Rewrite the Technological Landscape

Chips are actually a very wide range of things. From high-end to low-end, from civilian to military, there are thousands of chips, but for most people, the basics they know are computer chips and mobile phone chips.

Meng Qian worked as a software developer and architect in his previous life, but he has never worked in the chip industry after all, so he doesn't know much about it. Otherwise, he would have used a few low-end chips to practice his hands.

But fortunately, he is a technical geek, and he is a little more interested in the field of technology than ordinary people. Coupled with his own technical level, he reads a lot of content with exploration, so although he does not know much about the chip field. It may be so comprehensive, but more or less besides computer chips and mobile phone chips, I still understand some commonly used chips in other fields.

Just like a CMOS image sensor. Because the later application scenarios of this thing include mobile phones, digital cameras, automobiles, security and other common fields, it is also a quite influential chip, and in terms of the process itself, CMOS image sensors are also high-end chips.

Combining the relatively in-depth understanding of the previous life and the hard work of learning through various channels in this life, Meng Qian finally figured out this thing. Of course, there is another very important thing here, that is, the development of lithography machines It gave Meng Qian the opportunity to get to know some big names in the domestic semiconductor industry. The new design of this CMOS image sensor chip architecture is inseparable from the help of these big shots.

"Have you ever been exposed to CMOS image sensors at Sony?" Meng Qian asked, and everyone who came back from Sony nodded.

"Then do you know what advantages CMOS image sensors have over CCD image sensors?"

This time, everyone shook their heads.

Meng Qian once again confirmed that everyone really doesn't know much about this thing, "CMOS image sensor chip, as the name suggests, uses CMOS technology, so it can integrate the image acquisition unit and signal processing unit on the same chip, and CCD image sensor chip Compared with sensors, its advantages are quite obvious, low power consumption, low size, and low price.

In the past two years, why did CMOS image sensors suddenly gain momentum, especially why did Sony care about this thing so much? Because of the catalysis of a product, that is the mobile phone camera.

Of course, the advantages of CMOS image sensors are also applicable to other personal devices, but before the production was insufficient, the price advantage was not great. The emergence of mobile phones and the increase in production have made the advantages of CMOS image sensors more obvious.

However, CMOS image sensor technology has only been commercialized for only 7 years, and there are still many immature technologies, so the reason why I chose the core of this product is because it has a broad future, and because it now has many There is room for improvement. "

When Meng Qian said this, he began to draw on the blackboard, "CMOS image sensor is a modular design concept, including photosensitive area array, timing control, analog signal processing and analog-to-digital conversion modules, the most important of which is the photosensitive area. Array, whether it is Micron or Sony, the photosensitive area arrays of all CMOS image sensors all adopt front-illuminated technology.

The light enters the metal area after passing through the microlens and color filter, then reaches the light receiving surface, and then passes to the photodiode. Can anyone see the problem with this design? "

After waiting for half a minute, no one spoke. Sure enough, in any era, even the most professional people can hardly see what is wrong with the most advanced technology in this era, but from the perspective of future generations, those technologies are like mentally retarded people. , but at the moment, this is the most advanced technology.

"The biggest problem with the front-illumination technology is that the light will be affected by the metal wires, and the signal-to-noise is naturally very low, so I thought of a new method, the back-illumination technology, which puts the photodiode on the back of the metal, so that the light will not Affected by metal wires, but this design will involve another issue, frame rate.

To put it bluntly, the frame rate is how many photos can be displayed per second. After using the back-illuminated technology, the pixels will definitely increase, but if there is no matching high-speed readout and processing circuit, then the previous changes are useless, so I am very concerned about the sensor architecture. Made a very important change.

Now all architectures in the world have only one ADC analog-to-digital converter, but I plan to equip each pixel with its own ADC.

"

It took Meng Qian two hours to give a detailed explanation of this new architecture design to the company's employees.

After everyone had a preliminary understanding, Meng Qian issued a task, and everyone went back to work.

After arranging the CMOS image sensor, Meng Qian continued to go back to work on the database. At the end of April, the company carried out a large-scale promotion of middle management. The first batch of people who joined the company basically worked hard enough to be promoted to middle management. s position. No matter how bad it is, it is also a supervisor or team leader.

Improving the management post as soon as possible is also to further reduce Meng Qian's energy dispersion in management. For Meng Qian in 2002, he needed to invest a lot of energy in development and research and development.

Just like on April 29th, after taking out everything he could from the database, Meng Qian turned to the company's R\u0026D center.

Since the establishment of the R\u0026D center after the year, there has been an unavoidable running-in period. First, some people in the company applied for internal job transfers, wanting to switch from development posts to scientific research posts. In addition, external recruitment was screened again and again. The R \u0026 D center is considered to have an initial scale.

The R\u0026D center now has a total of 136 people, because Meng Qian’s requirements are relatively high and the recruitment efficiency is relatively low. Among them, 26 people were transferred from within the company, 50 people were introduced by others, and only 60 people were actually recruited by the company.

Of course, another reason for the lack of social recruitment is that the R\u0026D center currently does not have many projects, and personnel are recruited according to projects. There is no way to do this, it has to be done step by step. It is impossible for Meng Qian to throw hundreds of projects at the R\u0026D center all at once.

The first batch of projects he issued will be used in the next few years, such as architecture design in the semiconductor field, such as the development of core components in the computer field, such as touch screen technology in the mobile phone field, such as artificial intelligence and cloud computing.

During the following period, Meng Qian devoted himself from the commercial software department to the R\u0026D center, especially in the fields of artificial intelligence and cloud computing. He brought a lot of very precious things to researchers.

Dafeng Group is gradually developing into a real comprehensive technology company.

As soon as the time turned to May 6th, Meng Qian, who had slept in the company for another night, was woken up by He Yating, who had just come to work, and washed up briefly. "How did I forget about this?"

Meng Qian saw a piece of news in the newspaper: "Separation of Telecom Group".

The original Huaxia Telecom Group was divided into north and south regions, and the new Huaxia Telecom and China Netcom Groups were formally established. The backbone network of Huaxia Internet will be divided into two, with Netcom Group managing the north and Huaxia Telecom managing the south.