Chapter 110 Director Gao takes the greatest credit (3k)
The John Bull customs officer took the paper package with a smile on his face, waved his furry hand, and stamped it.
This batch of transistors finally arrived at a factory on Hong Kong Island. There was a sign hanging on the door of the factory, "Revitalizing Electronics".
Two nights later, these components were transported to the seaside and boarded a small boat under the cover of darkness.
A tall man on the boat shook hands with the customs declarer: "Comrade, thank you!"
In the dark night, the boat sailed towards the north.
——
After completing the theoretical training for maintenance personnel, Gao Zhendong switched to communication with the designers of the computing institute. As for the practical operation of maintenance personnel, , Gao Zhendong doesn’t need to worry about it. The people at the computing office get started very quickly. With them here, Gao Zhendong doesn’t need to continue to struggle with this work.
For designers, Gao Zhendong did not explain everything in the explanatory materials. They have a foundation and understand computers.
What Gao Zhendong mainly does is a few things.
The first thing to do is to put forward the von Neumann architecture in a systematic way.
The Von Neumann system and the Harvard system are the two most commonly used architectures in modern computers.
In fact, the earliest computers could only be used to complete specific tasks. Their programs were designed from the beginning to only do one thing or solve one problem. For example, early computers were a bit like today's functional calculators. They could be used to calculate various complex functions, but were they used for other purposes, such as word processing and playing games? Sorry, can't do it.
The inability to do this is not just as simple as not programming software for him, but that this thing is only prepared from the bottom up to do this kind of thing, because their program is completely solidified. This curing is curing on the physical level. Do we need to change the program? Then you have to change the circuit diagram directly.
Later, the concept of microprograms came into being. Microprograms can be regarded as solidifying some instruction groups inside the computer. These instruction groups can complete some simple functions, and then use these microprograms to combine and complete a larger task.
For example, the calculation of the function y=x+1 can be made into a microprogram, and then as long as the microprogram is called and x=1 is input, y=2 can be calculated.
At this time, the degree of solidification of the program has been reduced. What is solidified is the computer's instruction set and these microprograms. As for the main program, it can be changed to a certain extent.
Until 1978, domestic computers were actually still following this path. This was one of the reasons why domestic computers in that era ultimately had no general-purpose products.
When Gao Zhendong was searching for information on the Internet in his previous life, he came across a professional book on electronic computers published in 1978, and read it carefully with a bit of archeology. The content above was about this situation. Gao Zhendong almost didn't understand their examples. He always felt that there was a missing step in the middle. The reason was that the micro-programming step had long since disappeared by the time Gao Zhendong went to college.
At that time, the computer system parameters would directly include the languages supported by the computer system. This would be incredible to people in the future.
For example, DJS-6 supports ALGOL60 and FORTRAN.
TQ-16, supports ALGOL60, and is proudly marked as supporting the management program.
Waiting until DJS-130, I shyly added the operating system, but this operating system is not perfect. It only supports three languages, ALGOL60, FORTRAN, and BASIC.
With DJS-220, it seems to be a little more in line with modern computers: operating systems and algorithm languages, but by this time, it has completely lagged behind the world.
It can be seen that the microprogram method is still not flexible enough. In the end, the complete von Neumann architecture was perfected. However, this process sounds simple, but in fact it is more complex, intersecting and progressing. Some retreated.
The biggest feature of the von Neumann architecture is that programs and data are organized and called in the same physical memory, and the content of any part can be frequently modified, bringing the computer's flexibility to its peak. This is why contemporary general-purpose computers, The reason why this architecture is generally adopted is, of course, not without cost, that is, the efficiency will be lower.
Another common architecture is the Harvard architecture. The difference between this architecture and the von Neumann architecture is that its program memory and data memory are separated, and two independent data buses are used directly. Manage them separately, so the efficiency will be higher.
Most microcontrollers have a typical Harvard structure, including the ARM cores used extensively in smartphones that have become popular after the 21st century, which are Harvard structures.
Of course, with the development of technology, the distinction between the two is gradually getting smaller and will eventually disappear.
These are things for later. At this time in 1959, the difference between the two was still obvious.
Gao Zhendong doesn't care so much anymore. Anyway, my transistor computer adopts the von Neumann architecture, which also lays the foundation for the next step of computer generalization.
In Gao Zhendong’s view, sowing the seeds of generalization is even more significant than the transistor computer system itself.
Gao Zhendong first talked about the three principles of the von Neumann system: binary logic, program storage and execution, and the five major parts of the computer.
