After a long time, the generalist series among scientists has not written the last article until today, taking everyone to understand the open life of von Neumann.
Portals for the first two articles:
Let's talk about the "non-single" generalists among scientists: Newton
For example, the "non-specific" generalist scientists: Maxwell
Mathematical Foundations of Quantum Mechanics (1926), Computers and the Human Brain (1958), Operator Methods in Classical Mechanics, Game Theory and Economic Behavior (1944), And Continuous Geometry (1960), how would you feel if you told you that these books were written by one person? It is estimated that most people will feel that this person is too hard to fight for the title, and what kind of things dare to write ah...
The above joke, in fact, these works are indeed written by von Neumann alone, but he does not have to worry about his title, he is already one of the six distinguished professors of the Princeton Institute for Advanced Study in his early 30s, and he is also the youngest of them, you know, there is Einstein. So von Neumann is no ordinary genius, just look at the hair volume in the photo below to know that it is great!
His evaluation is more comprehensive and objective: one of the most important mathematicians of the 20th century, a scientific all-rounder in the fields of modern computers, game theory, and nuclear weapons, has been called "the father of computers" and "the father of game theory" by posterity. Well, let's classify him as a scientific generalist in this article, worthy of the name. Next, let's talk about his genius contributions in each of these aspects. The article is too long, or the old rule: divided into two parts, the first half.
His most well-known contribution is the fundamental contribution to modern computers, he put forward the two basic principles of computer manufacturing, that is, the use of binary logic and the working principle of the computer's stored programs, he also divided the computer into five parts (operator, controller, memory, input device, output device), which is known as the von Neumann architecture, until today's computer is still in use, so we generally call him the "father of modern computers". Let's talk about this history in detail.
As we generally know, the eniac machine is the world's first electronic computer. It began operations in Philadelphia on February 14, 1946. But the eniac has two obvious disadvantages: it has no memory, and it uses a wiring board for control, and just taking these wires is enough for a group of people to fiddle with all day, so its overall computing efficiency is certainly not high. Therefore, the eniac machine development team is obviously also very large, eager to improve the structure of the computer.
On the other hand, in 1944, Neumann was working on the development of the atomic bomb, which involved extremely difficult calculations (consider the darkness before dawn, the first atomic bomb was successfully tested the following year). In the study of the process of a nuclear reaction, a "yes" or "no" answer is given to the propagation of a reaction. Solving this problem usually requires billions of mathematical operations and logical instructions. His Los Alamos lab employs more than a hundred female calculators for this, and it's still too slow to count from morning to night with a desktop computer.
Von Neumann learned of the eniac computer by chance, and by chance, he threw himself into this magnificent undertaking and achieved the greatest achievement of his legendary life.
Von Neumann and Oppenheimer pose in front of the first computer
One day in the summer of 1944, Neumann, who was waiting at the train station, met Goldstein, the head of the U.S. Ballistic Laboratory's military, who was working on the Eniac computer, and had a brief conversation with him. During the conversation, Neumann quickly became interested in the development plan, and he realized the far-reaching significance of the work.
After that, von Neumann was introduced by lieutenant Goldstein, joined the eniac machine development team, and led a group of innovative young scientists and technicians to march towards a higher goal. After joining, von Neumann showed his strong mathematical knowledge and genius for exploring problems and comprehensive analysis, and on the basis of their joint discussion, Neumann wrote a report in handwriting, "The First Draft of the Edvac Report", which is an epoch-making document in the history of computer development.
In this report, von Neumann identified five parts of the new machine, including: arithmetic, controller, memory, input and output devices, and described the functions and interrelationships of these five parts, thus systematically introducing new ideas for the manufacture of electronic computers and programming.
Not only that, von Neumann also suggested the use of binary in electronic computers based on the characteristics of the dual stable operation of electronic components. The report mentions the advantages of binary and predicts that the adoption of binary will greatly simplify the logical lines of the machine. Time has proved the correctness of Neumann's predictions. Nowadays, the application of logic algebra has become an important means of designing electronic computers, and the main logic lines used in edvac have always been used, but the engineering methods for implementing logic lines and the analysis methods for logic circuits have been improved.
It was only for a few reasons that the report named only von Neumann, which later caused a lot of controversy, and the reasons for this were also controversial. What exactly happened is not the focus of this article and will be omitted here. But one thing is certain, the new computer architecture proposed in this report, known as the "von Neumann architecture", has been used to this day!
Extended Reading 1: Who invented the computer...
Turing and von Neumann are both Taishan Beidou characters, and they both have pioneering contributions to computers, but who are these two geniuses who are the real fathers of computers...
In a word, Turing theoretically expounded the possibilities of computers, he should be called the father of computer science; von Neumann defined the basic architecture and technical route of modern computers, he should be called the father of modern computers.
