1976: magnetic-core memory
Historical background
The first who searched for an alternative non-volatile memory was An Wang, who was born in Shanghai. His inventions were the 'pulse controlling device' and the 'write-after-read-cycle'. The first invention implies that the cores, which are formed like a ring, can store data with the aid of well-directed control of magnatic-discharges. His second invention solved the problem, that the informatoin is destroyed after its readout. Wang had his invention the 'write-after- read-cylce' patented. This patent became valid in 1955 and IBM (International Business Machines Corporation) had to buy it from Wang, because they already used it a few years before.
The magnetic-core memory reached its final form by further inventions. The most important was the 'coincident-current system' by Jay Forresters, which allowed to put many magnetic- cores in a memory with just a little number of wires (arranged in a matrix).
By the end of the 1950s, the magnetic-core memory got menufractured by hand with microscopes in asian factories. The manufracturing process was never automated. The cots of this process, which were high in the beginning, decreased with the market development. The access time also decreased over the years whereas the memory capacity rised. The magnetic- core memories were applied in the usual computing machines. A well-known producer was for example IBM. By the mid-1970s, magnetic-core memories were displaced by semiconductor memories.