motorola 68020 (32-bit; 114 pins used)	190,000[33]	1984	Motorola	2,000 nm	85 mm2
Intel 80386 (32-bit, 132-pin; no cache)	275,000	1985	Intel	1,500 nm	104 mm2
ARM 1 (32-bit; no cache)	25,000[33]	1985	Acorn	3,000 nm	50 mm2
Novix NC4016 (16-bit)	16,000[34]	1985[35]	Harris Corporation	3,000 nm[36]
SPARC MB86900 (32-bit; no cache)	110,000[37]	1986	Fujitsu	1,200 nm
NEC V60[38] (32-bit; no cache)	375,000	1986	NEC	1,500 nm
ARM 2 (32-bit, 84-pin; no cache)	27,000[39][33]	1986	Acorn	2,000 nm	30.25 mm2
Z80000 (32-bit; very small cache)	91,000	1986	Zilog
NEC V70[38] (32-bit; no cache)	385,000	1987	NEC	1,500 nm
Hitachi Gmicro/200[40]	730,000	1987	Hitachi	1,000 nm
Motorola 68030 (32-bit, very small caches)	273,000	1987	Motorola	800 nm	102 mm2
TI Explorer's 32-bit Lisp machine chip	553,000[41]	1987	Texas Instruments	2,000 nm[42]
DEC WRL MultiTitan	180,000[43]	1988	DEC WRL	1,500 nm	61 mm2
Intel i960 (32-bit, 33-bit memory subsystem, no cache)	250,000[44]	1988	Intel	1,500 nm[45]
Intel i960CA (32-bit, cache)	600,000[45]	1989	Intel	800 nm	143 mm2
Intel i860 (32/64-bit, 128-bit SIMD, cache, VLIW)	1,000,000[46]	1989	Intel
Intel 80486 (32-bit, 4 KB cache)	1,180,235	1989	Intel	1000 nm	173 mm2
ARM 3 (32-bit, 4 KB cache)	310,000	1989	Acorn	1,500 nm	87 mm2
Motorola 68040 (32-bit, 8 KB caches)	1,200,000	1990	Motorola	650 nm	152 mm2
R4000 (64-bit, 16 KB of caches)	1,350,000	1991	MIPS	1,000 nm	213 mm2
ARM 6 (32-bit, no cache for this 60 variant)	35,000	1991	ARM	800 nm
Hitachi SH-1 (32-bit, no cache)	600,000[47]	1992[48]	Hitachi	800 nm	10 mm2
Intel i960CF (32-bit, cache)	900,000[45]	1992	Intel		125 mm2
DEC Alpha 21064 (64-bit, 290-pin; 16 KB of caches)	1,680,000	1992	DEC	750 nm	233.52 mm2
Hitachi HARP-1 (32-bit, cache)	2,800,000[49]	1993	Hitachi	500 nm	267 mm2
Pentium (32-bit, 16 KB of caches)	3,100,000	1993	Intel	800 nm	294 mm2
ARM700 (32-bit; 8 KB cache)	578,977[50]	1994	ARM	700 nm	68.51 mm2
MuP21 (21-bit,[51] 40-pin; includes video)	7,000[52]	1994	Offete Enterprises	1200 nm
Motorola 68060 (32-bit, 16 KB of caches)	2,500,000	1994	Motorola	600 nm	218 mm2
SA-110 (32-bit, 32 KB of caches)	2,500,000[33]	1995	Acorn/DEC/Apple	350 nm	50 mm2
Pentium Pro (32-bit, 16 KB of caches;[53] L2 cache on-package, but on separate die)	5,500,000[54]	1995	Intel	500 nm	307 mm2
AMD K5 (32-bit, caches)	4,300,000	1996	AMD	500 nm	251 mm2
Hitachi SH-4 (32-bit, caches)	10,000,000[55]	1997	Hitachi	200 nm[56]	42 mm2[57]
Pentium II Klamath (32-bit, 64-bit SIMD, caches)	7,500,000	1997	Intel	350 nm	195 mm2
AMD K6 (32-bit, caches)	8,800,000	1997	AMD	350 nm	162 mm2
F21 (21-bit; includes e.g. video)	15,000	1997[52]	Offete Enterprises
AVR (8-bit, 40-pin; w/memory)	140,000 (48,000 excl. memory[58])	1997	Nordic VLSI/Atmel
Pentium II Deschutes (32-bit, large cache)	7,500,000	1998	Intel	250 nm	113 mm2
ARM 9TDMI (32-bit, no cache)	111,000[33]	1999	Acorn	350 nm	4.8 mm2
Pentium III Katmai (32-bit, 128-bit SIMD, caches)	9,500,000	1999	Intel	250 nm	128 mm2
Emotion Engine (64-bit, 128-bit SIMD, cache)	13,500,000[59]	1999	Sony/Toshiba	180 nm[60]	240 mm2[61]
Pentium II Mobile Dixon (32-bit, caches)	27,400,000	1999	Intel	180 nm	180 mm2
AMD K6-III (32-bit, caches)	21,300,000	1999	AMD	250 nm	118 mm2
AMD K7 (32-bit, caches)	22,000,000	1999	AMD	250 nm	184 mm2
Gekko (32-bit, large cache)	21,000,000[62]	2000	IBM/Nintendo	180 nm	43 mm2
Pentium III Coppermine (32-bit, large cache)	21,000,000	2000	Intel	180 nm	80 mm2
Pentium 4 Willamette (32-bit, large cache)	42,000,000	2000	Intel	180 nm	217 mm2
SPARC64 V (64-bit, large cache)	191,000,000[63]	2001	Fujitsu	130 nm[64]	290 mm2
Pentium III Tualatin (32-bit, large cache)	45,000,000	2001	Intel	130 nm	81 mm2
Pentium 4 Northwood (32-bit, large cache)	55,000,000	2002	Intel	130 nm	145 mm2
Itanium 2 McKinley (64-bit, large cache)	220,000,000	2002	Intel	180 nm	421 mm2
DEC Alpha 21364 (64-bit, 946-pin, SIMD, very large caches)	152,000,000[21]	2003	DEC	180 nm	397 mm2
Barton (32-bit, large cache)	54,300,000	2003	AMD	130 nm	101 mm2
AMD K8 (64-bit, large cache)	105,900,000	2003	AMD	130 nm	193 mm2
Itanium 2 Madison 6M (64-bit)	410,000,000	2003	Intel	130 nm	374 mm2
Pentium 4 Prescott (32-bit, large cache)	112,000,000	2004	Intel	90 nm	110 mm2
SPARC64 V+ (64-bit, large cache)	400,000,000[65]	2004	Fujitsu	90 nm	294 mm2
Itanium 2 (64-bit;9 MB cache)	592,000,000	2004	Intel	130 nm	432 mm2
Pentium 4 Prescott-2M (32-bit, large cache)	169,000,000	2005	Intel	90 nm	143 mm2
Pentium D Smithfield (32-bit, large cache)	228,000,000	2005	Intel	90 nm	206 mm2
Xenon (64-bit, 128-bit SIMD, large cache)	165,000,000	2005	IBM	90 nm
Cell (32-bit, cache)	250,000,000[66]	2005	Sony/IBM/Toshiba	90 nm	221 mm2
Pentium 4 Cedar Mill (32-bit, large cache)	184,000,000	2006	Intel	65 nm	90 mm2
Pentium D Presler (32-bit, large cache)	362,000,000	2006	Intel	65 nm	162 mm2
Core 2 Duo Conroe (dual-core 64-bit, large caches)	291,000,000	2006	Intel	65 nm	143 mm2
Dual-core Itanium 2 (64-bit, SIMD, large caches)	1,700,000,000[67]	2006	Intel	90 nm	596 mm2
AMD K10 quad-core 2M L3 (64-bit, large caches)	463,000,000[68]	2007	AMD	65 nm	283 mm2
ARM Cortex-A9 (32-bit, (optional) SIMD, caches)	26,000,000[69]	2007	ARM	45 nm	31 mm2
Core 2 Duo Wolfdale (dual-core 64-bit, SIMD, caches)	411,000,000	2007	Intel	45 nm	107 mm2
POWER6 (64-bit, large caches)	789,000,000	2007	IBM	65 nm	341 mm2
Core 2 Duo Allendale (dual-core 64-bit, SIMD, large caches)	169,000,000	2007	Intel	65 nm	111 mm2
Uniphier	250,000,000[70]	2007	Matsushita	45 nm	?
SPARC64 VI (64-bit, SIMD, large caches)	540,000,000	2007[71]	Fujitsu	90 nm	421 mm2
Core 2 Duo Wolfdale 3M (dual-core 64-bit, SIMD, large caches)	230,000,000	2008	Intel	45 nm	83 mm2
Core i7 (quad-core 64-bit, SIMD, large caches)	731,000,000	2008	Intel	45 nm	263 mm2
AMD K10 quad-core 6M L3 (64-bit, SIMD, large caches)	758,000,000[68]	2008	AMD	45 nm	258 mm2
Atom (32-bit, large cache)	47,000,000	2008	Intel	45 nm	24 mm2
SPARC64 VII (64-bit, SIMD, large caches)	600,000,000	2008[72]	Fujitsu	65 nm	445 mm2
Six-core Xeon 7400 (64-bit, SIMD, large caches)	1,900,000,000	2008	Intel	45 nm	503 mm2
Six-core Opteron 2400 (64-bit, SIMD, large caches)	904,000,000	2009	AMD	45 nm	346 mm2
SPARC64 VIIIfx (64-bit, SIMD, large caches)	760,000,000[73]	2009	Fujitsu	45 nm	513 mm2
16-core SPARC T3 (64-bit, SIMD, large caches)	1,000,000,000[74]	2010	Sun/Oracle	40 nm	377 mm2
Six-core Core i7 (Gulftown)	1,170,000,000	2010	Intel	32 nm	240 mm2
8-core POWER7 32M L3 (64-bit, SIMD, large caches)	1,200,000,000	2010	IBM	45 nm	567 mm2
Quad-core z196[75] (64-bit, very large caches)	1,400,000,000	2010	IBM	45 nm	512 mm2
Quad-core Itanium Tukwila (64-bit, SIMD, large caches)	2,000,000,000[76]	2010	Intel	65 nm	699 mm2
8-core Xeon Nehalem-EX (64-bit, SIMD, large caches)	2,300,000,000[77]	2010	Intel	45 nm	684 mm2
SPARC64 IXfx (64-bit, SIMD, large caches)	1,870,000,000[78]	2011	Fujitsu	40 nm	484 mm2
Quad-core + GPU Core i7 (64-bit, SIMD, large caches)	1,160,000,000	2011	Intel	32 nm	216 mm2
Six-core Core i7/8-core Xeon E5 (Sandy Bridge-E/EP) (64-bit, SIMD, large caches)	2,270,000,000[79]	2011	Intel	32 nm	434 mm2
10-core Xeon Westmere-EX (64-bit, SIMD, large caches)	2,600,000,000	2011	Intel	32 nm	512 mm2
Atom "Medfield" (64-bit)	432,000,000[80]	2012	Intel	32 nm	64 mm2
SPARC64 X (64-bit, SIMD, caches)	2,990,000,000[81]	2012	Fujitsu	28 nm	600 mm2
8-core AMD Bulldozer (64-bit, SIMD, caches)	1,200,000,000[82]	2012	AMD	32 nm	315 mm2
Quad-core + GPU AMD Trinity (64-bit, SIMD, caches)	1,303,000,000	2012	AMD	32 nm	246 mm2
Quad-core + GPU Core i7 Ivy Bridge (64-bit, SIMD, caches)	1,400,000,000	2012	Intel	22 nm	160 mm2
8-core POWER7+ (64-bit, SIMD, 80 MB L3 cache)	2,100,000,000	2012	IBM	32 nm	567 mm2
Six-core zEC12 (64-bit, SIMD, large caches)	2,750,000,000	2012	IBM	32 nm	597 mm2
8-core Itanium Poulson (64-bit, SIMD, caches)	3,100,000,000	2012	Intel	32 nm	544 mm2
61-core Xeon Phi (32-bit, 512-bit SIMD, caches)	5,000,000,000[83]	2012	Intel	22 nm	720 mm2
Apple A7 (dual-core 64/32-bit ARM64, "mobile SoC", SIMD, caches)	1,000,000,000	2013	Apple	28 nm	102 mm2
Six-core Core i7 Ivy Bridge E (64-bit, SIMD, caches)	1,860,000,000	2013	Intel	22 nm	256 mm2
12-core POWER8 (64-bit, SIMD, caches)	4,200,000,000	2013	IBM	22 nm	650 mm2
Xbox One main SoC (64-bit, SIMD, caches)	5,000,000,000	2013	Microsoft/AMD	28 nm	363 mm2
Quad-core + GPU Core i7 Haswell (64-bit, SIMD, caches)	1,400,000,000[84]	2014	Intel	22 nm	177 mm2
Apple A8 (dual-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	2,000,000,000	2014	Apple	20 nm	89 mm2
8-core Core i7 Haswell-E (64-bit, SIMD, caches)	2,600,000,000[85]	2014	Intel	22 nm	355 mm2
Apple A8X (tri-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	3,000,000,000[86]	2014	Apple	20 nm	128 mm2
15-core Xeon Ivy Bridge-EX (64-bit, SIMD, caches)	4,310,000,000[87]	2014	Intel	22 nm	541 mm2
18-core Xeon Haswell-E5 (64-bit, SIMD, caches)	5,560,000,000[88]	2014	Intel	22 nm	661 mm2
Quad-core + GPU GT2 Core i7 Skylake K (64-bit, SIMD, caches)	1,750,000,000	2015	Intel	14 nm	122 mm2
Dual-core + GPU Iris Core i7 Broadwell-U (64-bit, SIMD, caches)	1,900,000,000[89]	2015	Intel	14 nm	133 mm2
Apple A9 (dual-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	2,000,000,000+	2015	Apple	14 nm (Samsung)	96 mm2 (Samsung)
				16 nm (TSMC)	104.5 mm2 (TSMC)
Apple A9X (dual core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	3,000,000,000+	2015	Apple	16 nm	143.9 mm2
IBM z13 (64-bit, caches)	3,990,000,000	2015	IBM	22 nm	678 mm2
IBM z13 Storage Controller	7,100,000,000	2015	IBM	22 nm	678 mm2
32-core SPARC M7 (64-bit, SIMD, caches)	10,000,000,000[90]	2015	Oracle	20 nm
Qualcomm Snapdragon 835 (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	3,000,000,000[91][92]	2016	Qualcomm	10 nm	72.3 mm2
10-core Core i7 Broadwell-E (64-bit, SIMD, caches)	3,200,000,000[93]	2016	Intel	14 nm	246 mm2[94]
Apple A10 Fusion (quad-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	3,300,000,000	2016	Apple	16 nm	125 mm2
HiSilicon Kirin 960 (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	4,000,000,000[95]	2016	Huawei	16 nm	110.00 mm2
22-core Xeon Broadwell-E5 (64-bit, SIMD, caches)	7,200,000,000[96]	2016	Intel	14 nm	456 mm2
72-core Xeon Phi (64-bit, 512-bit SIMD, caches)	8,000,000,000	2016	Intel	14 nm	683 mm2
Zip CPU (32-bit, for FPGAs)	1,286 6-LUTs[97]	2016	Gisselquist Technology
Qualcomm Snapdragon 845 (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	5,300,000,000[98]	2017	Qualcomm	10 nm	94 mm2
Qualcomm Snapdragon 850 (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	5,300,000,000[99]	2017	Qualcomm	10 nm	94 mm2
Apple A11 Bionic (hexa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	4,300,000,000	2017	Apple	10 nm	89.23 mm2
Zeppelin SoC Ryzen (64-bit, SIMD, caches)	4,800,000,000[100]	2017	AMD	14 nm	192 mm2
Ryzen 5 1600 Ryzen (64-bit, SIMD, caches)	4,800,000,000[101]	2017	AMD	14 nm	213 mm2
Ryzen 5 1600 X Ryzen (64-bit, SIMD, caches)	4,800,000,000[102]	2017	AMD	14 nm	213 mm2
IBM z14 (64-bit, SIMD, caches)	6,100,000,000	2017	IBM	14 nm	696 mm2
IBM z14 Storage Controller (64-bit)	9,700,000,000	2017	IBM	14 nm	696 mm2
HiSilicon Kirin 970 (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	5,500,000,000[103]	2017	Huawei	10 nm	96.72 mm2
Xbox One X (Project Scorpio) main SoC (64-bit, SIMD, caches)	7,000,000,000[104]	2017	Microsoft/AMD	16 nm	360 mm2[104]
28-core Xeon Platinum 8180 (64-bit, SIMD, caches)	8,000,000,000[105][disputed – discuss]	2017	Intel	14 nm
POWER9 (64-bit, SIMD, caches)	8,000,000,000	2017	IBM	14 nm	695 mm2
Freedom U500 Base Platform Chip (E51, 4×U54) RISC-V (64-bit, caches)	250,000,000[106]	2017	SiFive	28 nm	~30 mm2
SPARC64 XII (12-core 64-bit, SIMD, caches)	5,450,000,000[107]	2017	Fujitsu	20 nm	795 mm2
Apple A10X Fusion (hexa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	4,300,000,000[108]	2017	Apple	10 nm	96.40 mm2
Centriq 2400 (64/32-bit, SIMD, caches)	18,000,000,000[109]	2017	Qualcomm	10 nm	398 mm2
32-core AMD Epyc (64-bit, SIMD, caches)	19,200,000,000	2017	AMD	14 nm	768 mm2
Qualcomm Snapdragon 710 (octa-core ARM64 "mobile SoC", SIMD, caches)	?	2018	Qualcomm	10 nm	?
Qualcomm Snapdragon 675 (octa-core ARM64 "mobile SoC", SIMD, caches)	?	2018	Qualcomm	11 nm	?
Qualcomm Snapdragon 855 (octa-core ARM64 "mobile SoC", SIMD, caches)	?	2018	Qualcomm	7 nm	73.27 mm2
Qualcomm Snapdragon 8cx / SCX8180 (octa-core ARM64 "mobile SoC", SIMD, caches)	8,500,000,000[110]	2018	Qualcomm	7 nm	112 mm2
Apple A12 Bionic (hexa-core ARM64 "mobile SoC", SIMD, caches)	6,900,000,000[111][112]	2018	Apple	7 nm	83.27 mm2
HiSilicon Kirin 980 (octa-core ARM64 "mobile SoC", SIMD, caches)	6,900,000,000[113]	2018	Huawei	7 nm	74.13 mm2
HiSilicon Kirin 990 5G	10,300,000,000[114]	2019	Huawei	7 nm	113.31 mm2
HiSilicon Kirin 990 4G	8,000,000,000[115]	2019	Huawei	7 nm	90.00 mm2
HiSilicon Kirin 710 (octa-core ARM64 "mobile SoC", SIMD, caches)	5,500,000,000[116]	2018	Huawei	12 nm
Apple A12X Bionic (octa-core 64/32-bit ARM64 "mobile SoC", SIMD, caches)	10,000,000,000[117]	2018	Apple	7 nm	122 mm2
Apple A13 (iPhone 11 Pro)	8,500,000,000[118][119]	2019	Apple	7 nm	98.48 mm2
Fujitsu A64FX (64/32-bit, SIMD, caches)	8,786,000,000[120]	2018[121]	Fujitsu	7 nm
Tegra Xavier SoC (64/32-bit)	9,000,000,000[122]	2018	Nvidia	12 nm	350 mm2
Samsung Exynos 9820 (octa-core ARM64 "mobile SoC", SIMD, caches)	?	2019	Samsung	8 nm	127 mm2
AMD Ryzen 7 3700X (64-bit, SIMD, caches, I/O die)	5,990,000,000[123][d]	2019	AMD	7&12 nm (TSMC)	199 (74+125) mm2
AMD Ryzen 9 3900X (64-bit, SIMD, caches, I/O die)	9,890,000,000[1][2]	2019	AMD	7 & 12 nm (TSMC)	273 mm2
AMD Epyc Rome (64-bit, SIMD, caches)	39,540,000,000[1][2]	2019	AMD	7 & 12 nm (TSMC)	1088 mm2
AWS Graviton2 (64-bit, 64-core ARM-based, SIMD, caches)[124][125]	30,000,000,000	2019	Amazon	7 nm

