The Apollo Guidance Computer (A.G.C.) was the onboard flight computer for many of the NASA Apollo space missions
from 1969 to 1972. It was designed by MIT Instrumentation Lab in 1962 and then built by Raytheon.
The 16-bit hardware of the A.G.C. consisted of approximately
4000 discrete integrated circuits (RTL 3-Input Dual NOR gates in a flat-pack package, made by Fairchild Semiconductor) connected
via wire wrap and encapsulated in epoxy resin to be hermetically sealed. The computer run from two-megahertz clock and had
2 kilobytes of RAM and 36 kilobytes of ROM, both built of magnetic cores.
Being more basic than the electronics in modern microwave
oven, the A.G.C. did not have to be too much powerful, but just reliable and appropriate to that task (guidance, navigation
and control system). It was the A.G.C. the computer that Neil Armstrong, Buzz Aldrin and Michael Collins used in the command
and lunar modules during the Apollo 11 mission, which landed on the moon on July 20, 1969: by entering simple commands by
typing in pairs of nouns and verbs to control the spacecraft, it took about 10500 keystrokes to complete a lunar mission!
The Apollo Guidance Computer is
known for being the first embedded computer made of single integrated circuits and its design has been a key driver for research
on integrated circuits. In fact all the digital logic functions which had been manufactured thereafter (from individual
gates to microprocessors) were synthesized from NOR and NAND gates.
|Resistor-Transistor Logic Technology - Fairchild Semiconductor
|Image credit: NASA
|Display and Keyboard (DSKY) interface of the Apollo Guidance Computer
74181 ALU Integrated Circuit
The ending term in "APOLLO181" comes from
SN74181, that is a 24 pin TTL digital integrated circuit made by Texas Instruments capable of arithmetic and logic operations,
which I chose to be the heart of my processor.
|APOLLO181 @ 2.5 MHz: 74LS181 ALU (Chip Data code: 1976 by Texas Instruments)
The 74181 chip is capable of adding, subtracting, left shifting, comparing
and performing all the possible sixteen logic functions (i.e. the number of different ways you can write down the different
choices of 0 and 1 for the four possible truth table rows: NOT, AND, OR, EXOR, NOR, Implication etc.); it also acts as a function
generator (generates specific values like zero, one, minus one, etc regardless of inputs). The 74181 performs in few nanoseconds
thirty two different types of arithmetic and logical operations on two 4-bit words: logic operations are performed on a bit
basis while arithmetic operations are performed on a word basis. Cascading it is as simple as going from Carry Out of one,
to Carry In on the next. It is worth remarking that not all 74181 functions have useful results, but APOLLO181 was anyway
designed to be able to perform all of them.
Consisting of the equivalent of 75 logic gates, that is a MSI Medium-scale
Integration chip, the 74181 (which is not a CPU as the 4-bit Intel 4004 was in 1971) doesn't need a signal trigger or a clock
to operate since has no internal state: it is just a combinatorial switching network with 14 inputs and 8 outputs where the
two 4-bit operands are operated upon according to 4-bit command and the outputs change a short time after the inputs have
No storage is provided, hence accumulation of temporary results always
requires an additional external register (4-bit latch called "Accumulator").
Launched by Texas Instruments in March 1970, the SN74181
integrated circuit is known for being the first complete ALU (Arithmetic Logic Unit) in a single package.
Introduction of the 74181: from conception to commercialization
"In March (1970), Texas Instruments
introduced the SN54/74181 arithmetic logic unit, claimed to be equivalent to seventy-five TTL gates ... it is the
closest thing yet to a 4-bit CPU in a package"
at: ACS Amateur Computer Society Newsletter @ Stephen B. Gray, Vol II, No. 6 May 1970]
"Fairchild Semiconductor has announced a high-speed, 4-bit arithmetic logic unit that can be expanded into
a low-cost 16-bit unit...The 9340 is designed to accept carry lookahead outputs from three other 9340 packages, producing
a 16-bit ALU without additional gates"
at: Computer World, pg. 42, 5 Ago 1970]
"To make the lowest-cost arithmetic logic unit with carry lookahead built-in,
Fairchild's new 9340 is the perfect arithmetic logic unit for almost every application."
[Found at: Computer Design Publishing Corporation, Volume 9, 1970]
I have not found any evidence that the Texas Instruments 74181
(or the equivalent 9341 Fairchild chip) was used before 1970
as some web sources report.
