News Archive - Report on Hayakawa Electric (Sharp) Developing Electronic Calculator Using LSI ICs
IC-based Electronic Calculator Under Development at Hayakawa Electric (Sharp)
Electronics Magazine November, 11, 1968
An article stating that Hayakawa Electric (Sharp) has developed a
working experimental electronic calculator (shown in the image above the article)
that is smaller than anything yet produced, using eleven very advanced MOS Large Scale Integration
integrated circuits for its logic. Prior to this experimental machine,
electronic calculators would use up to and beyond 100 small-scale
bipolar or MOS integrated circuits spread across anywhere from
three to five circuit boards. The experimental calculator had one
circuit board for the eleven logic ICs. The machine provided the four
basic mathematical functions, and automatically positioned the decimal point
for maximum accuracy. The experimental calculator was developed with the help of a $42,000
subsidy from the Japanese government's Ministry of International Trade
and Industry (MITI).
The prototype used MOS LSI integrated circuits that were
developed by the Japanese semiconductor company Mitsubishi, which
was a well-established producer of small and medium-scale bipolar integrated
circuits that were used in some of Sharp's earlier electronic
calculators, such as the Sharp Compet 32 (Model CS-32A).
Mitsubishi was a relative newcomer in the Japanese MOS integrated
circuit arena, with Japanese competitors Nippon Electric (NEC) and Hitachi
already well-established making small and medium-scale MOS integrated
circuits, with both companies each investing tremendous amounts of R&D
money in the development of Large Scale MOS ICs. It was a bit of a shock
to both NEC and Hitachi when it was announced that Mitsubishi produced
the Large-Scale MOS ICs in this experimental calculator.
Each of the chips in the calculator carried out a specific
function as part of the overall architecture of the machine, with
each IC containing somewhere between 300 and 400 MOS transistors elements.
Sharp's intention with this experimental calculator was to prove
that the technology was attainable. The real goal, though, was to make
a production calculator based on the basic design used in the
experimenal calculator. Sharp's goal was to
use roughly half the number of ICs in what became the production version of
this experimental calculator. This meant that each chip would need to
contain at least double the number of elements, which was unheard of in Japan
at the time.
Mitsubishi had been having difficulty being able to reliably produce the eleven MOS
chips that made up the prototype calculator's logic in anything but prototype
quantities. Hayakawa Electric's engineers
were concerned that Mitsubishi's MOS IC processes simply were not up to the task
of cutting down the number of chips by at least half, and decided to look to
US chip manufacturers, which were quite far ahead of Japanese MOS chipmakers
at the time.
US chipmakers had proven that they could reliably manufacture MOS large-scale
integration ICs, and had been for some time. As early as 1965, US chipmakers were creating
MOS ICs with as many as 50 logic elements on a chip, and by late 1968 when
Hayakawa Electric showed this experimental calculator, US chipmakers
had dramatically improved the technology to pack ever-more logic
elements on a single chip. By the time that Hayakawa Electric began its search
in the US to find a chipmaker to design and manufacture the chips for their
calculator design, companies such as AMI (American Micro-Systems, Inc), General Instrument,
Philco-Ford Microelectrics and others had developed MOS LSI arrays with more than
enough density to meet Hayakawa Electric's needs.
Hayakawa Electric sent a small team of engineers and
management representatives to a number of US-based MOS integrated circuit
manufacturers. They met with no success, even though they had money to
offer up-front to get a contract for development of the ICs off the ground,
as MITI had provided an additional $130,000 to Hayakawa Electric to
make the experimental calculator a production reality. It isn't really
clear why the US chipmakers were reluctant to work with Hayakawa Electric,
however it is suspected that these chipmakers already had chip development
deals going with US calculator manufacturers, as well as secret contracts
with US goverment intelligence agencies as well as the various branchs of
the military, and that partnering with the largest electronic calculator producer
in Japan, a major competitor to these US calculator makers, might ruffle some feathers.
The story goes that the folks from Hayakawa Electric, after failing in
their task, were at the airport waiting for a flight home, not knowing what
the future would hold for them or their calcultor project. While at the airport,
a notification came over the PA system saying that there was a call waiting for Mr. Sasaki,
who was the project leader of the team. The call turned out to be
from a representative of Autonetics, a subsidiary of the giant firm of
North American Rockwell. Autonetics specialized in designing and
fabricating high complexity MOS integrated circuits, mostly for the
military and national security sectors of the US market. The team
had visited Autonetics and made their pitch for development of the chips,
but they were turned away, just as they had been at every other
US chipmaker they visited.
