Rebirth: I carry a desktop computer inside me

After the National Day, Professor Luo Ruihua, who has always been haunted, finally returned.

According to him, he went to carry out a scientific experiment, about quantum computing.

When I saw Jiang Cheng, I directly took the school exam, and as usual, it was a blank piece of paper, which involved problems in machine learning and pattern recognition.

In the past few months, Jiang Cheng still spent most of his time on chip research and development, and the content research of machine learning was somewhat relaxed.

Fortunately, the problem was not too difficult, and Jiang Cheng took a while to do it.

Luo Ruihua looked at it and said, "Your study and research have fallen behind for a while.

Jiang Cheng was embarrassed and said honestly, "After the last project, I mainly focused on chip design, and I really didn't spend much time studying."

"Learning still can't be lost, especially the learning of cutting-edge knowledge." Luo Ruihua shook his head, "I know that you have opened your own company, but you must know that our research is not separated from application, but solves basic or mechanical problems, and you should take the initiative to find new research directions."

Jiang Cheng humbly taught, "Teacher, I understand.

"Your research is very broad, but I suggest you choose 1 or 2 directions, go deeper, and if you can, don't just be satisfied with standing on the shoulders of your predecessors and take a small step forward, but be able to open up new fields." Luo Ruihua's expectations for Jiang Cheng are still relatively high.

In his senses, this young man has a flexible mind and the energy to study.

However, the fly in the ointment is that the research direction is somewhat scattered and the frontier is not enough.

Coming out of Professor Luo Ruihua's office, Jiang Cheng looked pensive, did not look at the road at all, and then subconsciously walked to the library.

What direction should I choose to continue my research?

Can you combine your interests and open up a new direction?

His current research is more in-depth in the field of machine learning, including deep learning and convolutional neural networks, in addition, in software design and program development, circuit design and other aspects of research, especially applied research.

There are still not many cutting-edge research, just a little achievement in face recognition and image recognition.

It's still too far from the real big guy.

Artificial intelligence is the future direction, Jiang Cheng naturally cannot discard, but can he combine chip design for research?

Jiang Cheng's impression that AI chips are also a hot spot in the future.

Especially in addition to the traditional processor computing power, it will play a great role in the intelligent processing, recognition and optimization of mobile phones.

Heading to the latest international periodical room, Jiang Cheng stood and looked up.

It took a whole morning.

In his heart, Jiang Cheng was more sure of his next direction of study.

In the past two years of ILSVRC-ImageNet (computer vision competition), many CNN (convolutional neural network) structures have emerged, including AlexNet, ZFNet and other classical structures.

Jiang Cheng intends to combine neurobiology with the study of in vivo computers to propose a new neural network structure.

In the afternoon, the biology books in Jiang Chengcheng's library were first read into the internal computer, ready to study carefully.

Alas!

Jiang Cheng wished that he could also split a few more of himself like a multi-core!

......

----⊙_After the incorporation

of ⊙---- Maple Fire Technology Co., Ltd., the venue of Jinfeng Building has seized time to simply decorate.

Hired employees have moved in, and engineers have begun development of mobile phones.

Of course, for now, it is mainly designed for appearance (ID), structure (MD), motherboard and other components (HW).

Fast charging, baseband, RF chips can all use their own.

As for the battery, it is possible to purchase batteries from the company of Yadi.

Jiang Chenghui was looking at the design drawing on the 32-inch display screen.

5.3 inches in size, 149mm long, 74mm wide, 8.5mm thick preliminary design.

Of course, this depends on whether the final construction design can be implemented in hardware.

In appearance, the four corners have an arc design, the upper and lower border design is relatively narrow, and the screen-to-body ratio reaches 78%.

Judging from Jiang Cheng's aesthetics, the overall appearance is still good.

"Mr. Jiang, the appearance design and structural design of the new mobile phone, do you think how to adjust it?" Mo Tianqing asked next to him.

Jiang Cheng pointed to the border, "Find a way to make the border narrower, cancel the three buttons at the bottom, try to keep it narrow, and make the screen-to-body ratio higher."