GPUs[edit]

A graphics processing unit (GPU) is a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the building of images in a frame buffer intended for output to a display.

The designer refers to the technology company that designs the logic of the integrated circuit chip (such as Nvidia and AMD). The manufacturer refers to the semiconductor company that fabricates the chip using its semiconductor manufacturing process at a foundry (such as TSMC and Samsung Semiconductor). The transistor count in a chip is dependent on a manufacturer's fabrication process, with smaller semiconductor nodes typically enabling higher transistor density and thus higher transistor counts.

The random-access memory (RAM) that comes with GPUs (such as VRAM, SGRAM or HBM) greatly increase the total transistor count, with the memory typically accounting for the majority of transistors in a graphics card. For example, Nvidia's Tesla P100 has 15 billion FinFETs (16 nm) in the GPU in addition to 16 GB of HBM2 memory, totaling about 150 billion MOSFETs on the graphics card.[126] The following table does not include the memory. For memory transistor counts, see the Memory section below.

Processor	MOS transistor count	Date of introduction	Designer(s)	Manufacturer(s)	MOS process	Area	Ref
µPD7220 GDC	40,000	1982	NEC	NEC	5,000 nm		[127]
ARTC HD63484	60,000	1984	Hitachi	Hitachi			[128]
YM7101 VDP	100,000	1988	Sega	Yamaha			[129]
Tom & Jerry	750,000	1993	Flare	IBM			[129]
VDP1	1,000,000	1994	Sega	Hitachi	500 nm		[130][131]
Sony GPU	1,000,000	1994	Toshiba	LSI	500 nm		[132][133][134]
NV1	1,000,000	1995	Nvidia, Sega	SGS	500 nm	90 mm2	[130]
Reality Coprocessor	2,600,000	1996	SGI	NEC	350 nm	81 mm2	[135]
PowerVR	1,200,000	1996	VideoLogic	NEC	350 nm		[136]
Voodoo Graphics	1,000,000	1996	3dfx	TSMC	500 nm		[137][138]
Voodoo Rush	1,000,000	1997	3dfx	TSMC	500 nm		[137][138]
NV3	3,500,000	1997	Nvidia	SGS, TSMC	350 nm	90 mm2	[139][140]
PowerVR2 CLX2	10,000,000	1998	VideoLogic	NEC	250 nm	116 mm2	[55][141][142][57]
i740	3,500,000	1998	Intel, Real3D	Real3D	350 nm		[137][138]
Voodoo 2	4,000,000	1998	3dfx	TSMC	350 nm
Voodoo Rush	4,000,000	1998	3dfx	TSMC	350 nm
Riva TNT	7,000,000	1998	Nvidia	TSMC	350 nm		[137][140]
PowerVR2 PMX1	6,000,000	1999	VideoLogic	NEC	250 nm		[143]
Rage 128	8,000,000	1999	ATI	TSMC, UMC	250 nm	70 mm2	[138]
Voodoo 3	8,100,000	1999	3dfx	TSMC	250 nm		[144]
Graphics Synthesizer	43,000,000	1999	Sony, Toshiba	Sony, Toshiba	180 nm	279 mm2	[62][60][59][61]
NV5	15,000,000	1999	Nvidia	TSMC	250 nm		[138]
NV10	17,000,000	1999	Nvidia	TSMC	220 nm	111 mm2	[145][140]
Voodoo 4	14,000,000	2000	3dfx	TSMC	220 nm		[137][138]
NV11	20,000,000	2000	Nvidia	TSMC	180 nm	65 mm2	[138]
NV15	25,000,000	2000	Nvidia	TSMC	180 nm	81 mm2	[138]
Voodoo 5	28,000,000	2000	3dfx	TSMC	220 nm		[137][138]
R100	30,000,000	2000	ATI	TSMC	180 nm	97 mm2	[138]
Flipper	51,000,000	2000	ArtX	NEC	180 nm	106 mm2	[62][146]
PowerVR3 KYRO	14,000,000	2001	Imagination	ST	250 nm		[137][138]
PowerVR3 KYRO II	15,000,000	2001	Imagination	ST	180 nm
NV2A	60,000,000	2001	Nvidia	TSMC	150 nm		[137][147]
NV20	57,000,000	2001	Nvidia	TSMC	150 nm	128 mm2	[138]
R200	60,000,000	2001	ATI	TSMC	150 nm	68 mm2
NV25	63,000,000	2002	Nvidia	TSMC	150 nm	142 mm2
R300	107,000,000	2002	ATI	TSMC	150 nm	218 mm2
R360	117,000,000	2003	ATI	TSMC	150 nm	218 mm2
NV38	135,000,000	2003	Nvidia	TSMC	130 nm	207 mm2
R480	160,000,000	2004	ATI	TSMC	130 nm	297 mm2
NV40	222,000,000	2004	Nvidia	IBM	130 nm	305 mm2
Xenos	232,000,000	2005	ATI	TSMC	90 nm	182 mm2	[148][149]
RSX Reality Synthesizer	300,000,000	2005	Nvidia, Sony	Sony	90 nm	186 mm2	[150][151]
G70	303,000,000	2005	Nvidia	TSMC, Chartered	110 nm	333 mm2	[138]
R520	321,000,000	2005	ATI	TSMC	90 nm	288 mm2
R580	384,000,000	2006	ATI	TSMC	90 nm	352 mm2
G80	681,000,000	2006	Nvidia	TSMC	90 nm	480 mm2
G86 Tesla	210,000,000	2007	Nvidia	TSMC	80 nm	127 mm2
G84 Tesla	289,000,000	2007	Nvidia	TSMC	80 nm	169 mm2
R600	700,000,000	2007	ATI	TSMC	80 nm	420 mm2
G92	754,000,000	2007	Nvidia	TSMC, UMC	65 nm	324 mm2
G98 Tesla	210,000,000	2008	Nvidia	TSMC	65 nm	86 mm2
RV710	242,000,000	2008	ATI	TSMC	55 nm	73 mm2
G96 Tesla	314,000,000	2008	Nvidia	TSMC	55 nm	121 mm2
G94 Tesla	505,000,000	2008	Nvidia	TSMC	65 nm	240 mm2
RV730	514,000,000	2008	ATI	TSMC	55 nm	146 mm2
RV670	666,000,000	2008	ATI	TSMC	55 nm	192 mm2
RV770	956,000,000	2008	ATI	TSMC	55 nm	256 mm2
RV790	959,000,000	2008	ATI	TSMC	55 nm	282 mm2	[152][138]
GT200b Tesla	1,400,000,000	2008	Nvidia	TSMC, UMC	55 nm	470 mm2	[138]
GT200 Tesla	1,400,000,000	2008	Nvidia	TSMC	65 nm	576 mm2	[153][138]
GT218 Tesla	260,000,000	2009	Nvidia	TSMC	40 nm	57 mm2	[138]
GT216 Tesla	486,000,000	2009	Nvidia	TSMC	40 nm	100 mm2
GT215 Tesla	727,000,000	2009	Nvidia	TSMC	40 nm	144 mm2
RV740	826,000,000	2009	ATI	TSMC	40 nm	137 mm2
Juniper RV840	1,040,000,000	2009	ATI	TSMC	40 nm	166 mm2
Cypress RV870	2,154,000,000	2009	ATI	TSMC	40 nm	334 mm2	[154]
Cedar RV810	292,000,000	2010	AMD (formerly ATI)	TSMC	40 nm	59 mm2	[138]
Redwood RV830	627,000,000	2010	AMD	TSMC	40 nm	104 mm2
GF106 Fermi	1,170,000,000	2010	Nvidia	TSMC	40 nm	238 mm2
Barts RV940	1,700,000,000	2010	AMD	TSMC	40 nm	255 mm2
Cayman RV970	2,640,000,000	2010	AMD	TSMC	40 nm	389 mm2
GF100 Fermi	3,200,000,000	March 2010	Nvidia	TSMC	40 nm	526 mm2	[155]
GF110 Fermi	3,000,000,000	November 2010	Nvidia	TSMC	40 nm	520 mm2	[155]
GF119 Fermi	292,000,000	2011	Nvidia	TSMC	40 nm	79 mm2	[138]
Caicos RV910	370,000,000	2011	AMD	TSMC	40 nm	67 mm2
GF108 Fermi	585,000,000	2011	Nvidia	TSMC	40 nm	116 mm2
Turks RV930	716,000,000	2011	AMD	TSMC	40 nm	118 mm2
GF104 Fermi	1,950,000,000	2011	Nvidia	TSMC	40 nm	332 mm2
Tahiti	4,312,711,873	2011	AMD	TSMC	28 nm	365 mm2	[156]
GK107 Kepler	1,270,000,000	2012	Nvidia	TSMC	28 nm	118 mm2	[138]
Cape Verde	1,500,000,000	2012	AMD	TSMC	28 nm	123 mm2
GK106 Kepler	2,540,000,000	2012	Nvidia	TSMC	28 nm	221 mm2
Pitcairn	2,800,000,000	2012	AMD	TSMC	28 nm	212 mm2
GK104 Kepler	3,540,000,000	2012	Nvidia	TSMC	28 nm	294 mm2	[157]
GK110 Kepler	7,080,000,000	2012	Nvidia	TSMC	28 nm	561 mm2	[158][159]
Oland	1,040,000,000	2013	AMD	TSMC	28 nm	90 mm2	[138]
Bonaire	2,080,000,000	2013	AMD	TSMC	28 nm	160 mm2
Durango (Xbox One)	5,000,000,000	2013	AMD	TSMC	28 nm	363 mm2	[160]
Liverpool (PlayStation 4)	Unknown	2013	AMD	TSMC	28 nm	348 mm2	[161]
Hawaii	6,300,000,000	2013	AMD	TSMC	28 nm	438 mm2	[138]
GM107 Maxwell	1,870,000,000	2014	Nvidia	TSMC	28 nm	148 mm2
GM206 Maxwell	2,940,000,000	2014	Nvidia	TSMC	28 nm	228 mm2
Tonga	5,000,000,000	2014	AMD	TSMC, GlobalFoundries	28 nm	366 mm2
GM204 Maxwell	5,200,000,000	2014	Nvidia	TSMC	28 nm	398 mm2
GM200 Maxwell	8,000,000,000	2015	Nvidia	TSMC	28 nm	601 mm2
Fiji	8,900,000,000	2015	AMD	TSMC	28 nm	596 mm2
Polaris 11 "Baffin"	3,000,000,000	2016	AMD	Samsung, GlobalFoundries	14 nm	123 mm2	[138][162]
GP108 Pascal	4,400,000,000	2016	Nvidia	TSMC	16 nm	200 mm2	[138]
Durango 2 (Xbox One S)	5,000,000,000	2016	AMD	TSMC	16 nm	240 mm2	[163]
Neo (PlayStation 4 Pro)	5,700,000,000	2016	AMD	TSMC	16 nm	325 mm2	[164]
Polaris 10 "Ellesmere"	5,700,000,000	2016	AMD	Samsung, GlobalFoundries	14 nm	232 mm2	[165]
GP104 Pascal	7,200,000,000	2016	Nvidia	TSMC	16 nm	314 mm2	[138]
GP100 Pascal	15,300,000,000	2016	Nvidia	TSMC, Samsung	16 nm	610 mm2	[166]
GP108 Pascal	1,850,000,000	2017	Nvidia	Samsung	14 nm	74 mm2	[138]
Polaris 12 "Lexa"	2,200,000,000	2017	AMD	Samsung, GlobalFoundries	14 nm	101 mm2	[138][162]
GP107 Pascal	3,300,000,000	2017	Nvidia	Samsung	14 nm	132 mm2	[138]
Scorpio (Xbox One X)	7,000,000,000	2017	AMD	TSMC	16 nm	359 mm2	[167]
GP102 Pascal	11,800,000,000	2017	Nvidia	TSMC, Samsung	16 nm	471 mm2	[138]
Vega 10	12,500,000,000	2017	AMD	Samsung, GlobalFoundries	14 nm	484 mm2	[168]
GV100 Volta	21,100,000,000	2017	Nvidia	TSMC	12 nm	815 mm2	[3]
TU106 Turing	10,800,000,000	2018	Nvidia	TSMC	12 nm	445 mm2
Vega 20	13,230,000,000	2018	AMD	TSMC	7 nm	331 mm2	[138]
TU104 Turing	13,600,000,000	2018	Nvidia	TSMC	12 nm	545 mm2
TU102 Turing	18,600,000,000	2018	Nvidia	TSMC	12 nm	754 mm2	[169]
TU117 Turing	4,700,000,000	2019	Nvidia	TSMC	12 nm	200 mm2	[170]
TU116 Turing	6,600,000,000	2019	Nvidia	TSMC	12 nm	284 mm2	[171]
Navi 14	6,400,000,000	2019	AMD	TSMC	7 nm	158 mm2	[172]
Navi 10	10,300,000,000	2019	AMD	TSMC	7 nm	251 mm2	[173]