Apparently the word that the folks from Hayakawa Electric with money
to spend were turned away got up to higher levels of management within
North American Rockwell, and a decision was made that a deal would
be made with the Japanese, prompting the hurried call to Mr. Sasaki at the
airport.
Negotiations between Hayakawa Electric and Autonetics proceeded with speed and
little dispute, and the decision was made to go with Autonetics. While the decision
was made, there were some government hoops to go through. First, Japan's Ministry of
International Trade and Industry (MITI) would have to approve of the deal, as it was
going to be making grants to Hayakawa Electric to help in the development of a
commercially-viable version of the calculator, and the US Federal Trade Commission(FTC)
also had to give its stamp of approval for the deal. Both governmental approvals
came off with relatively few problems, but there was some contention by US
calculator manufacturers stating that allowing Hayakawa Electric access to advanced US
MOS integrated circuit technology could start a flood of other Japanese calculator
manufacturers rushing to get set up with US chipmakers, and be able to produce an electronic calculator
equivalent to a given US-made calculator for a less-expensive retail price. The
concern was that this could create a situation where US calculator manufacturers
could no longer effectively compete in the calculator marketplace.
While this concern very quickly came to pass, with the Japanese calculator manufacturers
taking over all but the most high-end (calculators with computer-like capabilities
such as those from Hewlett Packard and Wang Laboratories) calculator market.
The FTC and MITI had not all that long ago brokered a heated dispute over calculator
display technology (the Nixie tube), as well as complaints by Texas Instruments that Japanese
chip makers were infringing on TI's intellectual property rights by producing their own chips
that were being used in Japanese calculators that were being imported into the US without paying royalties.
These issues were quite contentious, and in the end required Japanese companies
to pay royalties for some devices that were clearly copies of US-held designs
In comparison, the Hayakawa/Autonetics deal was quite straightforward, with every aspect of the proposed calculator
except for the chip design and fabrication having been developed in Japan with careful
respect for US intellectual property rights, as well as all of the funding for the development
coming from Japan. While the US calculator manufacturers were right in their concern over the Hayakawa/Autonetics
deal, in reality, the deal resulted in a sizable amount of money coming into the US from Japan, money
that directly benefitted the US integrated circuit manufacturers. At the time, the US position on
integrated circuits was that it was of strategic importance that it keep its position as
the technological leader in integrated circuit development, and as such, any deals that would put
money in the chipmakers' pockets that could be used for further chip technology development were
good deals. Such concerns overrode the concerns of potential competition from Japan in the electronic
calculator marketplace.
Autonetics would go on to successfully develop the IC's
for the production calculator, which became the
Sharp QT-8D. It is
interesting to note that the article repeatedly states that the calculator
being developed was a pocket calculator, but the QT-8D was hardly
pocket-able. It also required mains power to operate, limiting its
portability. Despite these perhaps optimistic claims in the
article, Autonetics managed to pack all of the logic into four MOS LSI integrated circuits,
with a fifth bipolar clock driver chip used to generate the master clock
signals that were used to orchestrate the internal operations of the chips and
synchronizing the chips' communication with each other.
While the QT-8D was not technically the first MOS LSI electronic
calculator in the world (that title technically belongs to the
Victor 3900), the
QT-8D would go on to be historically recognized as claiming that title,
due to the fact that the Victor 3900 was too much of a stretch of the
technology at the time. The 3900 was plagued with constant delays for
delivery of the calculators due to problems manufacturing the chips
in production quantities, as well as reliability issues once in the hands
of customers. All told, the Victor 3900 ran the company that was contracted
to make the chips into financial ruin, resulting it it being sold off, and
the calculator discontinued not long after that.
Hayakawa Electric (which was renamed Sharp Corporation on January 1, 1970)
received a prestigious
IEEE Milestone Award
in December, 2005, recognizing the company's pioneering work between
1963 and 1974 in th
development of electronic calculator technology, with the Sharp QT-8D and its
MOS Large Scale Integration integrated circuits heralded as one of the
major innovations that Sharp created, among many others relating to electronic
calculator technology, that led to the IEEE granting the award.