Riding the wind of rebirth

"But if CompuBox is such a high-sounding name, then it would be a bit of a shame to limit it to a character terminal." Zhou Zhi continued: "Can this terminal be equipped with the function of interpreting the graphical interface? It should be possible to achieve this using the JAVA protocol."

"The storage space used by the terminal is used to pre-install MINICOS and JAVA, and the remaining space is used to install fonts and drivers." Mai Xiaomiao said.

"There are also a series of interfaces. We try to use TYPE-C as much as possible." Zhou Zhi's thinking became more and more divergent: "To match the graphical interface, we also need a mouse."

"If that's the case, what would this CompuBox look like?"

"My ideal is a 16-mo book, about six or seven centimeters thick, which can serve as the base of a monitor, use a TYPE-C interface as a power input, and also have many interfaces for monitors, printers, keyboards, mice and other peripherals. The front row is an indicator light. The system uses graphical Minicos, which can be seamlessly linked with COS on a remote server. The application system is developed in JAVA and works through frontpage access. This will maximize the availability of the machine and minimize the difficulty of migrating our original system."

"In a nutshell, it will be far superior to all current character terminals in terms of usage and will have graphics processing capabilities. Except for the server being deployed in a remote center, the user experience of this terminal is actually not much different from that of a computer."

"The key is price, because the computing power card has been developed, this is just an expansion of application, plus we are already producing memory and hard storage, so the equipment cost will not be too high, and the profit margin will be huge."

"With these functions, it is possible to build low-cost computer classrooms and information rooms for primary and secondary schools across the country, which is very important for improving the level of the country's information technology reserve force."

"Moreover, when these children grow up, they will become familiar with COS and MiniCOS users, which will also be very beneficial to our group."

"Are you thinking too far ahead?" Mai Xiaomiao found it a little funny, but the pie that Zhou Zhi had drawn did sound very attractive.

The thing designed by Zhou Zhi actually had no prototype in the previous life. If you have to say there was one, it was somewhat similar to a tablet computer, but the storage was stored on a server through the network, the screen was an ordinary monitor, and the peripherals were richer and more complex, with more interfaces, and the processor was not as fast as that of later tablet computers.

More similar things are happening in operating systems and JAVA-based application development.

But there is no doubt that this will be a very eye-catching step for the Clover operating system, and an intermediate version similar to the personal computer operating system and the later Android will appear, which is between the server-based COS and the industrial control-based MiniCOS.

But these are all things that can be achieved within Clover's existing patented architecture.

"Are terminals based on the VT100 protocol still being developed?" Mai Xiaomiao was a little regretful. "The technical agreement expired this year, and the intellectual property protection just happened to expire. We can get it for free, and it would be a pity not to use it."

"We can make this announcement to the outside world, just say that Clover Group is going to develop domestic terminals." Zhou Zhi rubbed Mai Xiaomiao's hands and smiled wickedly: "I guess there are many manufacturers at home and abroad who have the same idea as you. It can be used as a smokescreen for us."

In the next few days, Zhou Zhi began to contact Shanghai and Shudu, mainly to tell Li Guangnan about his ideas about smart terminals. Technically speaking, Zhou Zhi and Mai Xiaomiao's design plan would not be a technical obstacle for Clover Group, but Li Guangnan was not very optimistic about Zhou Zhi's idea.

From the perspective of cost reduction, developing a character terminal using the VT100 protocol, which is no longer under patent protection, is the simplest and lowest-cost solution. It can also meet most application scenarios. The only difference is that the interface is a little uglier. It only uses the up and down keys and the enter key to control the movement of the cursor on the screen composed of characters, which can meet basic needs.

However, Zhou Zhi believed that this alone was not enough, because the character-type VT100 terminal could not realize many document editing functions, without the help of a mouse, there was no free switching of multiple screens, no split-screen function, and most importantly, no image processing function, which made it impossible to complete many work scenarios.

However, with the addition of so many functions, the function of this terminal is actually almost the same as that of a computer. It can even be said that all you need to do is add a computing power card to replace the motherboard and CPU, add a hard drive with larger space, and then replace the processor, and it will become a computer in itself.

The processors used on computing power cards today are somewhat different from traditional CPUs. The CPU is the core processor used in computers. Its main functions are to interpret instructions, process data, and provide command and coordination, which is equivalent to the brain and heart of the computer.

The CPU is based on a low-latency design. It has a powerful arithmetic unit that can complete powerful calculations in a very short clock cycle, which requires it to have a huge cache and computing power unit.

However, Clover's computing power card takes a different approach, because it relies on building a neural network to perform calculations. Therefore, it faces large-scale data of uniform type but not too dependent on each other, as well as a pure computing environment that does not require interruption.

Therefore, at the beginning of its design, it adopted a design based on high throughput and used a large number of computing units, but greatly simplified the control logic and optimization circuits.

It handles complex problems by "transmitting data backwards" after calculation and by networking to implement a "domino effect" linkage method.

This method is actually very similar to a board that has graphics rendering capabilities - a graphics card.

This was indeed the case with technological development in the previous life. Early neural networks were trained using CPUs, but later on, all types of models were replaced with GPUs, or graphics processors, especially 3D hard-accelerated graphics processors, which eventually became large networking units for building neural networks.

Zhou Zhi was too lazy to take that detour, and the function of the product he developed was actually similar to that of the later 3D computing power card, but it was simply called a computing power card.

But now if we want to make the computing power card have the functions of a computer motherboard and CPU again, the design must be changed back. Fortunately, it is not really intended to be a personal computer. It is just a graphical processing terminal, which is highly consistent with the previous functions of the computing power card, so there is no need to change too much.

But the cost is higher than that of a pure character processing terminal anyway.

Zhou Zhi had no choice but to patiently persuade Li Guangnan that we now have the production capacity of computing power cards and the production capacity of the two chips. As for the other various interface protocols, we have very good relations with Sony, Samsung, and Panasonic. In addition, we also have the Type-C interface, so it won’t be a big problem for everyone to share resources. (End of this chapter)