FPGA[edit]

A field-programmable gate array (FPGA) is an integrated circuit designed to be configured by a customer or a designer after manufacturing.

FPGA	MOS transistor count	Date of introduction	Designer	Manufacturer	MOS process	Area	Ref
Virtex	70,000,000	1997	Xilinx
Virtex-E	200,000,000	1998	Xilinx
Virtex-II	350,000,000	2000	Xilinx		130 nm
Virtex-II PRO	430,000,000	2002	Xilinx
Virtex-4	1,000,000,000	2004	Xilinx		90 nm
Virtex-5	1,100,000,000	2006	Xilinx	TSMC	65 nm		[174]
Stratix IV	2,500,000,000	2008	Altera	TSMC	40 nm		[175]
Stratix V	3,800,000,000	2011	Altera	TSMC	28 nm		[176]
Arria 10	5,300,000,000	2014	Altera	TSMC	20 nm		[177]
Virtex-7 2000T	6,800,000,000	2011	Xilinx	TSMC	28 nm		[178]
Stratix 10 SX 2800	17,000,000,000	TBD	Intel	Intel	14 nm	560 mm2	[179][180]
Virtex-Ultrascale VU440	20,000,000,000	Q1 2015	Xilinx	TSMC	20 nm		[181][182]
Virtex-Ultrascale+ VU19P	35,000,000,000	2020	Xilinx	TSMC	16 nm	900 mm2 [e]	[183][184][185]
Versal VC1902	37,000,000,000	2H 2019	Xilinx	TSMC	7 nm		[186][187][188]
Stratix 10 GX 10M	43,300,000,000	Q4 2019	Intel	Intel	14 nm	1400 mm2 [e]	[189][190]

Memory[edit]

See also: Random-access memory § Timeline, flash memory § Timeline, and read-only memory § Timeline

Semiconductor memory is an electronic data storage device, often used as computer memory, implemented on integrated circuits. Nearly all semiconductor memory since the 1970s have used MOSFETs (MOS transistors), replacing earlier bipolar junction transistors. There are two major types of semiconductor memory, random-access memory (RAM) and non-volatile memory (NVM). In turn, there are two major RAM types, dynamic random-access memory (DRAM) and static random-access memory (SRAM), as well as two major NVM types, flash memory and read-only memory (ROM).

Typical CMOS SRAM consists of six transistors per cell. For DRAM, 1T1C, which means one transistor and one capacitor structure, is common. Capacitor charged or not is used to store 1 or 0. For flash memory, the data is stored in floating gate, and the resistance of the transistor is sensed to interpret the data stored. Depending on how fine scale the resistance could be separated, one transistor could store up to 3-bits, meaning eight distinctive level of resistance possible per transistor. However, the fine the scale comes with cost of repeatability therefore reliability. Typically, low grade 2-bits MLC flash is used for flash drives, so a 16 GB flash drive contains roughly 64 billion transistors.

For SRAM chips, six-transistor cells (six transistors per bit) was the standard.[191] DRAM chips during the early 1970s had three-transistor cells (three transistors per bit), before single-transistor cells (one transistor per bit) became standard since the era of 4 Kb DRAM in the mid-1970s.[192][193] In single-level flash memory, each cell contains one floating-gate MOSFET (one transistor per bit),[194] whereas multi-level flash contains 2, 3 or 4 bits per transistor.

Flash memory chips are commonly stacked up in layers, up to 128-layer in production,[195] and 136-layer managed,[196] and available in end-user devices up to 69-layer from manufacturers.

Random-access memory (RAM)

Chip name	Capacity (bits)	RAM type	Transistor count	Date of introduction	Manufacturer(s)	MOS process	Area	Ref
N/A	1-bit	SRAM (cell)	6	1963	Fairchild	N/A	N/A	[197]
N/A	1-bit	DRAM (cell)	1	1965	Toshiba	N/A	N/A	[198][199]
?	8-bit	SRAM (bipolar)	48	1965	SDS, Signetics	?	?	[197]
SP95	16-bit	SRAM (bipolar)	80	1965	IBM	?	?	[200]
TMC3162	16-bit	SRAM (TTL)	96	1966	Transitron	N/A	?	[193]
?	?	SRAM (MOS)	?	1966	NEC	?	?	[192]
	256-bit	DRAM (IC)	256	1968	Fairchild	?	?	[193]
	64-bit	SRAM (PMOS)	384	1968	Fairchild	?	?	[192]
	144-bit	SRAM (NMOS)	864	1968	NEC
1101	256-bit	SRAM (PMOS)	1,536	1969	Intel	12,000 nm	?	[201][202][203]
1102	1 Kb	DRAM (PMOS)	3,072	1970	Intel, Honeywell	?	?	[192]
1103	1 Kb	DRAM (PMOS)	3,072	1970	Intel	8,000 nm	10 mm2	[204][191][205][193]
μPD403	1 Kb	DRAM (NMOS)	3,072	1971	NEC	?	?	[206]
?	2 Kb	DRAM (PMOS)	6,144	1971	General Instrument	?	12.7 mm2	[207]
2102	1 Kb	SRAM (NMOS)	6,144	1972	Intel	?	?	[201][208]
?	8 Kb	DRAM (PMOS)	8,192	1973	IBM	?	18.8 mm2	[207]
5101	1 Kb	SRAM (CMOS)	6,144	1974	Intel	?	?	[201]
2116	16 Kb	DRAM (NMOS)	16,384	1975	Intel	?	?	[209][193]
2114	4 Kb	SRAM (NMOS)	24,576	1976	Intel	?	?	[201][210]
?	4 Kb	SRAM (CMOS)	24,576	1977	Toshiba	?	?	[202]
	64 Kb	DRAM (NMOS)	65,536	1977	NTT	?	35.4 mm2	[207]
		DRAM (VMOS)	65,536	1979	Siemens	?	25.2 mm2	[207]
	16 Kb	SRAM (CMOS)	98,304	1980	Hitachi, Toshiba	?	?	[211]
	256 Kb	DRAM (NMOS)	262,144	1980	NEC	1,500 nm	41.6 mm2	[207]
					NTT	1,000 nm	34.4 mm2	[207]
	64 Kb	SRAM (CMOS)	393,216	1980	Matsushita	?	?	[211]
	288 Kb	DRAM	294,912	1981	IBM	?	25 mm2	[212]
	64 Kb	SRAM (NMOS)	393,216	1982	Intel	1,500 nm	?	[211]
	256 Kb	SRAM (CMOS)	1,572,864	1984	Toshiba	1,200 nm	?	[211][203]
	8 Mb	DRAM	8,388,608	January 5, 1984	Hitachi	?	?	[213][214]
	16 Mb	DRAM (CMOS)	16,777,216	1987	NTT	700 nm	148 mm2	[207]
	4 Mb	SRAM (CMOS)	25,165,824	1990	NEC, Toshiba, Hitachi, Mitsubishi	?	?	[211]
	64 Mb	DRAM (CMOS)	67,108,864	1991	Matsushita, Mitsubishi, Fujitsu, Toshiba	400 nm
KM48SL2000	16 Mb	SDRAM	16,777,216	1992	Samsung	?	?	[215][216]
?	16 Mb	SRAM (CMOS)	100,663,296	1992	Fujitsu, NEC	400 nm	?	[211]
	256 Mb	DRAM (CMOS)	268,435,456	1993	Hitachi, NEC	250 nm
	1 Gb	DRAM	1,073,741,824	January 9, 1995	NEC	250 nm	?	[217][218]
					Hitachi	160 nm	?
		SDRAM	1,073,741,824	1996	Mitsubishi	150 nm	?	[211]
		SDRAM (SOI)	1,073,741,824	1997	Hyundai	?	?	[219]
	4 Gb	DRAM (4-bit)	1,073,741,824	1997	NEC	150 nm	?	[211]
		DRAM	4,294,967,296	1998	Hyundai	?	?	[219]
	8 Gb	SDRAM (DDR3)	8,589,934,592	April 2008	Samsung	50 nm	?	[220]
	16 Gb	SDRAM (DDR3)	17,179,869,184	2008
	32 Gb	SDRAM (HBM2)	34,359,738,368	2016	Samsung	20 nm	?	[221]
	64 Gb	SDRAM (HBM2)	68,719,476,736	2017
	128 Gb	SDRAM (DDR4)	137,438,953,472	2018	Samsung	10 nm	?	[222]
	?	RRAM[223] (3DSoC)[224]	?	2019	Skywater[225]	90 nm	?