The first datasheet was published in the "TTL Catalog
Supplement from Texas Instruments" dated 15th March 1970 (which I own) and all schematics of minicomputers that use that chip
are from the Seventies.
In fact the first DATA GENERAL minicomputers, which
were the NOVA (1969) and the SUPERNOVA (1970), were certainly not based upon the 74181 chip: in the NOVA the arithmetic and
logical operations were performed 4-bit at a time on a Texas Instruments 7483, 16-pin full adder . The SUPERNOVA was based
upon four Signetics 8260 24-pin arithmetic logic element .
Only the next series, which was issued in late 1970,
the NOVA 1200, started to use the ALU 74181 in the datapath.
| Images Copyright by single Manufacturers
| The 74181 ALU chips were used in the TTL CPUs of many third-generation 1970s minicomputers
|Image copyright McGraw-Hill, 1973 
|As far as I know Data General Nova 1200 was the first commercial computer using the 74181 (1970)
To be sure about the date on which the chip first appeared, I did a research on the rare previous
data book edition dated 1st August 1969 "TTL Integrated Circuits Catalog from Texas Instruments". A prompt response from a
librarian at the Technische Universität Chemnitz Universitätsbibliothek (Chemnitz, Germany) was: "I only found SN54180/74180
function. It is the last entry of the numeric index and in the "Table VIII. Series 54/74 Generalized Loading Factors". I received
similar proof from the Northeastern University Libraries (Boston, Massachusetts): the August 1969 Catalog's index mentions
52 standard TTL devices including 30 MSI functions, with L- and H- variants, but not the 74181 function. Same kind response
from the German National Library of Science and Technology (Hannover,Germany): "The SN 74181 is not mentioned at the index
of the TTL Integrated Circuits Catalog from Texas Instruments". Our copy really is the "1969-1970 Catalog" with number "Catalog
CC201" and the date "1 August 1969" on the main title page".
This apparently was satisfactory, until I was told by the Computer History Museum (Mountain View,
California) that they own a TI Catalog edition, the "CC201-R", dated "1st August 1969" in which "the 74181 is listed in the
numeric index and the MSI index" but that "the references for the 74181 are all to a supplemental catalog - that's the 1970".
After further research it seems that this particular TI edition, really marked 1st August 1969,
is a reprint (or a revision) nearly identical to all the other August editions (the same layout, same list of datasheets and
all copies marked on the back "C-4252 110M 89"), but which contains a more exhaustive index that refers the customers also
to the next complementary Supplement Catalog, the CC-301, that was released in March 1970. You may recognise the earlier 1969
edition because the catalog number is CC-201, without the "-R" at the end.
This could explain in part the doubtful dating of the ALU 74181: the chip
was listed in some TI revised editions, dated August 1969, but it was documented by a datasheet for the first time only
in March 1970, in the CC-301 "TTL Catalog Supplement from Texas Instruments" on page S7-1 in the section:
"NEW TTL Integrated Circuit from TI".
|©Texas Instruments TTL ICs Catalog - 1 August 1969
|Few '69 editions list the 74181 but the chip was documented only in the supplemental '70 TTL Catalog
|©Fairchild by courtesy of Computer History Museum
|1969 Fairchild 9341M three layer metal mask. Was this an earliest design of the 74181 function?
Having been launched in
early 1970, the 74181 had to be anyway designed at least in the previous year, i.e. in 1969.
doubtful dating of the ALU 74181 might derive also from a caption of a microphotograph of the 9341M which was reported in
the 1973 release of "A Solid State Of Progress" by Fairchild Camera and Instrument Corp. that includes many of Fairchild's
most important technical milestones.
Fairchild 9300 Series was the TTL/MSI counterpart of the Texas Instruments 7400 Series; in particular the Fairchild 9341 chip
was equivalent of the TI 74181 integrated circuit.
caption line of the 9341M image in the book is year “1969” and the cutline says: "THE INTERCONNECTION
MASKS ON THIS DEVICE WERE GENERATED ON FAIRCHILD's COMPUTER-AIDED DESIGN SYSTEM AND IMPLEMENTED WITH THE FIRST THREE LAYER
METAL PROCESS. THIS IS A 48-GATE CUSTOM TTL LOGIC ARRAY".
images in that book were part of the records that Steve Allen collected during his career as a photographer for Fairchild
Semiconductor and National Semiconductor. The Steve Allen photographs of Fairchild Semiconductor was donated by Steve Allen
to the Computer History Museum,
Mountain View, California
in November of 2007. The image, here published in low resolution by courtesy of the Computer History Museum, is titled: “9341
3 layer MSI”, with Accession Number 102710042 and it is dated 1970-03, exactly the same year and month of the issue
of the 74181’s datasheet in the “TTL Catalog Supplement from Texas Instruments".
some research about the 9341 on the "Fairchild Semiconductor Integrated Circuit Data Catalog 1970" (©1969, Fairchild Semiconductor).
