Luo Sheng came to the R&D department of Lens Semiconductor, and Xu Zhijun, the head of the company, was also there, and he took Luo Sheng to the Lh06xx server chip R&D department, which is the internal R&D code name of Lens Semiconductor, and the two entered the laboratory in professional overalls.

The engineers in the laboratory are burying their heads in their own things, Xu Zhijun said: "Compared with mobile phone SoCs, the difficulty and threshold of making server chips are much higher, and we are still making non-x86 architecture chips, in today's server market, x86 architecture processor market accounts for more than 90%, it is not easy to tear off a piece of meat from Intel in such a stable ecosystem." ”

Speaking of this, Xu Zhijun couldn't help but let out a long sigh, the x86 architecture is almost unshakable, and it takes great courage and determination to challenge such a hegemon.

However, Luo Sheng was quite optimistic, and said with a smile: "Not only us, the world x86 has been around for a long time, and there are also a number of Arm manufacturers in foreign countries who are trying to change this situation that makes many server application providers feel helpless, because they have little bargaining power anywhere in the x86 series." ”

It is worth mentioning that the R&D code name of Lh06xx within Lens Semiconductor is a server chip based on Arm architecture, which is the second generation of server chips developed by Lens Semiconductor.

Another R&D department of Lens Semiconductor is also working hard to develop and study its own underlying instruction set architecture, not only that, but the engineers of the Cte Coast Laboratory are also working together, and the engineers of the Cell chip after seeing it have transferred 127 people to lend to Huawei, most of which are the disciples taught by Luo Sheng during the [ACC plan project], which is an important part of the spare tire, and the Arm architecture is after all something from Europeans and Americans.

Now not only Huawei's server chips are based on the Arm architecture, but the "Blue Core L" series of Azure mobile phone processors on the Cote d'Azur are also based on the Arm architecture.

However, the self-developed underlying instruction set architecture is still a long way from replacing the application stage of the Arm architecture, and it is already a great success for the Huawei server chip to get rid of the Arm architecture in the fifth generation.

In other words, the third and fourth generations have to continue to use the Arm architecture.

Huawei's ability to spend money on scientific research is actually not vague than the Cote d'Azur, with total sales of US$32.5 billion last year, but R&D investment accounted for more than 30%, reaching an astonishing US$10.075 billion, second only to the Cote d'Azur (US$39.19 billion) and Bluestar Technology (US$12.1 billion) in China, ranking seventh in the world.

It is precisely because of the courage to invest heavily in R&D investment, and the profit is not as good as that of the Cote d'Azur, let alone Bluestar Technology, so Huawei's net profit last year was only 10.6 billion yuan.

But even so, Huawei still feels that R&D investment is far from enough, otherwise it would not be so eager to establish a joint venture with Luo Sheng again to build a Xiongxin Electronics, and set up an independent brand Glory mobile phone to expand new revenue growth points.

Lack of money!

Huawei's R&D investment is mainly focused on the research and development of communication baseband, and the server project was originally an edge business that did not have much R&D budget.

Xu Zhijun said with a smile: "In the past five years, Huawei has also made great progress in the research and development of Arm server chips, and detours are naturally unavoidable, and the first generation of chips launched is obviously unsatisfactory, it is such a garbage product, and you did not hesitate to give Huawei a super-large-scale order for seven data centers for Huawei servers, otherwise the server business sector may be half-dead now, without you to give people a list, how can there be such a satisfactory progress rate today." ”

This is the result of mutual fulfillment, the first generation of Huawei server is really bad, expensive and not as good as IBM's products, but Luo Sheng still insists on the full support of Bluestar Technology, it can be said that the only customer of Huawei server business is Bluestar Technology.

However, just such a customer can top dozens of other large customer groups, and Bluestar Technology is simply a super large customer.

To a certain extent, Bluestar Technology is constantly transfusing the Huawei server business group, and Luo Sheng wants to forcibly support it, and even completely replace IBM in North America.

And Huawei server department did not disappoint Luo Sheng, he got the order here, and the repatriated funds did not hesitate to smash into the research and development of the new generation of products, and gradually formed a virtuous positive cycle, which is obvious to all in the past five years, and the latest generation of products is leapfrogging compared with the previous generation.

With five years of R&D and market experience, the server department has passed the novice period and gradually entered the fast lane of rapid growth.

After walking around the laboratory a few times, the two also came out and sat down in the lounge of the R&D department to continue chatting and communicating.