Flash memory

Chip name	Capacity (bits)	Flash type	FGMOS transistor count	Date of introduction	Manufacturer(s)	MOS process	Area	Ref
?	256 Kb	NOR	262,144	1985	Toshiba	2,000 nm	?	[211]
	1 Mb	NOR	1,048,576	1989	Seeq, Intel	?
	4 Mb	NAND	4,194,304	1989	Toshiba	1,000 nm
	16 Mb	NOR	16,777,216	1991	Mitsubishi	600 nm
DD28F032SA	32 Mb	NOR	33,554,432	1993	Intel	?	280 mm2	[201][226]
?	64 Mb	NOR	67,108,864	1994	NEC	400 nm	?	[211]
		NAND	67,108,864	1996	Hitachi
	128 Mb	NAND	134,217,728	1996	Samsung, Hitachi	?
	256 Mb	NAND	268,435,456	1999	Hitachi, Toshiba	250 nm
	512 Mb	NAND	536,870,912	2000	Toshiba	?	?	[227]
	1 Gb	2-bit NAND	536,870,912	2001	Samsung	?	?	[211]
					Toshiba, SanDisk	160 nm	?	[228]
	2 Gb	NAND	2,147,483,648	2002	Samsung, Toshiba	?	?	[229][230]
	8 Gb	NAND	8,589,934,592	2004	Samsung	60 nm	?	[229]
	16 Gb	NAND	17,179,869,184	2005	Samsung	50 nm	?	[231]
	32 Gb	NAND	34,359,738,368	2006	Samsung	40 nm
THGAM	128 Gb	Stacked NAND	128,000,000,000	April 2007	Toshiba	56 nm	252 mm2	[232]
THGBM	256 Gb	Stacked NAND	256,000,000,000	2008	Toshiba	43 nm	353 mm2	[233]
THGBM2	1 Tb	Stacked 4-bit NAND	256,000,000,000	2010	Toshiba	32 nm	374 mm2	[234]
KLMCG8GE4A	512 Gb	Stacked 2-bit NAND	256,000,000,000	2011	Samsung	?	192 mm2	[235]
KLUFG8R1EM	4 Tb	Stacked 3-bit V-NAND	1,365,333,333,504	2017	Samsung	?	150 mm2	[236]
eUFS (1 TB)	8 Tb	Stacked 4-bit V-NAND	2,048,000,000,000	2019	Samsung	?	150 mm2	[4][237]

Read-only memory (ROM)

Chip name	Capacity (bits)	ROM type	Transistor count	Date of introduction	Manufacturer(s)	MOS process	Area	Ref
?	?	PROM	?	1956	Arma	N/A	?	[238][239]
	1 Kb	ROM (MOS)	1,024	1965	General Microelectronics	?	?	[240]
3301	1 Kb	ROM (bipolar)	1,024	1969	Intel	N/A	?	[240]
1702	2 Kb	EPROM (MOS)	2,048	1971	Intel	?	15 mm2	[241]
?	4 Kb	ROM (MOS)	4,096	1974	AMD, General Instrument	?	?	[240]
2708	8 Kb	EPROM (MOS)	8,192	1975	Intel	?	?	[201]
?	2 Kb	EEPROM (MOS)	2,048	1976	Toshiba	?	?	[242]
µCOM-43 ROM	16 Kb	PROM (PMOS)	16,000	1977	NEC	?	?	[243]
2716	16 Kb	EPROM (TTL)	16,384	1977	Intel	N/A	?	[204][244]
EA8316F	16 Kb	ROM (NMOS)	16,384	1978	Electronic Arrays	?	436 mm2	[240][245]
2732	32 Kb	EPROM	32,768	1978	Intel	?	?	[201]
2364	64 Kb	ROM	65,536	1978	Intel	?	?	[246]
2764	64 Kb	EPROM	65,536	1981	Intel	3,500 nm	?	[201][211]
27128	128 Kb	EPROM	131,072	1982	Intel	?
27256	256 Kb	EPROM (HMOS)	262,144	1983	Intel	?	?	[201][247]
?	256 Kb	EPROM (CMOS)	262,144	1983	Fujitsu	?	?	[248]
	512 Kb	EPROM (NMOS)	524,288	1984	AMD	1,700 nm	?	[211]
27512	512 Kb	EPROM (HMOS)	524,288	1984	Intel	?	?	[201][249]
?	1 Mb	EPROM (CMOS)	1,048,576	1984	NEC	1,200 nm	?	[211]
	4 Mb	EPROM (CMOS)	4,194,304	1987	Toshiba	800 nm
	16 Mb	EPROM (CMOS)	16,777,216	1990	NEC	600 nm
		MROM	16,777,216	1995	AKM, Hitachi	?	?	[218]

Transistor computers[edit]

Before transistors were invented, relays were used in early computers. The world's first working programmable, fully automatic digital computer,[250] the 1941 Z3 22-bit word length computer, had 2,600 relays, and operated at a clock frequency of about 4–5 Hz. The 1940 Complex Number Computer had fewer than 500 relays,[251] but it was not fully programmable.

The second generation of computers were transistor computers that featured boards filled with discrete transistors and magnetic memory cores. The experimental 1953 48-bit Transistor Computer, developed at the University of Manchester, is widely believed to be the first transistor computer to come into operation anywhere in the world (the prototype had 92 point-contact transistors and 550 diodes).[252] A later version the 1955 machine had a total of 250 junction transistors and 1300 point diodes. The Computer also used a small number of tubes in its clock generator, so it was not the first fully transistorized. The ETL Mark III, developed at the Electrotechnical Laboratory in 1956, may have been the first transistor-based electronic computer using the stored program method. It had about "130 point-contact transistors and about 1,800 germanium diodes were used for logic elements, and these were housed on 300 plug-in packages which could be slipped in and out.[253] The 1958 decimal architecture IBM 7070 was the first transistor computer to be fully programmable. It had about 30,000 alloy-junction germanium transistors and 22,000 germanium diodes, on approximately 14,000 Standard Modular System (SMS) cards. The 1959 MOBIDIC, short for "MOBIle DIgital Computer", at 12,000 pounds (6.0 short tons) mounted in the trailer of a semi-trailer truck, was a transistorized computer for battlefield data.

The third generation of computers used integrated circuits (ICs).[254] The 1962 15-bit Apollo Guidance Computer used "about 4,000 "Type-G" (3-input NOR gate) circuits" for about 12,000 transistors plus 32,000 resistors.[255] The first commercial IC-based computer was the IBM System/360 in 1964.[254] The 1965 12-bit PDP-8 CPU had 1409 transistors and over 10,000 diodes. It was not a microprocessor, as it used discrete transistors on many cards; but later microprocessors, such as the Intersil 6100 reimplemented it, see below.[256]

The next generation of computers were the microcomputers, also known as home computers or personal computers (PC), which used MOS microprocessors, in the 1970s. This list includes early transistorized computers (second generation) and IC-based computers (third generation) from the 1950s and 1960s.

Computer	Transistor count	Year	Manufacturer	Notes	Ref
Transistor Computer	92	1953	University of Manchester	Point-contact transistors	[252]
TRADIC	700	1954	Bell Labs	Point-contact transistors	[252]
Transistor Computer (full size)	250	1955	University of Manchester	Discrete point-contact transistors	[252]
ETL Mark III	130	1956	Electrotechnical Laboratory	Point-contact transistors	[252][253]
Metrovick 950	200	1956	Metropolitan-Vickers	Discrete junction transistors
NEC NEAC-2201	600	1958	NEC	Germanium transistors	[257]
Hitachi MARS-1	1,000	1958	Hitachi		[258]
IBM 7070	30,000	1958	IBM	Alloy-junction germanium transistors	[259]
Matsushita MADIC-I	400	1959	Matsushita	Bipolar transistors	[260]
NEC NEAC-2203	2,579	1959	NEC		[261]
Toshiba TOSBAC-2100	5,000	1959	Toshiba		[262]
IBM 7090	50,000	1959	IBM	Discrete germanium transistors	[263]
PDP-1	2,700	1959	Digital Equipment Corporation	Discrete transistors
Mitsubishi MELCOM 1101	3,500	1960	Mitsubishi	Germanium transistors	[264]
M18 FADAC	1,600	1960	Autonetics	Discrete transistors
D-17B	1,521	1962	Autonetics	Discrete transistors
NEC NEAC-L2	16,000	1964	NEC	Ge transistors	[265]
IBM System/360	?	1964	IBM	Integrated circuits
PDP-8/I	1409	1968	Digital Equipment Corporation	74 series TTL circuits
Apollo Guidance Computer Block II	12,300	1966	Raytheon / MIT Instrumentation Laboratory	4,100 ICs, each containing a 3-transistor, 3-input NOR gate

Logic functions[edit]

Transistor count for generic logic functions is based on static CMOS implementation.[266]

Function	Transistor count	Ref
NOT	2
Buffer	4
NAND 2-input	4
NOR 2-input	4
AND 2-input	6
OR 2-input	6
NAND 3-input	6
NOR 3-input	6
XOR 2-input	6
XNOR 2-input	8
MUX 2-input with TG	6
MUX 4-input with TG	18
NOT MUX 2-input	8
MUX 4-input	24
1-bit adder full	28
1-bit adder–subtractor	48
AND-OR-INVERT	6	[267]
Latch, D gated	8
Flip-flop, edge triggered dynamic D with reset	12
8-bit multiplier	3,000
16-bit multiplier	9,000
32-bit multiplier	21,000	[268]
small-scale integration	2–100	[269]
medium-scale integration	100–500	[269]
large-scale integration	500–20,000	[269]
very-large-scale integration	20,000–1,000,000	[269]
ultra-large scale integration	>1,000,000

Parallel systems[edit]

Historically, each processing element in earlier parallel systems—like all CPUs of that time—was a serial computer built out of multiple chips. As transistor counts per chip increases, each processing element could be built out of fewer chips, and then later each multi-core processor chip could contain more processing elements.[270]

Goodyear MPP: (1983?) 8 pixel processors per chip, 3,000 to 8,000 transistors per chip.[270]

Brunel University Scape (single-chip array-processing element): (1983) 256 pixel processors per chip, 120,000 to 140,000 transistors per chip.[270]

Cell Broadband Engine: (2006) with 9 cores per chip, had 234 million transistors per chip.[271]

Other devices[edit]

Device type	Device name	Transistor count	Date of introduction	Designer(s)	Manufacturer(s)	MOS process	Area	Ref
Deep learning engine / IPU[f]	Colossus GC2	23,600,000,000	2018	Graphcore	TSMC	16 nm	~800 mm2	[272][273][better source needed]
Deep learning engine / IPU	Wafer Scale Engine	1,200,000,000,000	2019	Cerebras	TSMC	16 nm	46,225 mm2	[5][6][7][8]

Transistor density

Semiconductor device fabrication
MOSFET scaling (process nodes) 10 µm – 1971 6 µm – 1974 3 µm – 1977 1.5 µm – 1981 1 µm – 1984 800 nm – 1987 600 nm – 1990 350 nm – 1993 250 nm – 1996 180 nm – 1999 130 nm – 2001 90 nm – 2003 65 nm – 2005 45 nm – 2007 32 nm – 2009 22 nm – 2012 14 nm – 2014 10 nm – 2016 7 nm – 2018 5 nm – ~2020 Future 3 nm – ~2021 2 nm – ~2024
Half-nodes Density CMOS Device (multi-gate) Moore's law Semiconductor Nanoelectronics
v t e

The transistor density is the number of transistors that are fabricated per unit area, typically measured in terms of the number of transistors per square millimeter (mm2). The transistor density usually correlates with the gate length of a semiconductor node (also known as a semiconductor manufacturing process), typically measured in nanometers (nm). As of 2019, the semiconductor node with the highest transistor density is TSMC's 5 nanometer node, with 171.3 million transistors per square millimeter.[274]

MOSFET nodes[edit]