This data book is absolutely rare here in Europe. With the help of a gentle librarian of
the University Library of the University of Colorado
at Boulder that owns the data book, we found that the ALU
9341 is not mentioned in the index. The index shows only twelve MSI functions and the last entry is the function 9328.
9341 was later described, for the first time, in the data book "Fairchild TTL Family October 1970", in which Fairchild presented
its TTL family to the market.
Laws, Semiconductor Curator for the Computer History
Museum, who worked in Silicon Valley semiconductor
companies including Fairchild Semiconductor for more than 40 years, illuminated me about Steve Allen's photo and the origin
of the 9341 ALU.
1968 both Fairchild and Texas Instruments had bipolar variable array programs in operation to provide quickly custom
designed circuits. Fairchild's program was called Micromatrix, while TI's was called Master Slice. Fairchild had available
4600 (48 Gates/Array) and 4700 (96 Gates/Array) TTL Micromatrix arrays, which contained six and twelve cells of TTL logic
respectively; each cell consisted of four TTL AND-OR-INVERT elements, counted at two-gate-per-element complexity . The
cells were interconnected by two layers of custom metallization to produce the customer's desired function.
arithmetic logic unit was initially designed by Fairchild a year or two before the 9341 or 74181 as the MSI 2-layer 4711 (Steve
Allen's photo with Accession Number 102710038, 30 Sept 1968), an example of the capability of Fairchild's custom Micromatrix
product line. The 4711 chip was a functional 4-bit ALU, but not optimized for yield, as were the successor devices .
Fairchild 9341 MSI function was a derived function from the 9340, a
TTL MSI 4-bit arithmetic logic unit with internal Carry Lookahead, capable of two arithmetic operations and six logic functions,
which Fairchild designed in a 24 pin DIP package: it was conceived in the applications group under Bob Ulrickson by either
or both Clive Ghest and John Nichols. Robert Ulrickson was a supervisor of Systems Engineering Group in the Applications Department
at Fairchild (where he worked from 1966 to 1973) where his team of engineers invented the first dozen TTL MSI devices which
were initially introduced by Fairchild Semiconductor as the 9300 series.
9340 die was very big, that made it expensive. Also not every application needed the carry look ahead built-in feature. Realizing
that this would limit sales, they divided the function into two chips, the 9341 and 9342. To test out the idea rapidly the
first version of the 9341 was implemented using Fairchild's computer-aided design system: this is the 3-layer chip shown in
Steve Allen's photo.
the functionality was verified, the function was redesigned onto a smaller more economical chip suitable for high volume production:
the 9341 ALU chip.
did not market a TTL family until late Sixties. When Fairchild entered the market, during a period in which total industry
sales of integrated circuits almost doubled, a battle for market leadership in TTL was already on between Sylvania (the first commercial manufacturer of TTL) and Texas Instruments. At the same time,
National, which under license from Texas Instruments acted as a second source supplier of its 54/74 TTL family assuring a
secure supply, began an aggressive cutting price campaign that helped increase the market share of Texas Instruments design
1968, improvement in lithography significantly increased the number of transistors that could be integrated on a chip. Desirous
to gain share in the TTL business, Fairchild (9300 Series) and Signetics (8200 Series) pioneered the design of TTL/MSI functions
(Medium Scale Integration - up to 100 logic gates per chip) such as counters, shift registers and arithmetic logic units .
For a long time Fairchild supplied the more advanced MSI chips, including the 9300 4-bit universal shift register and the
9316 4-bit binary counter, while Texas Instruments had been more volume-focused on TTL/SSI chips (Small Scale Integration
- up to 10 logic gates per chip) such as simple gates and flip flops.
computer market was opening up, demand for TTL/MSI grew explosively: allowing a superior way to assemble a minicomputer, TTL/MSI
caught on quickly, particularly in the 74xxx numbering system originated by Texas Instruments, who soon grabbed the lead during
the early Seventies.
was actually the sole source on the proprietary TTL/MSI series 9300, while there were a lot of sources on the series 7400.
