1. How long can the human brain run stably?
The human brain only has three working states of wakefulness, non-REG sleep and REM sleep , and there is no "shutdown" state, so the human brain does not have the restart process from the shutdown state to the working state. In other words, the human brain cannot be restarted .
At present, the average human life expectancy is more than 70 years, and some people can live to more than 100 years, which means that the human brain can run stably for 100 years without restarting .
However, the current personal computer has a high probability of malfunctioning after several months of continuous operation, and a server with excellent performance needs to be restarted actively after 3 to 5 years of operation .
Why can the human brain run stably and continuously for 100 years, while a computer with excellent performance can only run stably for 3 to 5 years? The ratio of the trouble-free running time between the two is 20:1 .
In fact, if human beings can live for 200 years, the human brain can also run stably for 200 years. In other words, as long as the human brain can survive, it can run stably for as long as possible ! What makes the human brain can run continuously and stably for such a long time?
2. The mystery of stable operation
The picture below shows an ordinary faucet and a control center of a hydropower station , both of which perform the "same" function: turning on and off the water.
Just ask: Which of these two devices has a higher probability of making mistakes during operation?
Obviously, ordinary faucets have a lower probability of failure, and may not fail in a lifetime . In fact, the two functions of turning on and off the water can always be completed normally within the working life of the faucet.
Why do ordinary faucets have a low probability of error and a longer stable operation time?
The reason is because the function is single and the operation is simple!
The simpler the better! The simpler the less error-prone! The simpler and more stable!
The law of entropy increase states that the greater the entropy, the more stable the system . The more unitary the system, the more uniform the energy distribution, and the more uniform the energy distribution, the greater the entropy, and the greater the entropy, the more stable the system. The first single responsibility principle, Single Responsibility Principle (SRP), among
the five principles of software design (SOLID), the single principle is the first among the five principles, illustrating its importance! Software design must abide by the single principle, realize single function, and simple logic. Only in this way can the software quality be guaranteed and the software can run more stably. The important reason why the human brain can operate continuously and stably is the use of a single principle .
3. How to implement the single principle
The important reason why the human brain can run stably as long as the brain can survive is that a single principle is used in all aspects . The specific implementation methods of the human brain nervous system are: split, partition, layer, and group .
Let's take a look at how the human brain system achieves this " four points "?
3.1 Split
The human brain system has multiple components, mainly including: cerebrum, cerebellum, diencephalon and brainstem.
1. Brain: The brain is the most advanced part of the central nervous system, which completes functions related to advanced emotions, spirit, vision, language, and body movements; 2. Diencephalon: completes functions related to sleep, food intake, language, and sensory density
. Function;
3. Cerebellum: mainly realizes functions related to balance and fine motor control;
4. Brain stem: mainly related to consciousness state, life center, etc.;
We divide the human brain into multiple components, and these components are called the divisions of the human brain . Each sub-body performs a specific function, and each sub-body performs its own duties. The functions performed by the sub-body and the sub-body are different, and the physical structure of the sub-body is also different.
Since each split body is only responsible for a specific type of function, these functions are relatively simple compared to the entire function of the human brain nervous system , thus ensuring the stable operation of the split body !
3.2 Partition
The human brain is composed of multiple divisions, each of which is divided into multiple functional partitions , each of which can independently complete a specific function. The brain is an important division of the human brain, and its divisions are as follows:
The 17th, 18th, and 19th areas of the cerebral cortex are visual divisions, which are specially used to process human vision.
Areas 1, 2, and 3 of the cerebral cortex are somatosensory divisions, which are specifically responsible for somatosensory areas.
Areas 41 and 42 of the cerebral cortex are auditory divisions, which are specialized in complex human hearing.
The 43rd area of the cerebral cortex is the taste partition, which is responsible for the human body's taste.
The partition strategy allows each functional partition to complete a specific function, and the partition only needs to focus on the functions it is responsible for , thus decomposing a complex function into multiple simple functions.
3.3 Layering
When the functions processed by some functional partitions are more complex, the partitions will split the complex functions according to the logical relevance of processing. We call this splitting hierarchical . Each layer completes a specific function, and multiple layers complete the work of the entire partition.
Taking the visual system in the nervous system as an example, the primary visual cortex is the part of the brain that processes vision, or the striate cortex. The arrangement of neurons in the striate cortex can be roughly divided into six layers. These cell layers are named I, II, III, IV, V, and VI. The cellular structure of the striated cortex is shown below:
These cell layers of the striated cortex are composed of axons and dendrites. Information received by the retina is transmitted to the LGN (lateral geniculate nucleus) and finally to the striatal cortex.