Further information: List of semiconductor scale examples

Semiconductor nodes

Node name	Transistor density (transistors/mm2)	Production year	Process	MOSFET	Manufacturer(s)	Ref
?	?	1960	20,000 nm	PMOS	Bell Labs	[275][276]
?	?	1960	20,000 nm	NMOS
?	?	1963	?	CMOS	Fairchild	[18]
?	?	1964	?	PMOS	General Microelectronics	[277]
?	?	1968	20,000 nm	CMOS	RCA	[278]
?	?	1969	12,000 nm	PMOS	Intel	[211][203]
?	?	1970	10,000 nm	CMOS	RCA	[278]
?	300	1970	8,000 nm	PMOS	Intel	[205][193]
?	?	1971	10,000 nm	PMOS	Intel	[279]
?	480	1971	?	PMOS	General Instrument	[207]
?	?	1973	?	NMOS	Texas Instruments	[207]
?	220	1973	?	NMOS	Mostek	[207]
?	?	1973	7,500 nm	NMOS	NEC	[24][23]
?	?	1973	6,000 nm	PMOS	Toshiba	[25][280]
?	?	1976	5,000 nm	NMOS	Hitachi, Intel	[207]
?	?	1976	5,000 nm	CMOS	RCA
?	?	1976	4,000 nm	NMOS	Zilog
?	?	1976	3,000 nm	NMOS	Intel	[281]
?	1,850	1977	?	NMOS	NTT	[207]
?	?	1978	3,000 nm	CMOS	Hitachi	[282]
?	?	1978	2,500 nm	NMOS	Texas Instruments	[207]
?	?	1978	2,000 nm	NMOS	NEC, NTT
?	2,600	1979	?	VMOS	Siemens
?	7,280	1979	1,000 nm	NMOS	NTT
?	7,620	1980	1,000 nm	NMOS	NTT
?	?	1983	2,000 nm	CMOS	Toshiba	[211]
?	?	1983	1,500 nm	CMOS	Intel	[207]
?	?	1983	1,200 nm	CMOS	Intel
?	?	1984	800 nm	CMOS	NTT
?	?	1987	700 nm	CMOS	Fujitsu
?	?	1989	600 nm	CMOS	Mitsubishi, NEC, Toshiba	[211]
?	?	1989	500 nm	CMOS	Hitachi, Mitsubishi, NEC, Toshiba
?	?	1991	400 nm	CMOS	Matsushita, Mitsubishi, Fujitsu, Toshiba
?	?	1993	350 nm	CMOS	Sony
?	?	1993	250 nm	CMOS	Hitachi, NEC
3LM	32,000	1994	350 nm	CMOS	NEC	[135]
?	?	1995	160 nm	CMOS	Hitachi	[211]
?	?	1996	150 nm	CMOS	Mitsubishi
TSMC 180 nm	?	1998	180 nm	CMOS	TSMC	[283]
CS80	?	1999	180 nm	CMOS	Fujitsu	[284]
?	?	1999	180 nm	CMOS	Intel, Sony, Toshiba	[201][60]
CS85	?	1999	170 nm	CMOS	Fujitsu	[285]
Samsung 140 nm	?	1999	140 nm	CMOS	Samsung	[211]
?	?	2001	130 nm	CMOS	Fujitsu, Intel	[284][201]
Samsung 100 nm	?	2001	100 nm	CMOS	Samsung	[211]
?	?	2002	90 nm	CMOS	Sony, Toshiba, Samsung	[60][229]
CS100	?	2003	90 nm	CMOS	Fujitsu	[284]
Intel 90 nm	1,450,000	2004	90 nm	CMOS	Intel	[286][201]
Samsung 80 nm	?	2004	80 nm	CMOS	Samsung	[287]
?	?	2004	65 nm	CMOS	Fujitsu, Toshiba	[288]
Samsung 60 nm	?	2004	60 nm	CMOS	Samsung	[229]
TSMC 45 nm	?	2004	45 nm	CMOS	TSMC
Elpida 90 nm	?	2005	90 nm	CMOS	Elpida Memory	[289]
CS200	?	2005	65 nm	CMOS	Fujitsu	[290][284]
Samsung 50 nm	?	2005	50 nm	CMOS	Samsung	[231]
Intel 65 nm	2,080,000	2006	65 nm	CMOS	Intel	[286]
Samsung 40 nm	?	2006	40 nm	CMOS	Samsung	[231]
Toshiba 56 nm	?	2007	56 nm	CMOS	Toshiba	[232]
Matsushita 45 nm	?	2007	45 nm	CMOS	Matsushita	[70]
Intel 45 nm	3,300,000	2008	45 nm	CMOS	Intel	[291]
Toshiba 43 nm	?	2008	43 nm	CMOS	Toshiba	[233]
TSMC 40 nm	?	2008	40 nm	CMOS	TSMC	[292]
Toshiba 32 nm	?	2009	32 nm	CMOS	Toshiba	[293]
Intel 32 nm	7,500,000	2010	32 nm	CMOS	Intel	[291]
?	?	2010	20 nm	CMOS	Hynix, Samsung	[294][231]
Intel 22 nm	15,300,000	2012	22 nm	CMOS	Intel	[291]
IMFT 20 nm	?	2012	20 nm	CMOS	IMFT	[295]
Toshiba 19 nm	?	2012	19 nm	CMOS	Toshiba
Hynix 16 nm	?	2013	16 nm	FinFET	SK Hynix	[294]
TSMC 16 nm	28,880,000	2013	16 nm	FinFET	TSMC	[296][297]
Samsung 10 nm	51,820,000	2013	10 nm	FinFET	Samsung	[298][299]
Intel 14 nm	37,500,000	2014	14 nm	FinFET	Intel	[291]
14LP	32,940,000	2015	14 nm	FinFET	Samsung	[298]
TSMC 10 nm	52,510,000	2016	10 nm	FinFET	TSMC	[296][300]
12LP	36,710,000	2017	12 nm	FinFET	GlobalFoundries, Samsung	[162]
N7FF	96,500,000	2017	7 nm	FinFET	TSMC	[301][302][303]
8LPP	61,180,000	2018	8 nm	FinFET	Samsung	[298]
7LPE	95,300,000	2018	7 nm	FinFET	Samsung	[302]
Intel 10 nm	100,760,000	2018	10 nm	FinFET	Intel	[304]
5LPE	126,530,000	2018	5 nm	FinFET	Samsung	[305][306]
N7FF+	113,900,000	2019	7 nm	FinFET	TSMC	[301][302]
CLN5FF	171,300,000	2019	5 nm	FinFET	TSMC	[274]
TSMC 3 nm	?	?	3 nm	?	TSMC	[307]
Samsung 3 nm	?	?	3 nm	GAAFET	Samsung	[308]

See also[edit]

• Gate count, an alternate metric
• Dennard scaling
• Electronics industry
• Integrated circuit
• List of best-selling electronic devices
• List of semiconductor scale examples
• MOSFET
• Semiconductor
• Semiconductor device
• Semiconductor device fabrication
• Semiconductor industry
• Transistor

Notes[edit]

1. ^ Declassified 1998
2. ^ 3,510 without depletion mode pull-up transistors
3. ^ 6,813 without depletion mode pull-up transistors
4. ^ 3,900,000,000 core chiplet die, 2,090,000,000 I/O die
5. ^ Jump up to:a b Estimate
6. ^ "Intelligence Processing Unit"

References[edit]