This represented a serious limit for Fairchild, since early computer makers, like Digital Equipment Corporation
(DEC), avoided to design with components available exclusively from a single source to reduce the risk of not being able to
sell their final products due to delivery problems with the sole supplier.
a result, Fairchild had to adopt the 54/74 numbering scheme, becoming just an alternate source to Texas Instruments
who was already a giant manufacturer .
of these products, the 7400 series from Texas Instruments became de facto standard, with many similar products being
produced: 74195 was the direct replacement of the Fairchild 9300 function, the 74161 of the 9316, the 74181 of the 9341, and
the 74182 of the 9342.
1970, after having introduced a much faster technology called "Schottky TTL", the design of Texas Instruments had become the
industry reference, and Sylvania had effectively withdrawn
from the semiconductor industry. The 74181 was successively implemented by Texas Instruments in Schottky S/TTL technology
in mid 1971: in that year Texas Instruments had 41% market share in TTL, and it would remain the market leader in bipolar
logic until after 1980.
per above historic reconstruction, the 9341/74181 was not the first conceived integrated ALU: a 4-bit ALU implemented on Fairchild's
computer-aided design system seems to have preceded it by some considerable time.
the Signetics 8260, TTL/MSI Arithmetic Logic Element with minimalist functionality (4-bit adder, XNOR, AND), was really the
first integrated ALU to be marketed, at least in 1969 as per “6947”
data code I recognised in a picture of a circuit over internet. The 8260, which was employed in DATA GENERAL SUPERNOVA,
is de facto the sole device that Texas Instruments cross-referenced to the 74181 (as a recommendation for new design) on its
first data book, in March 1970.
the time of their inventions, Fairchild did not patent the logic functions or circuit designs, their patent filings were much
more focused on semiconductor device process technology. Actually, there was no specific pattern of licensing in the early
semiconductor industry and patents and intellectual property rights were the subject of frequent never-ending costly litigation. Prior
to “The Semiconductor Chip Protection Act” of 1984, the application of copyright law to integrated circuits was
not clear and any form of intellectual property to adequately cover a chip did not exist . Also, the practice of "second
sourcing" to provide customers with the assurance of multiple suppliers for popular logic functions as an incentive to buy
their standard products, pushed manufacturers to cross-license their products. Additionally, the recurrent talent mobility
due to the large number of spin-off firms in the early semiconductor industry, gave companies an easier access to one another's
On the Web (mainly
at the Computer History Museum)
you can find interesting documents and interviews about the lack of patents filed by the early semiconductor companies in
Silicon Valley. Interesting are the conversations of Robert Wayne Ulrickson about the alleged
paternity of the MSI design of either Fairchild's 9300 product line or TI's early 74/54 series. In particular it is mentioned
the 9341/74181 ALU: you are invited to read the outcome of such discussion in this inteview .
|© Texas Instruments and Fairchild Semiconductor
|74181 was described in 1970 data books: on the left by Texas Instruments, on the right by Fairchild
|Click to enlarge
|TI is proud to announce to the market the 74S181 ( © Texas Instruments, Data Book II, 1971)
The 74181 in commercial
The 74181/74S181 chips were used in third-generation
minicomputers such as: Data General Nova 1200, a 16-bit machine issued in late 1970, which utilises an ALU datapath width
only four bits wide, passing through a single 74181 chip ; the Philips P850, a 16-bit machine in 1971 with 8-bit of datapath,
using a pair of 74181; the Xerox Alto I (1973) and Alto II (1975) both 16-bit machines with four 74181 restricted, so that
they could do only 16 arithmetic and logical functions ; the most of PDP-11 models, 16-bit minicomputers, sold by Digital
Equipment Corporation (DEC): the DEC PDP-11/10 (1972), 11/40 (1973), 11/04 (1975) used 74181 with 74182 carry lookahead; the
DEC PDP-11/34 (1976) used 74S181 with 74S182; the DEC PDP-11/45 (1972) and 11/60 (1976) used 74S181 with 74182 .
There were four 74LS181 with a 74182 in the CPU board
of the Texas Instruments Model 990/10, a 16-bit minicomputer (1975-76) which had a cycle time of about 250 nanoseconds. Model
990/10, at the time of its issue, was the most powerful member of the TI 990 family, being a TTL implementation of the 990/4
architecture which used the MOS N-channel TMS9900 (single chip 16-bit microprocessor) as its central processor .
