Autosave: 2024-09-20 08:00:03

This commit is contained in:
thomasabishop 2024-09-20 08:00:03 +01:00
parent 0a78340f65
commit d21357e9c1
4 changed files with 52 additions and 8 deletions

Binary file not shown.

1
zk/1726814727_LUWV.md Normal file
View file

@ -0,0 +1 @@
# Turing Mach

View file

@ -17,6 +17,24 @@ created: Wednesday, September 18, 2024
- It worked as follows: - It worked as follows:
- Imagine we have an 8-bit number. The number is input as a sequence of pulses - Imagine we have an 8-bit number. The number is input as a sequence of pulses
where a pulse is 1 and the absence of a pulse is 0. where a pulse is 1 and the absence of a pulse is 0.
- The pulses are converted into sound waves (as with a speaker) and sent - The pulses are converted into sound waves (using a transducer, as with a
through a mercury-filled tube. speaker) and sent through a mercury-filled tube.
- The length of the tube is calculated... - The length of the tube is calculated so that it takes exactly the time of
one machine cycle for a pulse to travel from one end to the other
- At the other end of the tube, another transducer converts the sound waves
back to electrical pulses
- This operates in a loop: the pulses go back into the input end of the tube
and the cycle runs again.
To read the data, the computer would "listen" to the pulses coming out of the
receiving end. To write new data, the computer would inject new pulses at the
precise moment when it wants to change a '0' to a '1' or vice versa.
For an 8-bit number there would be 8 pulses for each bit and it would be
necessary to keep track of 8 specific positions in the delay line where each bit
would be stored.
The limitations were timing - difficult to keep track of each precise bit and
when it would arrive at the output. Also it was sequential rather than random
access. Not every bit could be accessed at once, you had to wait for its time
slot to come around to access it.

View file

@ -370,8 +370,8 @@ the ABC to make ENIAC.
### Post-EDVAC devices ### Post-EDVAC devices
There were several notable devices which attempted to implement the architecture There were several notable devices which attempted to implement the architecture
described in the _First Draft_ using vacuum tubes and other fully-electronic described in the _First Draft_ using vacuum tubes for logic operations and a
methods for memory: variety of different electronic methods for memory:
- the Manchester "Baby" - the Manchester "Baby"
- the Cambridge EDSAC - the Cambridge EDSAC
@ -379,6 +379,8 @@ methods for memory:
#### Manchester Baby (1948) #### Manchester Baby (1948)
![The Manchester Baby Computer](../img/manchester_baby.png)
- An experimental computer intended to create the - An experimental computer intended to create the
[von Neumann architecture](CPU_architecture.md) using [von Neumann architecture](CPU_architecture.md) using
[Williams_Tube_RAM](Williams_Tube_RAM.md) [Williams_Tube_RAM](Williams_Tube_RAM.md)
@ -390,15 +392,17 @@ methods for memory:
#### EDSAC (1949) J.Wilkes et al. #### EDSAC (1949) J.Wilkes et al.
![EDSAC](../img/EDSAC.jpg)
- _Electronic Delay Storge Automatic Computer_ - _Electronic Delay Storge Automatic Computer_
- Constructed by Maurice Wilkes and others at the Mathematical Laboratory - Constructed by Maurice Wilkes and others at the Mathematical Laboratory of the
Cambridge University. University of Cambridge.
- The second digital stored-program computer after the Manchester Baby. - The second digital stored-program computer after the Manchester Baby.
- It used vacuum-tubes for the arithmetical operations in the ALU and mercury - It used vacuum-tubes for the arithmetical operations in the ALU and mercury
delay lines for the RAM. [delay line memory](Delay_line_memory.md) for the RAM.
- Designed to be used by relatively non-specialist practitioners from other - Designed to be used by relatively non-specialist practitioners from other
university departments who were expected to program it themselves. To this university departments who were expected to program it themselves. To this
@ -410,6 +414,27 @@ methods for memory:
operations) operations)
- diagnostics: techniques for verifying program code and its correctness - diagnostics: techniques for verifying program code and its correctness
![Programming Digital Electronic Computer book](../img/programming_a_digital_electronic_computter.jpg)
- The key players wrote the first textbook on programming in 1951: _The - The key players wrote the first textbook on programming in 1951: _The
Preparation of Programs for an Electronic Digital Computer_ (Wilkes, Wheeler, Preparation of Programs for an Electronic Digital Computer_ (Wilkes, Wheeler,
and Gill) and Gill)
#### The Automatic Computing Engine (ACE) (1946) Alan Turing
- Technical design provided by Alan #Turing working at the Mathematics Division
of the National Physical Laboratory. The product of Turing's theoretical work
in "On Computable Numbers" where he proposes the concept of a
[Turing_machine](Turing_machines.md) and based on his experience with early
single-purpose computing devices at Bletchley.
- Due to secrecy over wartime work, it was hard for get Turing to build a fully
electronic implementation since this would disclose advances made during the
War. Tommy Flowers was intended to be brought in to build it but again this
was scuppered over the classification of wartime technology. Had this not been
the case, it is likely it would've been seen as on par with the EDVAC in its
design, if not dwarfing it.
- Turing proposed a stored program architecture with high-speed memory. It would
be more perfomant than the EDVAC as a result. It used punched cards for the
input of data and programs