<s>
The	O
Harwell	B-Device
CADET	I-Device
was	O
the	O
first	O
fully	O
transistorised	B-Device
computer	I-Device
in	O
Europe	O
,	O
and	O
may	O
have	O
been	O
the	O
first	O
fully	O
transistorised	B-Device
computer	I-Device
in	O
the	O
world	O
.	O
</s>
<s>
The	O
electronics	O
division	O
of	O
the	O
Atomic	O
Energy	O
Research	O
Establishment	O
at	O
Harwell	O
,	O
UK	O
built	O
the	O
Harwell	B-Device
Dekatron	I-Device
Computer	I-Device
in	O
1951	O
,	O
which	O
was	O
an	O
automatic	O
calculator	O
where	O
the	O
decimal	O
arithmetic	O
and	O
memory	O
were	O
electronic	O
,	O
although	O
other	O
functions	O
were	O
performed	O
by	O
relays	O
.	O
</s>
<s>
E	O
.	O
H	O
.	O
Cooke-Yarborough	O
based	O
the	O
design	O
around	O
a	O
64-kilobyte	O
(	O
65,536	O
bytes	O
)	O
magnetic	B-General_Concept
drum	I-General_Concept
memory	I-General_Concept
store	O
with	O
multiple	O
moving	O
heads	O
that	O
had	O
been	O
designed	O
at	O
the	O
National	O
Physical	O
Laboratory	O
,	O
UK	O
.	O
</s>
<s>
By	O
1953	O
his	O
team	O
had	O
transistor	O
circuits	O
operating	O
to	O
read	O
and	O
write	O
on	O
a	O
smaller	O
magnetic	B-General_Concept
drum	I-General_Concept
from	O
the	O
Royal	O
Radar	O
Establishment	O
.	O
</s>
<s>
The	O
resulting	O
machine	O
was	O
called	O
CADET	B-Device
(	O
Transistor	O
Electronic	O
Digital	O
Automatic	O
Computer	O
–	O
backward	O
)	O
.	O
</s>
<s>
CADET	B-Device
used	O
324	O
point-contact	O
transistors	O
provided	O
by	O
the	O
UK	O
company	O
Standard	O
Telephones	O
and	O
Cables	O
,	O
which	O
were	O
the	O
only	O
ones	O
available	O
in	O
sufficient	O
quantity	O
when	O
the	O
project	O
started	O
;	O
76	O
junction	O
transistors	O
were	O
used	O
for	O
the	O
first	O
stage	O
amplifiers	O
for	O
data	O
read	O
from	O
the	O
drum	O
,	O
since	O
point-contact	O
transistors	O
were	O
too	O
noisy	O
.	O
</s>
<s>
CADET	B-Device
was	O
built	O
from	O
a	O
few	O
standardised	O
designs	O
of	O
circuit	O
boards	O
which	O
never	O
got	O
mounted	O
into	O
the	O
planned	O
desktop	O
unit	O
,	O
so	O
it	O
was	O
left	O
in	O
its	O
breadboard	O
form	O
.	O
</s>
<s>
From	O
August	O
1956	O
CADET	B-Device
was	O
offering	O
a	O
regular	O
computing	O
service	O
,	O
during	O
which	O
it	O
often	O
executed	O
continuous	O
computing	O
runs	O
of	O
80	O
hours	O
or	O
more	O
.	O
</s>
<s>
Cooke-Yarborough	O
described	O
CADET	B-Device
as	O
being	O
"	O
probably	O
the	O
second	O
fully	O
transistorised	B-Device
computer	I-Device
in	O
the	O
world	O
to	O
put	O
to	O
use	O
"	O
,	O
second	O
to	O
an	O
unnamed	O
IBM	O
machine	O
.	O
</s>
<s>
Both	O
the	O
Manchester	O
University	O
Transistor	B-Device
Computer	I-Device
and	O
the	O
Bell	O
Laboratories	O
TRADIC	B-Device
were	O
demonstrated	O
incorporating	O
transistors	O
before	O
CADET	B-Device
was	O
operational	O
,	O
although	O
both	O
required	O
some	O
thermionic	O
valves	O
to	O
supply	O
their	O
faster	O
clock	O
power	O
,	O
so	O
they	O
were	O
not	O
fully	O
transistorised	O
.	O
</s>
<s>
In	O
April	O
1955	O
IBM	O
announced	O
the	O
IBM	B-Device
608	I-Device
transistor	O
calculator	O
,	O
which	O
they	O
claim	O
was	O
"	O
the	O
first	O
all	O
solid-state	O
computing	O
machine	O
commercially	O
marketed	O
"	O
and	O
"	O
the	O
first	O
completely	O
transistorized	B-Device
computer	I-Device
available	O
for	O
commercial	O
installation	O
"	O
,	O
and	O
which	O
may	O
have	O
been	O
demonstrated	O
in	O
October	O
1954	O
,	O
before	O
the	O
CADET	B-Device
.	O
</s>
<s>
By	O
1956	O
,	O
Brian	O
Flowers	O
,	O
head	O
of	O
the	O
theoretical	O
physics	O
division	O
at	O
AERE	O
,	O
was	O
convinced	O
that	O
the	O
CADET	B-Device
provided	O
insufficient	O
computing	O
power	O
for	O
the	O
needs	O
of	O
his	O
numerical	B-General_Concept
analysts	I-General_Concept
and	O
ordered	O
a	O
Ferranti	B-Device
Mercury	I-Device
computer	O
.	O
</s>
<s>
In	O
1958	O
,	O
Mercury	B-Device
number	O
4	O
became	O
operational	O
at	O
AERE	O
to	O
accompany	O
the	O
CADET	B-Device
for	O
another	O
two	O
years	O
before	O
the	O
CADET	B-Device
was	O
retired	O
after	O
four	O
years	O
 '	O
operation	O
.	O
</s>