Then we can find a massive use of the 74S181 in the
DEC VAX 11/780 (1977) which had a cycle time of 200 nanosecounds. VAX-11 extended the PDP-11 to provide a large, 32-bit, virtual
address for each user process. The 74S181 were located in the Exponent, in the Address and in the Arithmetic Section of the
data path, which, together with the Data Section, could operate as independent units in parallel with the others . It is
relevant that in the 1980s VAX-11/780 was the dominant performance reference in benchmarking computers, which was called (erroneously)
a 1-MIPS computer for many years.
|PDP 11/45 Data Paths Board with four 74S181 chips
|Image Copyright © 2001 by John Holden (The University of Sydney, Psychology Department.)
The longest propagation delay of the Schottky 74S181
(the typical time to perform the "A=B" output) is 20 ns, which means that the 74S181 can theoretically operate at 50 MHz.
Obviously in 1970s computers clock speed was reduced to match the speed limit of the rest of the circuit, in particular of
the slower memory. For example, the extensive use of Schottky TTL in the PDP-11/45 data paths made possible a 150-nanosecond
cycle time (i.e. a frequency of 6,6 MHz).
In minicomputers that used the 74181 ALU in the data
path, the move time for register-to-register transfers varies as follows:
PHILIPS P850: 12,8 µsec, 4 clock cycles (1971)
NOVA 1200: 1,35 µsec (1971) 
PDP-11/10: 3,1 µsec (1972)
PDP-11/45: 0,3 µsec (1972)
PDP-11/40: 0,9 µsec (1973) 
TI 990/10: 1,0 µsec + 2,9 µsec RAM Memory Expansion
Board (1975-76) 
PDP-11/04: 2,91 µsec (1975)
PDP-11/34: 1,83 µsec (1976)
PDP-11/60: 0,34 µsec (1976) 
In accordance with them, the transfer timing between
registers in our multi-chip didactic CPU is of the same order of magnitude (but only for 4-bit data transfer):
APOLLO181 @2,5 MHz: 3,2 µsec, two instructions 4+4
APOLLO181 @3 MHz: 2,7 µsec, two instructions 4+4 clock
Considering monolithic CPU, the "MOV reg, reg" transfer
timing improved more than proportionally with the year of the introduction:
Intel 8080 @2 MHz: 2,5 µsec, single instruction 5 clock
TI TMS9900 @3 MHz: 4,67 µsec, single instruction 14
clock cycles (1975)
Zilog Z80 @2,5 MHz: 1,6 µsec, single instruction 4
clock cycles (1976)
Intel 8085 @3 MHz: 1,33 µsec, single instruction 4
clock cycles (1976)
Intel 8088 @4,77 MHz: 0,42 µsec, single instruction
2 clock cycles (1979)
Intel 80286 @6 MHz: 0,33 µsec, single instruction 2
clock cycles (1982)
Intel 80386 @16 MHz: 0,125 µsec, single instruction
2 clock cycles (1985)
Intel 80486 @25 MHz: 0,04 µsec, single instruction
1 clock cycle (1989)
Given the above speed performances,
we understand why in the Seventies the use of the ALU 74181 in computer design has continued to be competitive against contemporary
microprocessors. But when the monolithic processor (characteristic of the fourth generation of computers) became
faster than the fastest processor built by cascading several ALUs, the 74181 was no longer commercially competitive (early
The classic PERQ workstation, launched in 1980, was
probably the last famous based around 74181 ALUs: the main CPU, built from five 74S181 and one 74S182 carry-lookahead generator
chip, had a 20-bit wide data path and a 0,17 microseconds long microcycle time.
|PERQ 20-bit CPU built from five 74S181 and one 74S182 carry-lookahead generator chip
The SN74S181 was obsoleted in December 1997
by TI. It is amazing that, at this writing, the commercial SN74LS181 is still flagged "active" in the Texas Instruments web
site. Today, in fact, as it was in the Seventies, this interesting ALU is still purchased for educational purposes by hobbyists
In March 1970, when Texas Instruments introduced
the SN54/74181, it was sold at $16.50 in quantities of 100-999, but the 1-24 price was about 50% more . After few years,
the TTL prices dropped: the Bugbook Vol. I reports a unit cost for unspecified MSI ALUs of $4.25 to $5.50 in middle 1974 .
In 1974 in Italy, according to advertisement price list in magazines , it was sold at 2500 lire (equivalent at that time
to $3.80). Advertisement price list in the first issue of the Byte Magazine (September 1975) show a cost of the 74181 at $2.98
and $3.55, both in California, USA. Today you can easily find the 74LS181 in the few italian shops selling electronic components
to hobbyists at nearly two euros ($2.60). Around internet you may easily still find it at few bucks. I recently discover a
SN74S181 (with datacode of 1975 and perfectly working) at 4.30 euro in an Italian electronic shop.