Then explain clearly the five major parts of a computer: operation, control, storage, input, and output.
Following this path, Gao Zhendong talked eloquently. At this time, the designers of the computing center felt that a new world had opened up. In the past, when working on computers, it was a matter of scratching the eyebrows and beards, without clear theoretical guidance.
The entire research and development work is typically to build a target first and then build a gun. All the functions and performance of this gun are prepared for this target. If the target is changed, for example, the distance is increased by 200 meters, it is over. This gun Just out of reach.
As for the set of things Gao Zhendong brought out, I only want to make my own gun. This gun can hit various targets within a certain range. The price is just that I may have to change the type of ammunition. of.
Director Li and Chief Engineer Chi of the Computing Institute were also listening carefully below. They were listening and communicating at the same time.
Director Li sighed with emotion: "At first, I thought that Director Gao's transistor computer only used transistor technology. It was small, lightweight and low-cost. But I didn't expect that the first lesson given to the designers was to give me a hard blow. After learning a lesson, the things behind it are actually more amazing and more meaningful.”
Chief Engineer Chi sighed softly: "Yes, once this lesson is over, should we continue along the M-3 path?"
Director Li shook his head: "I can't go any further. , are you still willing to go on that path? Look at the staff here, are they still willing to go on?"
Chief Engineer Chi said with a solemn expression. "I'm not willing to give in! As the saying goes, 'I could have endured the darkness if I had never seen the light.' Director Gao, you deserve the most credit!"
If Gao Zhendong could hear what the two of them said, I will definitely tell them that this is right. The algorithms and software technology in the north have never been surpassed from now to the future, but when it comes to computers themselves, they have never risen to prominence.
The two of them had an idea in mind and continued to listen to the class attentively. It was one thing to take the route, but in this class, it would be a pity not to listen.
Gao Zhendong finished talking about the architecture and continued to talk about the details of the design work of the transistor computer.
In terms of the structure of the CPU, people at the Institute of Computing understand it very well. Although there is no such thing as a CPU at this time, all arithmetic units, controllers, and registers are composed of discrete components. However, these large Parts, they are still very clear.
What Gao Zhendong mainly talks about is how to transform from transistors to a computer system step by step.
This thing is much more complicated and careful to talk about.
Gao Zhendong started from the basic logic gate circuit formed by transistors, step by step.
From how transistors obtain AND gates, OR gates, and NOT gates, we can even further form compound gate circuits such as NAND gates, NOR gates, AND OR gates, and XOR gates.
From the above gate circuits, how to form basic functional units such as bistable flip-flops, R-S flip-flops, and D flip-flops, as well as their logical functions and state transition tables.
Then how to use the above gate circuits and basic functional units to obtain half adder, full adder, decoder, register, counter and other functional components.
Finally, the above functional components are used to form arithmetic units, controllers, memory controllers, and even a complete computer system.
This is a relatively long process. Director Li and Chief Engineer Chi were surprised as they listened. This part can even be mentioned in isolation. Please Gao Zhendong to add more details about logical operations. Can form a course on its own.
If Gao Zhendong could know their thoughts, he would definitely say, ah, yes, it's okay. I passed the "Digital Circuits" exam and my grades were okay.
Not only "Digital Circuits", but also "Computer Principles and Interface Technology" in front of it, and there will be more in the future.
Director Li and Chief Engineer Chi quietly explained the matter to several staff members of the computing institute who were also attending the lecture. It was a surprise for Director Gao, they thought.
Director Gao has done too much, and we have done too little. It feels a bit embarrassing not to do something for him.
While the designers at the computing center are studying hard and making progress every day, the ESR project team of the third steel rolling mill is experiencing mixed feelings.
Mr. Chen took the test records submitted by Liang and found them and chewed his teeth while reading them.
"Hiss, the results of Zhendong's experiments under the given conditions are all good. Why do the other ones always feel less interesting?"
Liang Fafa nodded: "We are still far behind Director Gao. As long as we change the product type, we will always have some uncertainty about things like melting voltage, melting current, consumable electrode length, input power, slag volume, and depth of the slag pool. Well, there are always some fluctuations and surprises in product results."
Mr. Chen shook his head: "It's not your fault. This is a new thing. It's already good that you can do this. As for the slight difference, it’s not a big problem. Zhendong is temporarily borrowed by the computing institute. Now, it’s our turn to use it.”
Second update.
This chapter was actually very difficult to write. If I’m not careful, it will become a class handout. I tried my best, sadly.
(End of this chapter)