1. ^ Jump up to:a b c Broekhuijsen, Niels (October 23, 2019). "AMD's 64-Core EPYC and Ryzen CPUs Stripped: A Detailed Inside Look". Retrieved October 24, 2019.
2. ^ Jump up to:a b c Mujtaba, Hassan (October 22, 2019). "AMD 2nd Gen EPYC Rome Processors Feature A Gargantuan 39.54 Billion Transistors, IO Die Pictured in Detail". Retrieved October 24, 2019.
3. ^ Jump up to:a b Durant, Luke; Giroux, Olivier; Harris, Mark; Stam, Nick (May 10, 2017). "Inside Volta: The World's Most Advanced Data Center GPU". Nvidia developer blog.
4. ^ Jump up to:a b Manners, David (January 30, 2019). "Samsung makes 1TB flash eUFS module". Electronics Weekly. Retrieved June 23,2019.
5. ^ Jump up to:a b Hruska, Joel (August 2019). "Cerebras Systems Unveils 1.2 Trillion Transistor Wafer-Scale Processor for AI". extremetech.com. Retrieved September 6, 2019.
6. ^ Jump up to:a b Feldman, Michael (August 2019). "Machine Learning chip breaks new ground with waferscale integration". nextplatform.com. Retrieved September 6, 2019.
7. ^ Jump up to:a b Cutress, Ian (August 2019). "Hot Chips 31 Live Blogs: Cerebras' 1.2 Trillion Transistor Deep Learning Processor". anandtech.com. Retrieved September 6, 2019.
8. ^ Jump up to:a b "A Look at Cerebras Wafer-Scale Engine: Half Square Foot Silicon Chip". WikiChip Fuse. November 16, 2019. Retrieved December 2, 2019.
9. ^ "John Gustafson's answer to How many individual transistors are in the world's most powerful supercomputer?". Quora. Retrieved August 22, 2019.
10. ^ "13 Sextillion & Counting: The Long & Winding Road to the Most Frequently Manufactured Human Artifact in History". Computer History Museum. April 2, 2018. Retrieved July 28, 2019.
11. ^ Jump up to:a b c Moskowitz, Sanford L. (2016). Advanced Materials Innovation: Managing Global Technology in the 21st century. John Wiley & Sons. pp. 165–168. ISBN 9780470508923.
12. ^ "1960 - Metal Oxide Semiconductor (MOS) Transistor Demonstrated". The Silicon Engine. Computer History Museum.
13. ^ "Who Invented the Transistor?". Computer History Museum. December 4, 2013. Retrieved July 20, 2019.
14. ^ "Transistors Keep Moore's Law Alive". EETimes. December 12, 2018. Retrieved July 18, 2019.
15. ^ Jump up to:a b "Tortoise of Transistors Wins the Race – CHM Revolution". Computer History Museum. Retrieved July 22, 2019.
16. ^ Hittinger, William C. (1973). "Metal-Oxide-Semiconductor Technology". Scientific American. 229 (2): 48–59. Bibcode:1973SciAm.229b..48H. doi:10.1038/scientificamerican0873-48. ISSN 0036-8733. JSTOR 24923169.
17. ^ Bassett, Ross Knox (2007). To the Digital Age: Research Labs, Start-up Companies, and the Rise of MOS Technology. Johns Hopkins University Press. p. 22. ISBN 9780801886393.
18. ^ Jump up to:a b "1963: Complementary MOS Circuit Configuration is Invented". Computer History Museum. Retrieved July 6, 2019.
19. ^ Jump up to:a b "1971: Microprocessor Integrates CPU Function onto a Single Chip". The Silicon Engine. Computer History Museum. Retrieved September 4, 2019.
20. ^ Jump up to:a b Holt, Ray. "World's First Microprocessor". Retrieved March 5,2016. 1st fully integrated chip set microprocessor
21. ^ Jump up to:a b "Alpha 21364 - Microarchitectures - Compaq - WikiChip". en.wikichip.org. Retrieved September 8, 2019.
22. ^ Ryoichi Mori; Hiroaki Tajima; Morihiko Tajima; Yoshikuni Okada (October 1977). "Microprocessors in Japan". Euromicro Newsletter. 3 (4): 50–7. doi:10.1016/0303-1268(77)90111-0.
23. ^ Jump up to:a b "NEC 751 (uCOM-4)". The Antique Chip Collector's Page. Archived from the original on May 25, 2011. Retrieved June 11,2010.
24. ^ Jump up to:a b "1970s: Development and evolution of microprocessors"(PDF). Semiconductor History Museum of Japan. Retrieved June 27,2019.
25. ^ Jump up to:a b "1973: 12-bit engine-control microprocessor (Toshiba)"(PDF). Semiconductor History Museum of Japan. Retrieved June 27,2019.
26. ^ "Low Bandwidth Timeline – Semiconductor". Texas Instruments. Retrieved June 22, 2016.
27. ^ "The MOS 6502 and the Best Layout Guy in the World". research.swtch.com. January 3, 2011. Retrieved September 3, 2019.
28. ^ "Digital History: ZILOG Z8000 (APRIL 1979)". OLD-COMPUTERS.COM : The Museum. Retrieved June 19, 2019.
29. ^ "Chip Hall of Fame: Motorola MC68000 Microprocessor". IEEE Spectrum. Institute of Electrical and Electronics Engineers. June 30, 2017. Retrieved June 19, 2019.
30. ^ Microprocessors: 1971 to 1976 Christiansen
31. ^ "Microprocessors 1976 to 1981". weber.edu. Retrieved August 9,2014.
32. ^ "W65C816S 16-bit Core". www.westerndesigncenter.com. Retrieved September 12, 2017.
33. ^ Jump up to:a b c d e Demone, Paul (November 9, 2000). "ARM's Race to World Domination". real world technologies. Retrieved July 20, 2015.
34. ^ Hand, Tom. "The Harris RTX 2000 Microcontroller" (PDF). mpeforth.com. Retrieved August 9, 2014.
35. ^ "Forth chips list". UltraTechnology. March 15, 2001. Retrieved August 9, 2014.
36. ^ Koopman, Philip J. (1989). "4.4 Architecture of the Novix NC4016". Stack Computers: the new wave. Ellis Horwood Series in Computers and Their Applications. Carnegie Mellon University. ISBN 978-0745804187. Retrieved August 9, 2014.
37. ^ "Fujitsu SPARC". cpu-collection.de. Retrieved June 30, 2019.
38. ^ Jump up to:a b Kimura S, Komoto Y, Yano Y (1988). "Implementation of the V60/V70 and its FRM function". IEEE Micro. 8 (2): 22–36. doi:10.1109/40.527.
39. ^ "VL2333 - VTI - WikiChip". en.wikichip.org. Retrieved August 31,2019.
40. ^ Inayoshi H, Kawasaki I, Nishimukai T, Sakamura K (1988). "Realization of Gmicro/200". IEEE Micro. 8 (2): 12–21. doi:10.1109/40.526.
41. ^ Bosshart, P.; Hewes, C.; Mi-Chang Chang; Kwok-Kit Chau; Hoac, C.; Houston, T.; Kalyan, V.; Lusky, S.; Mahant-Shetti, S.; Matzke, D.; Ruparel, K.; Ching-Hao Shaw; Sridhar, T.; Stark, D. (October 1987). "A 553K-Transistor LISP Processor Chip". IEEE Journal of Solid-State Circuits. 22 (5): 202–3. doi:10.1109/ISSCC.1987.1157084.
42. ^ Fahlén, Lennart E.; Stockholm International Peace Research Institute (1987). "3. Hardware requirements for artificial intelligence § Lisp Machines: TI Explorer". Arms and Artificial Intelligence: Weapon and Arms Control Applications of Advanced Computing. SIPRI Monograph Series. Oxford University Press. p. 57. ISBN 978-0-19-829122-0.
43. ^ Jouppi, Norman P.; Tang, Jeffrey Y. F. (July 1989). "A 20-MIPS Sustained 32-bit CMOS Microprocessor with High Ratio of Sustained to Peak Performance". IEEE Journal of Solid-State Circuits. 24 (5): i. Bibcode:1989IJSSC..24.1348J. CiteSeerX 10.1.1.85.988. doi:10.1109/JSSC.1989.572612. WRL Research Report 89/11.
44. ^ "The CPU shack museum". CPUshack.com. May 15, 2005. Retrieved August 9, 2014.
45. ^ Jump up to:a b c "Intel i960 Embedded Microprocessor". National High Magnetic Field Laboratory. Florida State University. March 3, 2003. Retrieved June 29, 2019.
46. ^ Venkatasawmy, Rama (2013). The Digitization of Cinematic Visual Effects: Hollywood's Coming of Age. Rowman & Littlefield. p. 198. ISBN 9780739176214.
47. ^ "SH Microprocessor Leading the Nomadic Era" (PDF). Semiconductor History Museum of Japan. Retrieved June 27, 2019.
48. ^ "SH2: A Low Power RISC Micro for Consumer Applications"(PDF). Hitachi. Retrieved June 27, 2019.
49. ^ "HARP-1: A 120 MHz Superscalar PA-RISC Processor" (PDF). Hitachi. Retrieved June 19, 2019.
50. ^ "ARM7 Statistics". Poppyfields.net. May 27, 1994. Retrieved August 9, 2014.
51. ^ "Forth Multiprocessor Chip MuP21". www.ultratechnology.com. Retrieved September 6, 2019. MuP21 has a 21-bit CPU core, a memory coprocessor, and a video coprocessor
52. ^ Jump up to:a b "F21 CPU". www.ultratechnology.com. Retrieved September 6, 2019. F21 offers video I/O, analog I/O, serial network I/O, and a parallel I/O port on chip. F21 has a transistor count of about 15,000 vs about 7,000 for MuP21.
53. ^ "Intel Pentium Pro 180". hw-museum.cz. Retrieved September 8, 2019.
54. ^ "PC Guide Intel Pentium Pro ("P6")". PCGuide.com. April 17, 2001. Retrieved August 9, 2014.[dead link]
55. ^ Jump up to:a b "Remembering the Sega Dreamcast". Bit-Tech. September 29, 2009. Retrieved June 18, 2019.
56. ^ "Entertainment Systems and High-Performance Processor SH-4" (PDF). Hitachi Review. Hitachi. 48 (2): 58–63. 1999. Retrieved June 27, 2019.
57. ^ Jump up to:a b Hagiwara, Shiro; Oliver, Ian (November–December 1999). "Sega Dreamcast: Creating a Unified Entertainment World". IEEE Micro. IEEE Computer Society. 19 (6): 29–35. doi:10.1109/40.809375. Retrieved June 27, 2019.
58. ^ Ulf Samuelsson. "Transistor count of common uCs?". www.embeddedrelated.com. Retrieved September 8, 2019. IIRC, The AVR core is 12,000 gates, and the megaAVR core is 20,000 gates. Each gate is 4 transistors. The chip is considerably larger since the memory uses quite a lot.
59. ^ Jump up to:a b Hennessy, John L.; Patterson, David A. (May 29, 2002). Computer Architecture: A Quantitative Approach (3 ed.). Morgan Kaufmann. p. 491. ISBN 978-0-08-050252-6. Retrieved April 9,2013.
60. ^ Jump up to:a b c d "EMOTION ENGINE® AND GRAPHICS SYNTHESIZER USED IN THE CORE OF PLAYSTATION® BECOME ONE CHIP"(PDF). Sony. April 21, 2003. Retrieved June 26, 2019.
61. ^ Jump up to:a b Diefendorff, Keith (April 19, 1999). "Sony's Emotionally Charged Chip: Killer Floating-Point "Emotion Engine" To Power PlayStation 2000" (PDF). Microprocessor Report. 13 (5). Retrieved June 19,2019.
62. ^ Jump up to:a b c "NVIDIA GeForce 7800 GTX GPU Review". PC Perspective. June 22, 2005. Retrieved June 18, 2019.
63. ^ Ando, H.; Yoshida, Y.; Inoue, A.; Sugiyama, I.; Asakawa, T.; Morita, K.; Muta, T.; otokurumada, T.; Okada, S.; Yamashita, H.; Satsukawa, Y.; Konmoto, A.; Yamashita, R.; Sugiyama, H. (2003). A 1.3GHz fifth eneration SPARC64 microprocessor. Design Automation Conference. pp. 702–705. doi:10.1145/775832.776010. ISBN 1-58113-688-9.
64. ^ Krewell, Kevin (21 October 2002). "Fujitsu's SPARC64 V Is Real Deal". Microprocessor Report.
65. ^ Fujitsu Limited (August 2004). SPARC64 V Processor For UNIX Server.
66. ^ "A Glimpse Inside The Cell Processor". Gamasutra. July 13, 2006. Retrieved June 19, 2019.
67. ^ "PRESS KIT — Dual-core Intel Itanium Processor". Intel. Retrieved August 9, 2014.
68. ^ Jump up to:a b Toepelt, Bert (January 8, 2009). "AMD Phenom II X4: 45nm Benchmarked — The Phenom II And AMD's Dragon Platform". TomsHardware.com. Retrieved August 9, 2014.
69. ^ "ARM (Advanced RISC Machines) Processors". EngineersGarage.com. Retrieved August 9, 2014.
70. ^ Jump up to:a b "Panasonic starts to sell a New-generation UniPhier System LSI". Panasonic. October 10, 2007. Retrieved July 2, 2019.
71. ^ "SPARC64 VI Extensions" page 56, Fujitsu Limited, Release 1.3, 27 March 2007
72. ^ Morgan, Timothy Prickett (17 July 2008). "Fujitsu and Sun Flex Their Quads with New Sparc Server Lineup". The Unix Guardian, Vol. 8, No. 27.
73. ^ Takumi Maruyama (2009). SPARC64 VIIIfx: Fujitsu's New Generation Octo Core Processor for PETA Scale computing(PDF). Proceedings of Hot Chips 21. IEEE Computer Society. Archived from the original (PDF) on October 8, 2010. Retrieved June 30, 2019.
74. ^ Stokes, Jon (February 10, 2010). "Sun's 1 billion-transistor, 16-core Niagara 3 processor". ArsTechnica.com. Retrieved August 9,2014.
75. ^ "IBM to Ship World's Fastest Microprocessor". IBM. September 1, 2010. Retrieved August 9, 2014.
76. ^ "Intel to deliver first computer chip with two billion transistors". AFP. February 5, 2008. Archived from the original on May 20, 2011. Retrieved February 5, 2008.
77. ^ "Intel Previews Intel Xeon 'Nehalem-EX' Processor." May 26, 2009. Retrieved on May 28, 2009.
78. ^ Morgan, Timothy Prickett (November 21, 2011), "Fujitsu parades 16-core Sparc64 super stunner", The Register, retrieved December 8, 2011
79. ^ Angelini, Chris (November 14, 2011). "Intel Core i7-3960X Review: Sandy Bridge-E And X79 Express". TomsHardware.com. Retrieved August 9, 2014.
80. ^ "IDF2012 Mark Bohr, Intel Senior Fellow" (PDF).
81. ^ "Images of SPARC64" (PDF). fujitsu.com. Retrieved August 29,2017.
82. ^ "Intel's Atom Architecture: The Journey Begins". AnandTech. Retrieved April 4, 2010.
83. ^ "Intel Xeon Phi SE10X". TechPowerUp. Retrieved July 20, 2015.
84. ^ Shimpi, Lal. "The Haswell Review: Intel Core i7-4770K & i5-4670K Tested". anandtech. Retrieved November 20, 2014.
85. ^ "Dimmick, Frank (August 29, 2014). "Intel Core i7 5960X Extreme Edition Review". Overclockers Club. Retrieved August 29, 2014.
86. ^ "Apple A8X". NotebookCheck. Retrieved July 20, 2015.
87. ^ "Intel Readying 15-core Xeon E7 v2". AnandTech. Retrieved August 9, 2014.
88. ^ "Intel Xeon E5-2600 v3 Processor Overview: Haswell-EP Up to 18 Cores". pcper. Retrieved January 29, 2015.
89. ^ "Intel's Broadwell-U arrives aboard 15W, 28W mobile processors". TechReport. Retrieved January 5, 2015.
90. ^ http://www.enterprisetech.com/2014/08/13/oracle-cranks-cores-32-sparc-m7-chip/
91. ^ "Qualcomm Snapdragon 835 (8998)". NotebookCheck. Retrieved September 23, 2017.
92. ^ Takahashi, Dean (January 3, 2017). "Qualcomm's Snapdragon 835 will debut with 3 billion transistors and a 10nm manufacturing process". VentureBeat.
93. ^ "Broadwell-E: Intel Core i7-6950X, 6900K, 6850K & 6800K Review". Tom's Hardware. May 30, 2016. Retrieved April 12,2017.
94. ^ "The Broadwell-E Review". PC Gamer. July 8, 2016. Retrieved April 12, 2017.
95. ^ "HUAWEI TO UNVEIL KIRIN 970 SOC WITH AI UNIT, 5.5 BILLION TRANSISTORS AND 1.2 GBPS LTE SPEED AT IFA 2017". firstpost.com. September 1, 2017. Retrieved November 18,2018.
96. ^ "Broadwell-EP Architecture - Intel Xeon E5-2600 v4 Broadwell-EP Review". Tom's Hardware. March 31, 2016. Retrieved April 4,2016.