So the 74181 today is definitely not a rare
chip (maybe it will be in the future, in the year 2070 when, after a century, it will be moved from "vintage" to "antique"
|© Scuola Radio Elettra - Popular Electronics
|RADIORAMA March 1973: MAKE AN ARITHMETIC LOGIC TRAINER. The 74181 was forever popular among hobbyist
Evolution of the 74181 function
Following function 74181, there were less known but
interesting TTL functional variants:
74181 24 pin, 4 bit arithmetic logic unit and function
generator with 16 logic and 16 arithmetic type operations including left shift (the 74181 can't right shift), with comparator
74281 24 pin, 4 bit parallel binary accumulator with
8 arithmetic and 7 logic functions including B Minus A and A Minus B with full shifting capability
74381 20 pin, 4 bit arithmetic logic unit with 8 binary
functions selected specifically to simplify system implementation including B Minus A and A Minus B
74481 48 pin, 4 bit parallel binary micro/macroprogrammable
processor element with full function ALU, magnitude and overflow decision capabilities, double-length accumulator with full
shifting capability and sign-bit handling. The 74S481 was used in the Texas Instruments Model 990/12 CPU in 1979, which incorporated
floating point arithmetic, byte string operation, bit-array instructions and multiprecision integer and decimal conversion
74581 (never manufactured, see 74582 ALU)
74582 24 pin, 4 bit BCD Arithmetic Logic Unit with
four BCD functions: addition, subtraction, comparison and binary to BCD conversion
74681 20 pin, 4 bit parallel binary accumulator with
two synchronous registers: one simple storage register and a second storage/shift/accumulator register with 16 arithmetic
and 16 logic functions including B Minus A and A Minus B
74881 24 pin, same as 74181 plus a status check on
the input words in the logic mode
741181 24 pin, a faster version of 74181
All the above mentioned chips are easily expandable
since have outputs for look-ahead carry cascading: for this reason the 74181 is also called "a bit slice arithmetic logic
Then, it is worth mentioning the Monolithic
Memories Incorporated 5701/6701 slice processor family which was introduced in 1974. This was a complete bipolar LSI
4-bit slice processor element, equivalent to 1000 Schottky Gate complexity, which replaced 25 TTL MSI packages on a single
chip. It could perform multiple 4-bit nano-intructions such as Subtract, Shift and Store in one cycle (200 ns) and it
was used to upgrade systems using the 74181, 9340, 9341, 74S281 Arithmetic Logic Units. Pricing was 195.00 USD for 1-24
Very similar in design to the MMI 5701/6701 was the later 4-bit slice ALU AM2901 (1975), which
was the core of the famous AM2900 family of bipolar bit-slice processor elements created by Advanced Micro Devices
in the Seventies. The AM2901 and AM2903 were probably the main competitors of the more economic but significantly
less powerful SN74181 chip.
Here below my working model implementation of the 74181
chip, which I made on the excellent simulator ISIS Professional Proteus 7 DEMO by Labcenter Electronics Ltd.
|Click to enlarge
|My model implementation of the 74181 on ISIS Professional Proteus DEMO by Labcenter Electronics Ltd.
Many companies around the world manufactured
the TTL 74181 (commercial-grade):
The military-grade TTL unit was the SN54181.
The pin-to-pin compatible CMOS version was the CD40181 (but very slow: 1000 ns to perform A=B).
|The 74181 die measures 0.1 inches
| I have uncapped the TTL 74181 to take picture and measurement of the die
SN74181 chip die
I didn't find in any datasheet the schema of the 74181, in order to count
the exact number of transistors on the chip. But according to the TI schematic of the 4-bit Binary Full Adders SN7483, which
consists of 36 gates and 6 inverters and it is made of 92 transistors, we can estimate that the SN74181, which has a complexity
of 75 equivalent gates, should be built out of 180-200 transistors. Since the TI 74181 die measures 0.1 inches per side, this
my estimation fits with the density of about 20.000 components per square inch, typical of the early 1970s chips .
|74181 should be built out of 180-200 transistors
| Close-up picture I made of the TTL 74181 die taken after having uncapped the chip
| Photo taken with Canon Powershot A1100 IS
|Macro photo of the TTL 74181 die reveals details which can't be seen with the naked eye