And according to an untestable rumor (at least I have not found a specific source), von Neumann, when talking about his theories and ideas, modestly said that the basis of these theories and ideas came from the ideas of the British mathematicians Turing and Boole. Although this statement is very modest, it is also true. In a paper published in 1936, "On computable numbers, with an application to the entscheidungsproblem," Turing rigorously described the logical structure of the computer, first proposed a general model of the computer, the "Turing machine", and theoretically proved the possibility of such an abstract computer.
Unfortunately, not many people attended Turing's speech at that time, and when the speech was over, only two people asked if they could reprint his article. In that era, Turing's ideas were considered too radical to be accepted by the vast majority of people. Today, the idea of a universal Turing machine — an abstract computer that multitasks by changing software — has long been identified as the precursor to contemporary computers, the common "ancestor" of the various computer products we use today from the first generation of cathode arrays to today.
We know that many of the top scientists of the 20th century have thought about the origin and evolution of life, and in this article, I have sorted out in more detail who have made significant contributions in the history of modern science in this article from equilibrium to away from equilibrium and how life has gone from zero to one. Von Neumann, who is a genius among geniuses, is certainly no exception, but his starting point is a little different.
After completing the design project for the edvac computer, his interest turned to the exploration of computing itself, and similar to Turing, he wanted to develop a general theory of information processing, and in the process, he was the first and most deeply aware of the connection between life and computation. Compared with the title of the father of computers, von Neumann's research and contribution to the theory of "computation" itself may not sound so dazzling, but I think its significance is more profound and far-reaching, and the following is introduced to you from two aspects:
a) Self-replicating automata
Inspired by the reproduction of life, von Neumann became interested in "how to achieve self-replication.". In the late 1940s, von Neumann had already initially solved the problem. He was giving a series of talks at a Hexon Seminar in California, and the central question to be solved was " How can machines replicate themselves ?" He believes that any system that can reproduce itself should have two basic functions at the same time:
First, it must be able to build a next generation where all constituent elements and structures are identical to its own;
Second, it needs to be able to pass on its description of itself to the next generation.
He called these two parts "universal constructors" and "descriptors", respectively, and the descriptors included a "general machine" and descriptive information stored on a medium that the general purpose machine could read. You see, his points of interest and thinking about the problem are related to computers and their specific implementations, that is, Von Neumann actually wants to make the self-propagation function into an automaton.
In this way, as long as the right raw material is available, the universal constructor can produce the next machine according to the instructions of the descriptor, and pass the described information to this new machine. Then, the new machine starts, then enters the next cycle, and then the children and grandchildren are infinitely poor.
You know, von Neumann made these thoughts before the discovery of genetic material! A few years later, in 1953, Watson and Crick found that dna was perfectly in line with two of von Neumann's demands. So from this point of view, life itself is an automaton. Not only that, but von Neumann also proposed a complete design of how to build a machine capable of self-replication! The tools used by von Neumann for the study of self-replicating machines are the following: cellular automata.
b) Cellular automata
The von Neumann architecture just mentioned is actually a centralized computing system, the core of the entire computer system is the CPU, and the program is stored in the memory for the CPU to execute sequentially. In contrast, the cellular automaton mentioned here is a decentralized computing system, which is generally composed of a large number of "cellular", and the state of each "cellular" is regulated by the state of the cells around it, and all the cellular follow the same rules to determine their own state.
But the seemingly simple rule, even if there are more cells that make up the system, should be lifeless. wrong! The overall behavior that cellular automata systems can exhibit is extremely complex and rich, even if they do not have any central control unit. If you'd like to learn more, take a look at the brief introduction in the Extended Reading section below:
Extended Reading 2: Cellular Automata
Light Bulb Array: An Implementation of a Cellular Automata, excerpted from the book "Complex"
The array of bulbs shown in the figure above can be used to understand the concept of a cellular automaton, each lamp has two states of "on" and "off", each light is connected to the surrounding 8 lights (the lamp on the side will be considered connected to the other side, such as the leftmost lamp, will be considered connected to the rightmost lamp. This way all the lights are connected to 8 lights). In the initial stage, some lights are turned on and part lights are turned off. The cellular automata performs calculations step by step like a CPU.
Each cellular automaton has a rule that explains how each lamp determines its next state based on the previous 8 lights around it and its own state (for example, a rule can be: adopt a neighbor-majority state). The figure shows how this rule changes next in this cellular auto chance.)
Similarly, in my fantasy of how life goes from zero to one from equilibrium to away from equilibrium (part II), I also mentioned a simplified two-dimensional cellular automaton: the content of the game of life, which everyone is interested in, and will not be expanded here.
Von Neumann proved that its cellular automata is equivalent to a general-purpose Turing machine, that is, a cellular automata also has the power of universal computing! It is another computer with a completely different structure from the von Neumann architecture! Interestingly, computer architecture is divided into "von Neumann" architecture and "non-von Neumann architecture", and the result is that the most important of the "non-von Neumann architecture" is still proposed by von Neumann... So the computer called him Father, not wronged.
Because the contributions to computing and computing are so dazzling and relevant, I will talk about them together in this article, and in the next one, I will talk to you about Von Neumann's contributions in other fields, which are equally radiant.