Mo Tianqing was puzzled, "Then there is no button, how can the user interact?" Or like fruit and bamboo, set a home button and a small round key? "

Neither." Jiang Cheng shook his head, "The whole screen, except for 1 front camera at the top and distance sensing, don't do anything else, focus on simplicity."

"Okay."

No way, the boss asked, and the following employees will rectify in this aspect.

This is also good that Jiang Cheng mentioned it, if the appearance designer is mentioned, the structural design and hardware design people will not break the mouth.

Just know that opening your mouth to make some unrealistic requirements, how does this make people design?

As for the chip, Jiang Cheng has not made up his mind now, and currently has 2 chips to choose from.

It has been entrusted to MediaTek's MT6592 to release the chip not long ago.

Said to be octa-core, 28nm manufacturing process, core frequency between 1.7-2GHz.

But the performance is actually relatively average.

However, Brother Lianfa is still very chicken thief, the motherboard entrainment chip package the whole set, the price is still cheap, since this year, a lot of miscellaneous cottage machines have emerged on the market.

All of them use MediaTek chips.

As for Warwick's chips, under contact, Warwick expressed the idea of not being ready to sell the latest chips for the time being.

But for the K2V3 model, Warwick still has some inventory, which says it can be sold to Jiang Cheng's Fenghuo.

But this chip, not up or down, Jiang Cheng is a little uncomfortable.

Later, Jiang Cheng looked at the progress of the new system development team for the Android 4.2 version.

Jiang Cheng guided the concept of design as usual, and at the system level design, Jiang Cheng was not very professional, but he could only roughly look at the progress and put forward the requirements.

In mid-October, Kushuai Mobile took the lead in purchasing a full set of modules for baseband and RF chips from Maple Fire Semiconductor.

They purchased MediaTek's chips, this chip is not like Lambcom requires the purchase of chips but also entrained baseband, RF and other series of family barrels, in order to improve the signal, seize the opportunity of 4G full Netcom, after research and judgment, finally reached a cooperation agreement with Maple Semiconductor.

The purchase intention is 30 million sets of baseband + RF chips, and the first batch of 5 million sets are delivered.

The average price of the first batch of complete sets of devices is 108.68 yuan, and it will be reduced to about 90 yuan / set according to factors such as procurement cost and yield improvement.

The first batch is a bulk purchase of 540 million yuan!

Of course, this is only a cooperation agreement, and Kushuai will not pay all immediately, but order it in batches, and if the mobile phone sells well, they will also purchase the next batch soon.

In today's era of Internet mobile phone industry, these hardware costs have been compressed to extremely low profits.

Most of them have to make money by volume.

For Jiangcheng's Maple Fire Semiconductor, with the increase of the production capacity of the foundry, these types of chips are mass-produced, and the average cost will be compressed quickly.

Jiang Cheng estimates that the profit of 6700 million yuan can be completed.

Cool Shuai made a good start, and also dispelled the uneasy atmosphere that Maple Semiconductor had not unveiled for most of the year.

Guo Xiaoming is very happy, how can there be a good product, but no one comes to buy the reason!

Find a foundry and keep going!

---- (ˉ'ˉ) ----

afternoon, Jiang Cheng and Guo Xiaoming arrived at the Kyoto Economic Development Zone by sea.

The design company of the tender had already issued the design drawings in the first half of the year, and the construction company of the tender, after the various certificates were issued, began to enter the construction.

Jiang Cheng wore a hard hat and checked the progress of the construction.

Xiong Weiyong, a manager of China Construction Third Bureau, is accompanying this young big customer.

You know, the construction period is at least 2 years, and the construction cost of the main building alone needs more than 300 million.

"Mr. Jiang, the land has basically leveled at present, and we will take advantage of the favorable weather to speed up the construction progress in the next step." Xiong Weiyong pointed to the large piece of ground in front of him that had been leveled and introduced.

Jiang Cheng nodded, "Well, speed up the progress according to the drawings, money is not a problem."

"Okay."

Jiang Cheng turned his head and saw the BOE headquarters on the side of the road.

At present, the building that has been topped out has been completed, and Jiang Cheng has some emotion in his heart.

Facing each other, there is a cause, turn back and have money, get some shares.

This BOE screen behind the rise, is a contentious master, and it is estimated that there is a lot of cooperation with him behind!