Sensing stimuli, different types of ganglion cells in the retina transmit the information to different layers of the striate cortex for processing.
The vision we see actually includes the perception of different characteristics of objects, such as color, shape, movement, etc. These characteristics are processed in parallel by different cells in the striate cortex, and different layers in the striate cortex complete a process. special information processing .
This layering strategy can subdivide complex functions into some relatively simple functions , and each layer only needs to focus on some simple functions, thus ensuring the stability of operation.
3.4 Grouping
Some functions in the human brain system are composed of a series of identical sub-functions. At this time, the human brain nervous system will group this series of sub-functions, and each group only needs to focus on a specific sub-function .
For example, in the human auditory system, there are Coty's organs on the cochlea, and these Coty's organs contain auditory receptors.
Sound signals of different frequencies are received by sensory cells in different parts of the cochlea.
Sound signals of different frequencies are eventually passed on to different cells in the auditory cortex for processing.
Grouping can subdivide the same function into a series of sub-functions with similar functions . These sub-functions are relatively simple, and each group only focuses on one simple sub-function, thus ensuring the stability of operation.
3.5 Dedicated access
I talked about the division, partition, layering, and grouping used by the human brain in order to realize the single principle. In addition, in order to realize the single principle, most of the information pathways in the human brain use dedicated information pathways , which means that each information pathway is only responsible for one kind of information.
auditory pathway
motor pathway
taste pathway
Summary:
The human brain system "infinitely subdivides" the problem through operations such as splitting, partitioning, layering, and grouping, so that the underlying cells only need to complete a simple task, ensuring system stability.
This kind of operation strategy is like managing a country. It is very difficult to manage a huge country. Therefore, we divide the country into different levels and divide the huge country into smaller administrative regions, so as to achieve scientific and effective management.
Taking the great China as an example , China divides the whole country into several provinces, each province is divided into multiple cities, each city is divided into multiple counties, each county is divided into multiple towns, each town is divided into multiple a village.
4. How software engineers can learn from the human brain operation strategy
4.1 Split and Partition
Split and partition are essentially the same , and they both use the " divide and conquer " strategy, which is to split the entire software into multiple tasks or modules according to their functions. Each task completes a specific function, and each task Only focus on the functions that you are responsible for , and not on the functions of other tasks.
A task needs to include all relevant codes to complete a specific function inside it, so that the task can run completely independently without relying on other tasks, and achieve high cohesion.
The figure below is the engineering file of a gateway project. It can be seen from the figure that the whole project adopts the "divide and conquer" strategy. We divide the whole project into multiple tasks, each task completes an independent function, and each task does not Dependent on other tasks.
4.2 Layering
A complex software function can be partitioned horizontally into reasonably many subsystem layers . Related parts of the system are grouped together in the same independent layer.
The realization of the function of the upper layer needs to call the function of the lower layer and get a reply, and any two adjacent layers conform to the relationship between the client and the server. The lower layer provides services for the upper layer and returns results for the upper layer calls, that is, the functions of the lower layer serve the upper layer .
Layered structure rules:
1. The nth layer only depends on the n-1th layer below
. 2. The nth layer does not depend on the n+1th layer.
3. The nth layer only provides services for the n+1 layer.
4. The nth layer Layering of services provided by layer n-1 through interfaces
Continuing to take the engineering of the gateway project as an example, it can be seen from the figure that each task adopts a layered strategy. Usually, we divide the tasks into three layers . The first layer is responsible for completing transactions related to processor registers, the second layer is responsible for completing driver control-related transactions, and the third layer is responsible for completing and business-related transactions.
The layered architecture is very suitable for maintenance and transplantation . If the task needs to be transplanted to other processor platforms, only the first layer needs to be modified; if the processor platform has not changed but only the peripherals have been modified, we only need to modify the second layer; If the processor and peripherals have not changed and the business logic needs to be modified, we only need to change the third layer.
4.3 Grouping
Some operations contain a series of similar operations, we can group these similar operations , and select the corresponding grouping according to actual needs.
For example, we can group the addition of different data types: this operation is called overloading
in java .
5. References
1. "Neuroscience-Exploring the Brain" - Mark F. Bear, Barry W. Connors, Michael A. Paradiso
2. "In-depth Understanding of Computer Systems" - Randal E.Bryant / David O'Hallaron
3. "Computer Composition Design and Design" - David A. Patterson (David A. Patterson) / John L. Hennessy (John L. Hennessy)
4. "Software Architecture and Patterns" - Joachim Goll
