IBM's history stretches back more than 100 years, during which time it has helped forge the modern information age.
The firm's museum at its UK base in Hursley, near Winchester, charts the storied timeline of the company, right back to its early electromechanical machines that read and stored data on punched cards.
Punched card tabulating machines were invented in the late 19th century as a way of collating census data on the fast-growing US population, which had passed 60 million at the time.
The machine was devised by Herman Hollerith, a man who would go on to found one of the companies that merged to form the Computing-Tabulating Recording Company in 1911, later renamed IBM.
The Hollerith tabulating machine was able to automatically transcribe and tabulate - arrange in tables - census data using a punched card system.
To use the system, a card would be punched with holes, with each hole corresponding to a particular characteristic the census office wanted to record, such as gender or age.
To read the census data off each punched card, the operator would place each card inside a press attached to Hollerith's machine. Inside each press were electrified pins that pushed against the card - if the pin passed through a hole in the card it completed a circuit and advanced a mechanical dial that recorded an instance of a particular characteristic. Early machines could process cards punched with 40 holes, allowing each press to count some 40 characteristics at a time. When the machine had finished processing the cards, an operator would read numbers off dials on the front of the machine and record the final tally of all the characteristics the machine had counted.
Throughout the early days of computing, from 1950s through the 1970s, punched cards were the primary way corporations and governments stored and accessed information.
This is the IBM 26 card punch, a programmable punch used to record information on punched cards.
The IBM 26 card punch keyboard
The punched card seen here is 80 columns wide, with each column representing a different number or character.
The number of columns on the cards is the reason that early green screen video displays were 80 columns wide.
Before vacuum tubes, or thermionic valves, became the norm for logic switching in computers IBM built equipment from electromechanical devices, such as the counter seen here.
A famous example of electromechanical computing was the IBM Automatic Sequence Controlled Calculator (Harvard Mark I) that came into use in 1944, during the final days of World War II. It was used by the US Navy during the war to run repetitive calculations for the production of mathematical tables.
The machine was enormous by modern standards, 51 feet long and eight feet high, and made up of a series of interlocking panels of small gears, counters, switches and control circuits, all only a few inches deep.
The ASCC was computing writ large: using 500 miles of wire with three million connections, 3,500 multi-pole relays with 35,000 contacts, 2,225 counters, 1,464 tenpole switches and tiers of 72 adding machines, each with 23 significant numbers.
Vacuum tubes are sealed glass tubes containing multiple electrodes that are able to represent binary data.
Computers built using vacuum tubes could operate far faster than those using mechanical or electromechanical components, as the tubes could switch between storing a binary 1 or a 0 just by turning an electrical circuit on or off, or more accurately by switching between a high and low voltage.
These are vacuum tubes, or electronic thermionic valves, from early IBM computers in the 1950s.
This is a Transformer Read Only Store memory developed for the SCAMP Project in 1960.
Read only control stores were developed as a way of reducing the number of expensive transistors needed for control logic in a computer.
Information on the setting of logic gates for the execution of instructions was permanently stored in the read-only memory.
The read-only control stores were later adopted as a design feature of System/360 processors.
An SMS Printer Buffer from 1965, a standard module system with ferrite cores.
These cards from the 1970s are packed with integrated circuits rather than discrete transistors.
For the information age to get started, computers had to become cheap and powerful enough to spread throughout society, and integrated circuits played a crucial role in this happening.
Even after the transistor replaced vacuum tubes as the engine of logic processing, computers still weighed in tonnes, filled entire rooms and cost more money than most people would earn in a lifetime.
The problem was that the more transistors and other electronic components that were packed onto a computer circuit board, the longer each board took to hand-build and the bigger the machine would become. Using transistor technology to build a computer anywhere near as powerful as a modern PC would have taken an army of workers years and the resulting machine would have struggled to fit onto a football pitch.
The integrated circuit solved this problem by slimming down bulky circuit boards to a chip no bigger than a fingernail. On an integrated circuit, the circuit board's complex mesh of hand-wired electronic components - transistors, capacitors, diodes and the like - are reduced to nothing more than imperceptible peaks and troughs on the surface of a tiny semiconductor chip. These smaller circuits could be packed together to create powerful computer processors that didn't need a box the size of a car to house them and that could be manufactured automatically.
The integrated circuit was invented by two men at roughly the same time during a period from 1958 to 1959 - Jack Kilby, an engineer with Texas Instruments, and Robert Noyce, also an engineer but working for Fairchild Semiconductor Corporation.
This is core storage, an early form of Random Access Memory (RAM), the form of memory storage found inside computers today.
This core storage card was found inside the System/360 Model 40 mainframe computer.
The System/360 family of machines was launched by IBM around the mid-1960s and was the first family of computers architected to be compatible with each other.
IBM Hursley was responsible for the development of the low-end System/360s, including the Model 40.
This hulking beast is part of series of hard drives codenamed Gulliver.
The first Gulliver-series drive shipped in 1974 with a capacity of 5MB, less than a single high quality MP3 track, while subsequent models cold hold 10MB and 14MB.
It used a rotary actuator with one 14-inch disc. It was the first hard drive to ship more than 100,000 units, with 177,000 sold in its lifetime.
Gulliver-series drives were used in the IBM 3600, 3790 and 3730.
Codenamed Dolphin this massive hard drive was announced in September 1969 as a feature of the IBM System/3 computer.
It drives a removable IBM 5440 Disk Pack, which contains a single platter with capacity of up to 2MB.
Dynamic RAM from 1966.
A floppy disk from 1971.
Solid logic technology from 1974.
A Large Scale Integration Circuit from 1974.