97. ^ "About the ZipCPU". zipcpu.com. Retrieved September 10,2019. As of ORCONF, 2016, the ZipCPU used between 1286 and 4926 6-LUTs, depending upon how it is configured.
98. ^ "Qualcomm Snapdragon 1000 for laptops could pack 8.5 billion transistors". techradar. Retrieved September 23, 2017.
99. ^ "Spotted: Qualcomm Snapdragon 8cx Wafer on 7nm". AnandTech. Retrieved December 6, 2018.
100. ^ Cutress, Ian (February 22, 2017). "AMD Launches Zen". Anandtech.com. Retrieved February 22, 2017.
101. ^ "Ryzen 5 1600 - AMD". Wikichip.org. April 20, 2018. Retrieved December 9, 2018.
102. ^ "Ryzen 5 1600X – AMD". Wikichip.org. October 26, 2018. Retrieved December 9, 2018.
103. ^ "Kirin 970 – HiSilicon". Wikichip. March 1, 2018. Retrieved November 8, 2018.
104. ^ Jump up to:a b Leadbetter, Richard (April 6, 2017). "Inside the next Xbox: Project Scorpio tech revealed". Eurogamer. Retrieved May 3,2017.
105. ^ "Intel Xeon Platinum 8180". TechPowerUp. December 1, 2018. Retrieved December 2, 2018.
106. ^ Lee, Y. "SiFive Freedom SoCs : Industry's First Open Source RISC V Chips" (PDF). HotChips 29 IOT/Embedded.
107. ^ "Documents at Fujitsu" (PDF). fujitsu.com. Retrieved August 29,2017.
108. ^ Schmerer, Kai (November 5, 2018). "iPad Pro 2018: A12X-Prozessor bietet deutlich mehr Leistung". ZDNet.de (in German).
109. ^ "Qualcomm Datacenter Technologies Announces Commercial Shipment of Qualcomm Centriq 2400 – The World's First 10nm Server Processor and Highest Performance Arm-based Server Processor Family Ever Designed". Qualcomm. Retrieved November 9, 2017.
110. ^ "Qualcomm Snapdragon 8180: 7nm SoC SDM1000 With 8.5 Billion Transistors To Challenge Apple A12 Bionic Chipset". dailyhunt. Retrieved September 21, 2018.
111. ^ Yang, Daniel; Wegner, Stacy (September 21, 2018). "Apple iPhone Xs Max Teardown". TechInsights. Retrieved September 21, 2018.
112. ^ "Apple's A12 Bionic is the first 7-nanometer smartphone chip". Engadget. Retrieved September 26, 2018.
113. ^ "Kirin 980 – HiSilicon". Wikichip. November 8, 2018. Retrieved November 8, 2018.
114. ^ Friedman, Alan. "5nm Kirin 1020 SoC tipped for next year's Huawei Mate 40 line". Phone Arena. Retrieved December 23,2019.
115. ^ Frumusanu, Andrei. "The Huawei Mate 30 Pro Review: Top Hardware without Google?". AnandTech. Retrieved January 2,2020.
116. ^ "HiSilicon Kirin 710". Notebookcheck. September 19, 2018. Retrieved November 24, 2018.
117. ^ Zafar, Ramish (October 30, 2018). "Apple's A12X Has 10 Billion Transistors, 90% Performance Boost & 7-Core GPU". Wccftech.
118. ^ Zafar, Ramish (September 10, 2019). "Apple A13 For iPhone 11 Has 8.5 Billion Transistors, Quad-Core GPU". Wccftech. Retrieved September 11, 2019.
119. ^ Introducing iPhone 11 Pro — Apple Youtube Video, retrieved September 11, 2019
120. ^ "Fujitsu began to produce Japan's billions of super-calculations with the strongest ARM processor A64FX". firstxw.com. April 16, 2019. Retrieved June 19, 2019.
121. ^ "Fujitsu Successfully Triples the Power Output of Gallium-Nitride Transistors". Fujitsu. August 22, 2018. Retrieved June 19, 2019.
122. ^ "Hot Chips 30: Nvidia Xavier SoC". fuse.wikichip.org. September 18, 2018. Retrieved December 6, 2018.
123. ^ "AMD Ryzen 9 3900X and Ryzen 7 3700X Review: Zen 2 and 7nm Unleashed". Tom's Hardware. July 7, 2019. Retrieved October 19,2019.
124. ^ CPUs, Arne Verheyde 2019-12-05T19:12:44Z. "Amazon Compares 64-core ARM Graviton2 to Intel's Xeon". Tom's Hardware. Retrieved December 6, 2019.
125. ^ Morgan, Timothy Prickett (December 3, 2019). "Finally: AWS Gives Servers A Real Shot In The Arm". The Next Platform. Retrieved December 6, 2019.
126. ^ Williams, Chris. "Nvidia's Tesla P100 has 15 billion transistors, 21TFLOPS". www.theregister.co.uk. Retrieved August 12, 2019.
127. ^ "Famous Graphics Chips: NEC µPD7220 Graphics Display Controller". IEEE Computer Society. Institute of Electrical and Electronics Engineers. August 22, 2018. Retrieved June 21, 2019.
128. ^ "GPU History: Hitachi ARTC HD63484. The second graphics processor". IEEE Computer Society. Institute of Electrical and Electronics Engineers. Retrieved June 21, 2019.
129. ^ Jump up to:a b "30 Years of Console Gaming". Klinger Photography. August 20, 2017. Retrieved June 19, 2019.
130. ^ Jump up to:a b "Diamond Edge 3D (nVidia NV1+Sega Saturn)". Naver. February 24, 2017. Retrieved June 19, 2019.
131. ^ "Sega Saturn". MAME. Retrieved July 18, 2019.
132. ^ "ASIC CHIPS ARE INDUSTRY'S GAME WINNERS". The Washington Post. September 18, 1995. Retrieved June 19, 2019.
133. ^ "Is it Time to Rename the GPU?". Jon Peddie Research. IEEE Computer Society. July 9, 2018. Retrieved June 19, 2019.
134. ^ "FastForward Sony Taps LSI Logic for PlayStation Video Game CPU Chip". FastForward. Retrieved January 29, 2014.
135. ^ Jump up to:a b "Reality Co-Processor − The Power In Nintendo64" (PDF). Silicon Graphics. August 26, 1997. Retrieved June 18, 2019.
136. ^ "Imagination PowerVR PCX2 GPU". VideoCardz.net. Retrieved June 19, 2019.
137. ^ Jump up to:a b c d e f g h Lilly, Paul (May 19, 2009). "From Voodoo to GeForce: The Awesome History of 3D Graphics". PC Gamer. Retrieved June 19, 2019.
138. ^ Jump up to:a b c d e f g h i j k l m n o p q r s t u v w x y z aa ab ac ad "3D accelerator database". Vintage 3D. Retrieved July 21, 2019.
139. ^ "RIVA128 Datasheet". SGS Thomson Microelectronics. Retrieved July 21, 2019.
140. ^ Jump up to:a b c Singer, Graham (April 3, 2013). "History of the Modern Graphics Processor, Part 2". TechSpot. Retrieved July 21, 2019.
141. ^ Weinberg, Neil (September 7, 1998). "Comeback kid". Forbes. Retrieved June 19, 2019.
142. ^ Charles, Bertie (1998). "Sega's New Dimension". Forbes. Forbes Incorporated. 162 (5–9): 206. The chip, etched in 0.25-micron detail — state-of-the-art for graphics processors — fits 10 million transistors
143. ^ "VideoLogic Neon 250 4MB". VideoCardz.net. Retrieved June 19, 2019.
144. ^ Shimpi, Anand Lal (November 21, 1998). "Fall Comdex '98 Coverage". AnandTech. Retrieved June 19, 2019.
145. ^ "NVIDIA NV10 A3 GPU Specs". TechPowerUp. Retrieved June 19, 2019.
146. ^ IGN Staff (November 4, 2000). "Gamecube Versus PlayStation 2". IGN. Retrieved November 22, 2015.
147. ^ "NVIDIA NV2A GPU Specs". TechPowerUp. Retrieved July 21,2019.
148. ^ "ATI Xenos GPU Specs". TechPowerUp. Retrieved June 21,2019.
149. ^ International, GamesIndustry (July 14, 2005). "TSMC to manufacture X360 GPU". Eurogamer. Retrieved August 22, 2006.
150. ^ "NVIDIA Playstation 3 RSX 65nm Specs". TechPowerUp. Retrieved June 21, 2019.
151. ^ "PS3 Graphics Chip Goes 65nm in Fall". Edge Online. June 26, 2008.
152. ^ "The Radeon HD 4850 & 4870: AMD Wins at $199 and $299". AnandTech.com. Retrieved August 9, 2014.
153. ^ "NVIDIA's 1.4 Billion Transistor GPU: GT200 Arrives as the GeForce GTX 280 & 260". AnandTech.com. Retrieved August 9,2014.
154. ^ "Radeon 5870 specifications". AMD. Retrieved August 9, 2014.
155. ^ Jump up to:a b Glaskowsky, Peter. "ATI and Nvidia face off-obliquely". CNET. Retrieved August 9, 2014.[dead link]
156. ^ Woligroski, Don (December 22, 2011). "AMD Radeon HD 7970". TomsHardware.com. Retrieved August 9, 2014.
157. ^ "Whitepaper: NVIDIA GeForce GTX 680" (PDF). NVIDIA. 2012. Archived from the original (PDF) on April 17, 2012.
158. ^ http://www.nvidia.com/content/PDF/kepler/NVIDIA-Kepler-GK110-Architecture-Whitepaper.pdf
159. ^ Smith, Ryan (November 12, 2012). "NVIDIA Launches Tesla K20 & K20X: GK110 Arrives At Last". AnandTech.
160. ^ "AMD Xbox One GPU". www.techpowerup.com. Retrieved February 5, 2020.
161. ^ "AMD PlayStation 4 GPU". www.techpowerup.com. Retrieved February 5, 2020.
162. ^ Jump up to:a b c Schor, David (July 22, 2018). "VLSI 2018: GlobalFoundries 12nm Leading-Performance, 12LP". WikiChip Fuse. Retrieved May 31, 2019.
163. ^ "AMD Xbox One S GPU". www.techpowerup.com. Retrieved February 5, 2020.
164. ^ "AMD PlayStation 4 Pro GPU". www.techpowerup.com. Retrieved February 5, 2020.
165. ^ Smith, Ryan (June 29, 2016). "The AMD RX 480 Preview". Anandtech.com. Retrieved February 22, 2017.
166. ^ Harris, Mark (April 5, 2016). "Inside Pascal: NVIDIA's Newest Computing Platform". Nvidia developer blog.
167. ^ "AMD Xbox One X GPU". www.techpowerup.com. Retrieved February 5, 2020.
168. ^ "Radeon's next-generation Vega architecture" (PDF).
169. ^ "NVIDIA TURING GPU ARCHITECTURE: Graphics Reinvented" (PDF). Nvidia. 2018. Retrieved June 28, 2019.
170. ^ "NVIDIA GeForce GTX 1660 Ti". www.techpowerup.com. Retrieved February 5, 2020.
171. ^ "NVIDIA GeForce GTX 1650". www.techpowerup.com. Retrieved February 5, 2020.
172. ^ "AMD Radeon RX 5500 XT". www.techpowerup.com. Retrieved February 5, 2020.
173. ^ "AMD Radeon RX 5700 XT". www.techpowerup.com. Retrieved February 5, 2020.
174. ^ "Taiwan Company UMC Delivers 65nm FPGAs to Xilinx." SDA-ASIA Thursday, November 9, 2006.
175. ^ ""Altera's new 40nm FPGAs — 2.5 billion transistors!". pldesignline.com.
176. ^ "Altera unveils 28-nm Stratix V FPGA family". April 20, 2010. Retrieved April 20, 2010.
177. ^ "Design of a High-Density SoC FPGA at 20nm" (PDF). 2014. Retrieved July 16, 2017.
178. ^ Maxfield, Clive (October 2011). "New Xilinx Virtex-7 2000T FPGA provides equivalent of 20 million ASIC gates". EETimes. AspenCore. Retrieved September 4, 2019.
179. ^ Greenhill, D.; Ho, R.; Lewis, D.; Schmit, H.; Chan, K. H.; Tong, A.; Atsatt, S.; How, D.; McElheny, P. (February 2017). "3.3 A 14nm 1GHz FPGA with 2.5D transceiver integration". 2017 IEEE International Solid-State Circuits Conference (ISSCC): 54–55. doi:10.1109/ISSCC.2017.7870257. ISBN 978-1-5090-3758-2.
180. ^ "3.3 A 14nm 1GHz FPGA with 2.5D transceiver integration | DeepDyve". May 17, 2017. Archived from the original on May 17, 2017. Retrieved September 19, 2019.
181. ^ Santarini, Mike (May 2014). "Xilinx Ships Industry's First 20-nm All Programmable Devices" (PDF). Xcell journal. No. 86. Xilinx. p. 14. Retrieved June 3, 2014.
182. ^ Gianelli, Silvia (January 2015). "Xilinx Delivers the Industry's First 4M Logic Cell Device, Offering >50M Equivalent ASIC Gates and 4X More Capacity than Competitive Alternatives". www.xilinx.com. Retrieved August 22, 2019.
183. ^ Sims, Tara (August 2019). "Xilinx Announces the World's Largest FPGA Featuring 9 Million System Logic Cells". www.xilinx.com. Retrieved August 22, 2019.
184. ^ Verheyde, Arne (August 2019). "Xilinx Introduces World's Largest FPGA With 35 Billion Transistors". www.tomshardware.com. Retrieved August 23, 2019.
185. ^ Cutress, Ian (August 2019). "Xilinx Announces World Largest FPGA: Virtex Ultrascale+ VU19P with 9m Cells". www.anandtech.com. Retrieved September 25, 2019.
186. ^ Abazovic, Fuad (May 2019). "Xilinx 7nm Versal taped out last year". Retrieved September 30, 2019.
187. ^ Cutress, Ian (August 2019). "Hot Chips 31 Live Blogs: Xilinx Versal AI Engine". Retrieved September 30, 2019.
188. ^ Krewell, Kevin (August 2019). "Hot Chips 2019 highlights new AI strategies". Retrieved September 30, 2019.
189. ^ Leibson, Steven (November 6, 2019). "Intel announces Intel Stratix 10 GX 10M FPGA, worlds highest capacity with 10.2 million logic elements". Retrieved November 7, 2019.
190. ^ Verheyde, Arne (November 6, 2019). "Intel Introduces World's Largest FPGA With 43.3 Billion Transistors". Retrieved November 7, 2019.
191. ^ Jump up to:a b The DRAM memory of Robert Dennard history-computer.com
192. ^ Jump up to:a b c d "Late 1960s: Beginnings of MOS memory" (PDF). Semiconductor History Museum of Japan. January 23, 2019. Retrieved June 27, 2019.
193. ^ Jump up to:a b c d e f "1970: Semiconductors compete with magnetic cores". Computer History Museum. Retrieved June 19, 2019.
194. ^ "2.1.1 Flash Memory". TU Wien. Retrieved June 20, 2019.
195. ^ Shilov, Anton. "SK Hynix Starts Production of 128-Layer 4D NAND, 176-Layer Being Developed". www.anandtech.com. Retrieved September 16, 2019.
196. ^ "Samsung Begins Production of 100+ Layer Sixth-Generation V-NAND Flash". PC Perspective. August 11, 2019. Retrieved September 16, 2019.
197. ^ Jump up to:a b "1966: Semiconductor RAMs Serve High-speed Storage Needs". Computer History Museum. Retrieved June 19, 2019.
198. ^ "Specifications for Toshiba "TOSCAL" BC-1411". Old Calculator Web Museum. Archived from the original on July 3, 2017. Retrieved May 8, 2018.
199. ^ "Toshiba "Toscal" BC-1411 Desktop Calculator". Old Calculator Web Museum. Archived from the original on May 20, 2007.
200. ^ "IBM first in IC memory". Computer History Museum. Retrieved June 19, 2019.
201. ^ Jump up to:a b c d e f g h i j k l m "A chronological list of Intel products. The products are sorted by date" (PDF). Intel museum. Intel Corporation. July 2005. Archived from the original (PDF) on August 9, 2007. Retrieved July 31, 2007.
202. ^ Jump up to:a b "1970s: SRAM evolution" (PDF). Semiconductor History Museum of Japan. Retrieved June 27, 2019.
203. ^ Jump up to:a b c Pimbley, J. (2012). Advanced CMOS Process Technology. Elsevier. p. 7. ISBN 9780323156806.
204. ^ Jump up to:a b "Intel: 35 Years of Innovation (1968–2003)" (PDF). Intel. 2003. Retrieved June 26, 2019.
205. ^ Jump up to:a b Lojek, Bo (2007). History of Semiconductor Engineering. Springer Science & Business Media. pp. 362–363. ISBN 9783540342588. The i1103 was manufactured on a 6-mask silicon-gate P-MOS process with 8 μm minimum features. The resulting product had a 2,400 µm2 memory cell size, a die size just under 10 mm2, and sold for around $21.
206. ^ "Manufacturers in Japan enter the DRAM market and integration densities are improved" (PDF). Semiconductor History Museum of Japan. Retrieved June 27, 2019.
