Coming out of the restaurant, Han Feng took a taxi directly and went back to the dormitory to get the computer first. ，

This meal was a bit inexplicable, Han Feng didn't expect Yang Xiwen to spend a lot of effort to let him meet the so-called sisters of the Central Academy of Fine Arts, but the final result turned out to be like this.

Yang Xiwen doesn't look like Lala?

Han Feng can still feel this, after all, he has been in contact with her many times, and he has a little understanding.

However, who can say this kind of thing, especially Han Feng, he is still not good at this kind of thing, and he doesn't need to worry about this nosy, just Lala, it has nothing to do with him anyway.

Back at the rental office, Han Feng began to work hard to perfect his newly conceived file system.

Now, all the data in his super system is based on binary, which is completely simulated according to the idea of the computer.

But Han Feng is very clear that this coding mechanism is very unreasonable now.

Computers don't have to use binary, and the reason why they use binary to encode data is determined by hardware and cost.

In the early days, computers were not binary, but had many other bases, such as decimal, octal, ternary, and so on.

The reason why binary was finally chosen is that on the one hand, binary is simple (only 0 and 1), which only needs two states to be represented, and simple high and low voltages can be represented, while other decimal systems and other base systems require more states than this. Therefore, the use of binary to represent numbers is much simpler than other decimal systems in hardware design, which can greatly reduce costs. Of course, binary has drawbacks. For example, the number 9, which has a binary representation of 1001, requires four digits to represent it, while if it is a decimal, it only needs one digit. If the number is larger, the binary that is obviously needed is longer...... In other words, with binary, the storage efficiency is sacrificed at the expense of computing speed.

For example, although the storage efficiency is so low, it is still possible to create a digital matrix that can represent dozens of bits on a very small chip, and a hard disk is also very small, but the data that can be stored can reach TB (TB=1024GB).

So. Hardware and cost determine the base system used by the computer.

However, now the hardware of Han Fengdi has changed. into his body, his brain, his neural network system.

In terms of cost, Han Feng can be described at any cost. Therefore, it is no longer necessary to use binary to represent data. With binary, the storage efficiency is needless to say, very low. It doesn't improve the speed of computing, but it may slow down, which doesn't do any good at all.

Some time ago, Han Feng carefully studied most of the research results of scientists in the field of computer science in artificial neural networks, and then based on his close observation of brain neuronal axon dendrites and other structures from the microscopic level, Han Feng made a major discovery and breakthrough, and proposed a new information storage coding algorithm, which can greatly improve the storage capacity of the brain disk.

This is an algorithm based on a real neural network, not a so-called "artificial" neural network that scientists themselves have created or envisioned.

This neural network system. It is composed of a large number of neurons, each with inputs and outputs, "dendrites" and "axons", which are interconnected to form a rather complex and powerful system. Dendrites receive signals, neuronal cells process signals, and axons output signals. Although the computing power of each neuron is very simple, and the signal transmission rate is quite low, the function enhancement of this system formed by the extremely parallel interconnection of a large number of neurons is not as simple as 1 plus 1 equals 2, but increases geometrically, which is why an ordinary person's brain can complete a task that requires at least billions of processing steps in a current computer in one second.

Often, scientists think that axons and dendrites only have the function of transmitting information, that is, transmitting various signals between neuronal cells.

But Han Feng has long discovered that axons and dendrites can not only transmit information, but also use to store information, they have many states, according to their strength, he roughly divided these states into two, strong and weak, and then encoded them with binary, successfully realized the storage of information.

This also shows the difference between the brain and the computer in terms of hardware.

Usually, the memory and combinator of a computer are independent of each other, for example, they are divided into hard disks and central processing units, that is, the storage of data and data computing have nothing to do with each other, and only by writing programs can they be able to communicate, exchange and process data.

But in the brain, the storage and computing areas are integrated, and they are both neurons (neurons include neuronal cells and the dendrites and axons connected to them).

The advantages of this special structure are very great, when the data is exchanged, there is almost no need to consume any energy in the transmission process, and the speed will not be limited due to the bottleneck of the hardware.

Although the neuronal cells used by Hanpho are basically in a "dormant" state, and it is still necessary to transmit static data to active nerve cells for data processing, due to the high efficiency of the neural network itself, the loss and transmission time in the transmission process are almost negligible.

Han Feng now has the confidence to upgrade the way supersystems are encoded, and most importantly, he has made another major discovery in the representation of the signal state of axons and dendrites.

Previously, he simply divided the strength of their signal into two types and expressed them in binary, but now he found that this signal can be further subdivided, as if it were broadcast FM, which can be adjusted to different channels to represent more information.

These days, he's been using computers to simulate just to figure out how many states dendrites and axons can represent.

However, after so many days of observation, research and simulation, he found that there seemed to be more than a hundred frequency bands of dendrites and axons!

******

It's hard for Han Feng to imagine how the brain processes such complex information encoding, decimal system?

Completely mimicking the way nerve cells in the brain store information has become an impossible task.

If you can't get 100% simulation, then keep it simple.

Simplicity is beauty!

As a result, Han Feng unified axons and dendrites into "coding cells", and roughly divided the strength of such cells into ten levels, that is, his super system will be based on the decimal system in the future.

In addition, according to the characteristics of neural networks, this new decimal based file system will be stored on a new three-dimensional structure, which is a big breakthrough for the previous two-dimensional structure, for example, in the past, it was to put things on the desktop, but now, it is to install things in vertical cabinets, and the storage capacity has been greatly improved.

The combination of the two new features, the new brain disk storage capacity, will be increased by at least 100,000 times more than the previous one, for the time being, there is basically no need to worry about the storage capacity will be insufficient.

The problem of the capacity of the brain disk was solved, and Han Feng began to design the file system according to the characteristics of the brain disk, because the current brain disk is a three-dimensional structure, and the new file system not only requires a plane index function, but also must be able to carry out vertical indexing.

Fortunately, this is just pure software design, Han Feng has a lot of experience in this, and has thought about it for a long time, and it took a while to perfect it, and the "building file system" format is freshly baked.

In terms of data encoding, Hanfeng did not redesign a set of encoding system, but simply converted according to the binary to decimal mapping, thus forming the current data encoding.

In order to avoid falling into the situation of paralysis again, Han Feng did not rush to overwrite the previous file system, but opened up another space in his brain and initially formatted it according to the format of the "Mansion File System".

Due to the radical innovation from the bottom, the initial formatting had to be experienced by Han Feng.

This file system is too complicated, and Han Feng spent almost all of his time on formatting for the two days of the weekend.

This is a meticulous job, and you have to set the initialization state according to the decimal bits, one by one, according to the format of your own design.

Han Feng didn't get much space in the early stage, only made a small block, and then he used the new coding rules to first write a system with very simple functions that can't even be called an operating system, and it has only one task, that is, formatting.

After this system was put into operation, the formatting work was completely taken care of by it, and Han Feng finally got out of the sea of suffering.