Luo Sheng said: "The LH06xx version of the chip is a high K45 nanometer process, and Xiongxin Electronics' second super factory in Pengcheng is expected to complete the establishment of the production line in the first quarter of next year. ”

Xu Zhijun nodded affirmatively: "You don't have to worry about this, I can pack the ticket, there will definitely be no problem, now we have begun to discuss the naming of the official release of the second-generation server chip internally, when it comes to this, Mr. Luo, what do you suggest?" ”

LH06xx is only an internal R&D codename, while its final version is identified as LH0611.

Luo Sheng began to lay out the whole industrial chain of domestic semiconductors, and now as long as it is not the most high-end products, most of the production environment can be realized in the mainland supply chain, and it is still being improved.

Hearing this, Luo Sheng thought for a while and said: "On the side of Bluestar Technology, we are going to build four submarine data centers from 2012 to 2015, one a year, respectively, the Bohai Rim Economic Belt, the Yangtze River Delta Economic Belt, the Pearl River Delta Economic Belt and the bottom of Qinghai Lake in the western region of the country to sink down, and the data center of Qinghai Lake will be able to meet the needs of the western region for at least five to eight years. ”

The life cycle of a subsea data center is designed to be a five-year cycle.

Xu Zhijun was surprised and said, "What? Isn't this underwater data center project a secret project that you started just two years ago? Ready for commercial use so soon? That's not what you said in previous media interviews. ”

Lao Xu was really surprised, this R&D and output ability made his heart feel a huge shock, this is something that no one has done before.

Luo Sheng said with a smile: "The original plan was to go to 2014 or 2015, but there is a genius college student who is still studying for a doctorate at HKUST who has solved several key core technical pain points in half a year since the beginning of this year, which has greatly advanced the plan." ”

Xu Zhijun's eyes lit up, and he subconsciously said, "Oh? What is the genius of HKUST called? ”

Luo Sheng glanced at him lightly: "You don't have the idea of beating this person for me, he is already my person, but he is just raised in the University of Science and Technology first." The other one who fought with me for financial resources, yes, is that the group of people at Microsoft have already knelt down and sang about conquest. ”

Xu Zhijun was stunned for a while, and suddenly couldn't help but laugh: "Haha, look at what you said, a gentleman loves talent but doesn't win people's love, but you need to borrow it and use it, you can't shirk it, otherwise I can sue you for breach of contract." ”

Luo Sheng couldn't help but smile and nodded silently.

After the two sides reached a strategic partnership in 2004, they have established an agreement to break through the talent barriers of each other, so to a certain extent, your talent is my talent.

Over the years, it has been a common thing to borrow and transfer people from each other, and it is precisely because of this paper contract that the human resources of both parties have been greatly optimized, and the return is more efficient R&D output capacity and more money-saving.

You know, it is very expensive to rob people, for example, if you rob people with Google, you will give each other five times the treatment at every turn.

Luo Sheng said: "How about the official release of the name 'Kunpeng'? ”

"Okay, I can still make a decision on the naming rights." Xu Zhijun nodded very dryly.

In fact, when Luo Sheng said that he would build four underwater data centers, he knew that the other party would say this name, because the underwater data center project of Bluestar Technology was made public very early, and the name was "Kunpeng Project", and the second-generation server should first be applied to the underwater data center, which is a perfect match.

Xu Zhijun continued: "It is no surprise that the 'Kunpeng' server will be launched early next year, and we have been doing the ecological construction of Arm servers before and after the launch, in the face of the strong x86 series ecology, if you want to get customer orders, you definitely can't just rely on one or two processor chips and two or three server products, the penetration of industry organizations and platforms to participate in the construction, the collaboration of related hardware and software, etc., are very important, and it is also the most difficult to do." ”

Luo Sheng nodded and said: "Arm server manufacturers still have something to rely on, that is, the market has this demand, and it is necessary to break the monopoly of toothpaste factories." ”

Xu Zhijun said with a smile: "Sure enough, the hero sees the same." ”

Luo Sheng has made it clear that he wants to guarantee the order contract for the four submarine large-scale data centers of the 'Kunpeng' server, and as for whether it can open up new business markets, it is a matter for the business group.

Bluestar Technology's orders can only guarantee survival, but whether you can live a famous life depends on your own business ability.

However, Luo Sheng is still very confident in Huawei.

……

（Ps：emmm…… There have been three circuit breakers in the history of the United States, and the last circuit breaker event was traced back to March 9, 2020, when I became one of the witnesses of history, and I still remember it very deeply, as if it happened just a few days ago...... [Manual Funny])

()

Chinese Net