207. ^ Jump up to:a b c d e f g h i j k l m n Gealow, Jeffrey Carl (August 10, 1990). "Impact of Processing Technology on DRAM Sense Amplifier Design" (PDF). CORE. Massachusetts Institute of Technology. pp. 149–166. Retrieved June 25, 2019.
208. ^ "Silicon Gate MOS 2102A". Intel. Retrieved June 27, 2019.
209. ^ "One of the Most Successful 16K Dynamic RAMs: The 4116". National Museum of American History. Smithsonian Institution. Retrieved June 20, 2019.
210. ^ Component Data Catalog (PDF). Intel. 1978. pp. 3–94. Retrieved June 27, 2019.
211. ^ Jump up to:a b c d e f g h i j k l m n o p q r s t "Memory". STOL (Semiconductor Technology Online). Retrieved June 25, 2019.
212. ^ "The Cutting Edge of IC Technology: The First 294,912-Bit (288K) Dynamic RAM". National Museum of American History. Smithsonian Institution. Retrieved June 20, 2019.
213. ^ "Computer History for 1984". Computer Hope. Retrieved June 25, 2019.
214. ^ "Japanese Technical Abstracts". Japanese Technical Abstracts. University Microfilms. 2 (3–4): 161. 1987. The announcement of 1M DRAM in 1984 began the era of megabytes.
215. ^ "KM48SL2000-7 Datasheet". Samsung. August 1992. Retrieved June 19, 2019.
216. ^ "Electronic Design". Electronic Design. Hayden Publishing Company. 41 (15–21). 1993. The first commercial synchronous DRAM, the Samsung 16-Mbit KM48SL2000, employs a single-bank architecture that lets system designers easily transition from asynchronous to synchronous systems.
217. ^ Breaking the gigabit barrier, DRAMs at ISSCC portend major system-design impact. (dynamic random access memory; International Solid-State Circuits Conference; Hitachi Ltd. and NEC Corp. research and development) Highbeam Business, January 9, 1995
218. ^ Jump up to:a b "Japanese Company Profiles" (PDF). Smithsonian Institution. 1996. Retrieved June 27, 2019.
219. ^ Jump up to:a b "History: 1990s". SK Hynix. Retrieved July 6, 2019.
220. ^ "Samsung 50nm 2GB DDR3 chips are industry's smallest". SlashGear. September 29, 2008. Retrieved June 25, 2019.
221. ^ Shilov, Anton (July 19, 2017). "Samsung Increases Production Volumes of 8 GB HBM2 Chips Due to Growing Demand". AnandTech. Retrieved June 29, 2019.
222. ^ "Samsung Unleashes a Roomy DDR4 256GB RAM". Tom's Hardware. September 6, 2018. Retrieved June 21, 2019.
223. ^ "First 3D Nanotube and RRAM ICs Come Out of Foundry". IEEE Spectrum: Technology, Engineering, and Science News. July 19, 2019. Retrieved September 16, 2019. This wafer was made just last Friday… and it's the first monolithic 3D IC ever fabricated within a foundry
224. ^ "Three Dimensional Monolithic System-on-a-Chip". www.darpa.mil. Retrieved September 16, 2019.
225. ^ "DARPA 3DSoC Initiative Completes First Year, Update Provided at ERI Summit on Key Steps Achieved to Transfer Technology into SkyWater's 200mm U.S. Foundry". Skywater Technology Foundry(Press release). July 25, 2019. Retrieved September 16, 2019.
226. ^ "DD28F032SA Datasheet". Intel. Retrieved June 27, 2019.
227. ^ "TOSHIBA ANNOUNCES 0.13 MICRON 1Gb MONOLITHIC NAND FEATURING LARGE BLOCK SIZE FOR IMPROVED WRITE/ERASE SPEED PERFORMANCE". Toshiba. September 9, 2002. Retrieved March 11, 2006.
228. ^ "TOSHIBA AND SANDISK INTRODUCE A ONE GIGABIT NAND FLASH MEMORY CHIP, DOUBLING CAPACITY OF FUTURE FLASH PRODUCTS". Toshiba. November 12, 2001. Retrieved June 20, 2019.
229. ^ Jump up to:a b c d "Our Proud Heritage from 2000 to 2009". Samsung Semiconductor. Samsung. Retrieved June 25, 2019.
230. ^ "TOSHIBA ANNOUNCES 1 GIGABYTE COMPACTFLASH™CARD". Toshiba. September 9, 2002. Retrieved March 11, 2006.
231. ^ Jump up to:a b c d "History". Samsung Electronics. Samsung. Retrieved June 19, 2019.
232. ^ Jump up to:a b "TOSHIBA COMMERCIALIZES INDUSTRY'S HIGHEST CAPACITY EMBEDDED NAND FLASH MEMORY FOR MOBILE CONSUMER PRODUCTS". Toshiba. April 17, 2007. Retrieved November 23, 2010.
233. ^ Jump up to:a b "Toshiba Launches the Largest Density Embedded NAND Flash Memory Devices". Toshiba. August 7, 2008. Retrieved June 21,2019.
234. ^ "Toshiba Launches Industry's Largest Embedded NAND Flash Memory Modules". Toshiba. June 17, 2010. Retrieved June 21,2019.
235. ^ "Samsung e·MMC Product family" (PDF). Samsung Electronics. December 2011. Retrieved July 15, 2019.
236. ^ Shilov, Anton (December 5, 2017). "Samsung Starts Production of 512 GB UFS NAND Flash Memory: 64-Layer V-NAND, 860 MB/s Reads". AnandTech. Retrieved June 23, 2019.
237. ^ Tallis, Billy (October 17, 2018). "Samsung Shares SSD Roadmap for QLC NAND And 96-layer 3D NAND". AnandTech. Retrieved June 27, 2019.
238. ^ Han-Way Huang (December 5, 2008). Embedded System Design with C805. Cengage Learning. p. 22. ISBN 978-1-111-81079-5. Archived from the original on April 27, 2018.
239. ^ Marie-Aude Aufaure; Esteban Zimányi (January 17, 2013). Business Intelligence: Second European Summer School, eBISS 2012, Brussels, Belgium, July 15-21, 2012, Tutorial Lectures. Springer. p. 136. ISBN 978-3-642-36318-4. Archived from the original on April 27, 2018.
240. ^ Jump up to:a b c d "1965: Semiconductor Read-Only-Memory Chips Appear". Computer History Museum. Retrieved June 20, 2019.
241. ^ "1971: Reusable semiconductor ROM introduced". The Storage Engine. Computer History Museum. Retrieved June 19, 2019.
242. ^ Iizuka, H.; Masuoka, F.; Sato, Tai; Ishikawa, M. (1976). "Electrically alterable avalanche-injection-type MOS READ-ONLY memory with stacked-gate structure". IEEE Transactions on Electron Devices. 23(4): 379–387. Bibcode:1976ITED...23..379I. doi:10.1109/T-ED.1976.18415. ISSN 0018-9383.
243. ^ µCOM-43 SINGLE CHIP MICROCOMPUTER: USERS' MANUAL (PDF). NEC Microcomputers. January 1978. Retrieved June 27, 2019.
244. ^ "2716: 16K (2K x 8) UV ERASABLE PROM" (PDF). Intel. Retrieved June 27, 2019.
245. ^ "1982 CATALOG" (PDF). NEC Electronics. Retrieved June 20,2019.
246. ^ Component Data Catalog (PDF). Intel. 1978. pp. 1–3. Retrieved June 27, 2019.
247. ^ "27256 Datasheet" (PDF). Intel. Retrieved July 2, 2019.
248. ^ "History of Fujitsu's Semiconductor Business". Fujitsu. Retrieved July 2, 2019.
249. ^ "D27512-30 Datasheet" (PDF). Intel. Retrieved July 2, 2019.
250. ^ "A Computer Pioneer Rediscovered, 50 Years On". The New York Times. April 20, 1994. Archived from the original on November 4, 2016.
251. ^ "History of Computers and Computing, Birth of the modern computer, Relays computer, George Stibitz". history-computer.com. Retrieved August 22, 2019. Initially the 'Complex Number Computer' performed only complex multiplication and division, but later a simple modification enabled it to add and subtract as well. It used about 400-450 binary relays, 6-8 panels, and ten multiposition, multipole relays called "crossbars" for temporary storage of numbers.
252. ^ Jump up to:a b c d e "1953: Transistorized Computers Emerge". Computer History Museum. Retrieved June 19, 2019.
253. ^ Jump up to:a b "ETL Mark III Transistor-Based Computer". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
254. ^ Jump up to:a b "Brief History". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
255. ^ "1962: Aerospace systems are first the applications for ICs in computers | The Silicon Engine | Computer History Museum". www.computerhistory.org. Retrieved September 2, 2019.
256. ^ "PDP-8 (Straight 8) Computer Functional Restoration". www.pdp8.net. Retrieved August 22, 2019. backplanes contain 230 cards, approximately 10,148 diodes, 1409 transistors, 5615 resistors, and 1674 capacitors
257. ^ "【NEC】 NEAC-2201". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
258. ^ "【Hitachi and Japanese National Railways】 MARS-1". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
259. ^ The IBM 7070 Data Processing System. Avery et al. (page 167)
260. ^ "【Matsushita Electric Industrial】 MADIC-I transistor-based computer". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
261. ^ "【NEC】 NEAC-2203". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
262. ^ "【Toshiba】 TOSBAC-2100". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
263. ^ 7090 Data Processing System
264. ^ "【Mitsubishi Electric】 MELCOM 1101". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
265. ^ "【NEC】NEAC-L2". IPSJ Computer Museum. Information Processing Society of Japan. Retrieved June 19, 2019.
266. ^ Jan M. Rabaey, Digital Integrated Circuits, Fall 2001: Course Notes, Chapter 6: Designing Combinatorial Logic Gates in CMOS, retrieved October 27, 2012.
267. ^ Richard F. Tinder (January 2000). Engineering Digital Design. Academic Press. ISBN 978-0-12-691295-1.
268. ^ Asadi, P. (2007). "Energy-efficient 32 × 32-bit multiplier in tunable near-zero threshold CMOS" (PDF). World Applied Sciences Journal. 2 (4): 341–7. ISSN 1818-4952.
269. ^ Jump up to:a b c d Engineers, Institute of Electrical Electronics (2000). IEEE Standard 100: The Authoritative Dictionary of IEEE Standards Terms(7th ed.). doi:10.1109/IEEESTD.2000.322230. ISBN 978-0-7381-2601-2. IEEE Std 100-2000.
270. ^ Jump up to:a b c Smith, Kevin (August 11, 1983). "Image processor handles 256 pixels simultaneously". Electronics.
271. ^ Kanellos, Michael (February 9, 2005). "Cell chip: Hit or hype?". CNET News. Archived from the original on October 25, 2012.
272. ^ Kennedy, Patrick (June 2019). "Hands-on With a Graphcore C2 IPU PCIe Card at Dell Tech World". servethehome.com. Retrieved December 29, 2019.
273. ^ "Colossus – Graphcore". en.wikichip.org. Retrieved December 29, 2019.
274. ^ Jump up to:a b Schor, David (April 6, 2019). "TSMC Starts 5-Nanometer Risk Production". WikiChip Fuse. Retrieved April 7, 2019.
275. ^ "1960: Metal Oxide Semiconductor (MOS) Transistor Demonstrated". Computer History Museum. Retrieved July 17,2019.
276. ^ Lojek, Bo (2007). History of Semiconductor Engineering. Springer Science & Business Media. pp. 321–3. ISBN 9783540342588.
277. ^ "1964: First Commercial MOS IC Introduced". Computer History Museum. Retrieved July 17, 2019.
278. ^ Jump up to:a b Lojek, Bo (2007). History of Semiconductor Engineering. Springer Science & Business Media. p. 330. ISBN 9783540342588.
279. ^ Lambrechts, Wynand; Sinha, Saurabh; Abdallah, Jassem Ahmed; Prinsloo, Jaco (2018). Extending Moore's Law through Advanced Semiconductor Design and Processing Techniques. CRC Press. p. 59. ISBN 9781351248655.
280. ^ Belzer, Jack; Holzman, Albert G.; Kent, Allen (1978). Encyclopedia of Computer Science and Technology: Volume 10 – Linear and Matrix Algebra to Microorganisms: Computer-Assisted Identification. CRC Press. p. 402. ISBN 9780824722609.
281. ^ "Intel® Microprocessor Quick Reference Guide". Intel. Retrieved June 27, 2019.
282. ^ "1978: Double-well fast CMOS SRAM (Hitachi)" (PDF). Semiconductor History Museum of Japan. Retrieved July 5, 2019.
283. ^ "0.18-micron Technology". TSMC. Retrieved June 30, 2019.
284. ^ Jump up to:a b c d 65nm CMOS Process Technology
285. ^ Diefendorff, Keith (15 November 1999). "Hal Makes Sparcs Fly". Microprocessor Report, Volume 13, Number 5.
286. ^ Jump up to:a b Cutress, Ian. "Intel's 10nm Cannon Lake and Core i3-8121U Deep Dive Review". AnandTech. Retrieved June 19, 2019.
287. ^ "Samsung Shows Industry's First 2-Gigabit DDR2 SDRAM". Samsung Semiconductor. Samsung. September 20, 2004. Retrieved June 25, 2019.
288. ^ Williams, Martyn (July 12, 2004). "Fujitsu, Toshiba begin 65nm chip trial production". InfoWorld. Retrieved June 26, 2019.
289. ^ Elpida's presentation at Via Technology Forum 2005 and Elpida 2005 Annual Report
290. ^ Fujitsu Introduces World-class 65-Nanometer Process Technology for Advanced Server, Mobile Applications
291. ^ Jump up to:a b c d "Intel Now Packs 100 Million Transistors in Each Square Millimeter". IEEE Spectrum: Technology, Engineering, and Science News. Retrieved November 14, 2018.
292. ^ "40nm Technology". TSMC. Retrieved June 30, 2019.
293. ^ "Toshiba Makes Major Advances in NAND Flash Memory with 3-bit-per-cell 32nm generation and with 4-bit-per-cell 43nm technology". Toshiba. February 11, 2009. Retrieved June 21,2019.
294. ^ Jump up to:a b "History: 2010s". SK Hynix. Retrieved July 8, 2019.
295. ^ Shimpi, Anand Lal (June 8, 2012). "SandForce Demos 19nm Toshiba & 20nm IMFT NAND Flash". AnandTech. Retrieved June 19, 2019.
296. ^ Jump up to:a b Schor, David (April 16, 2019). "TSMC Announces 6-Nanometer Process". WikiChip Fuse. Retrieved May 31, 2019.
297. ^ "16/12nm Technology". TSMC. Retrieved June 30, 2019.
298. ^ Jump up to:a b c "VLSI 2018: Samsung's 8nm 8LPP, a 10nm extension". WikiChip Fuse. July 1, 2018. Retrieved May 31, 2019.
299. ^ "Samsung Mass Producing 128Gb 3-bit MLC NAND Flash". Tom's Hardware. April 11, 2013. Retrieved June 21, 2019.
300. ^ "10nm Technology". TSMC. Retrieved June 30, 2019.
301. ^ Jump up to:a b Jones, Scotten (May 3, 2019). "TSMC and Samsung 5nm Comparison". Semiwiki. Retrieved July 30, 2019.
302. ^ Jump up to:a b c Nenni, Daniel (January 2, 2019). "Samsung vs TSMC 7nm Update". Semiwiki. Retrieved July 6, 2019.
303. ^ "7nm Technology". TSMC. Retrieved June 30, 2019.
304. ^ Schor, David (June 15, 2018). "A Look at Intel's 10nm Std Cell as TechInsights Reports on the i3-8121U, finds Ruthenium". WikiChip Fuse. Retrieved May 31, 2019.
305. ^ Jones, Scotten, 7nm, 5nm and 3nm Logic, current and projected processes
306. ^ Shilov, Anton. "Samsung Completes Development of 5nm EUV Process Technology". AnandTech. Retrieved May 31, 2019.
307. ^ "TSMC Plans New Fab for 3nm". EE Times. December 12, 2016. Retrieved September 26, 2019.
308. ^ Armasu, Lucian (January 11, 2019), "Samsung Plans Mass Production of 3nm GAAFET Chips in 2021", www.tomshardware.com