Although his gentle attitude towards talents cannot be called a polite corporal, it will not disappoint anyone.
The initial communication between the two parties was smooth.
Wen Liang expressed his expectation for Lin Tong to join.
Lin Tong also expressed his expectation to join Bolang.
Everyone is happy.
In fact, Wen Liang only knew the general research direction of Lin Tong when he was studying in Berkeley, California. He did not pay much attention to the specific research projects that Lin Tong would be responsible for after joining.
Two factors.
First, the technical directions covered by Bolang Laboratory are very rich, including a large part of science and engineering, and even literature, history and philosophy.
Second, it is rare for Qin Zheng to formally recommend talents to the company. Lin Tong is a doctor. At this level, Bolang Laboratory always has positions that meet his research direction.
Of course, as the direct person in charge of Bolang Laboratory, with Lin Tong's academic level, Wen Liang always needs to ask a few questions about the process.
After finalizing the basic intentions, Wen Liang said enthusiastically: "Dr. Lin, on behalf of Bolang Laboratory, I welcome you to join."
Lin Tong responded with a smile.
Wen Liang turned to get straight to the point: "Sorry, our process is a bit out of place today. Now we have to talk about something mundane. I wonder what the annual salary Dr. Lin expects is?"
“Are separate group studios required and where are the expected work locations?”
"What is your intended research direction?"
He asked three questions in a very straightforward manner.
Qin Zheng on the side, including Lin Tong himself, only just realized that the basic process of joining a company was not discussed.
Lin Tong scratched his head slightly and smiled: "Sorry."
Then he answered the questions raised by Wen Liang one by one: "I hope to work with Qin Zheng and be responsible for the independent research and development of the graphics processor (GPU) in the mobile phone processing SoC. This is also the goal of Qin Zheng inviting me to join Bolang."
“It’s less than a year since I graduated with my Ph.D., and the average annual salary in Silicon Valley for people with corresponding experience is US$75,000. I’m looking forward to an annual salary of 400,000 RMB.”
Although Lin Tong has a doctorate, it can be said that he has no work experience.
The salary that Yang Xingshan and Dai Li can get in the Silicon Valley market is almost US$50,000 a year lower than that.
However, his academic qualifications are higher than those of Yang and Dai, who both have master's degrees.
Of course, Lin Tong is not the first Ph.D. to join Bolang Company, but he is indeed the first Ph.D. to have overseas study experience.
In fact, Bolang has many high-quality research talents, and if we count participation in Bolang research projects, more than 1,000 PhDs have participated so far.
There are also 39 academicians of the two academies.
Bolang invested the first batch of 1.6 billion research budget funds in the name of Bolang Terminal, which is still very impressive. The extra 100 million budget funds were allocated to Dr. Zhang Xingqiang’s EUV lithography light source project.
However, the expected salary level proposed by Lin Tong far exceeds the average salary level of domestic doctoral graduates, roughly doubled, and even exceeds the average salary level of American doctoral graduates.
Although according to the Silicon Valley market salary that Lin Tong can get, he has already lowered his worth in order to join Bolang and to match the domestic environment.
Wen Liang's focus was on the research direction mentioned by Lin Tong. He looked at Qin Zheng and then at Lin Tong: "Doctor, your intention is to independently develop mobile phone GPUs?"
"Yes." Lin Tong nodded seriously, "I have accumulated a lot of research in this area and have participated in joint experiments related to female IDIA."
Qin Zheng said from the side: "Dr. Lin has this ability. During our visit, we often exchanged academic experiences and learned a lot."
Wen Liang suddenly felt heartbroken.
He felt that Xiao Qin was pretending to be very mellow this time.
This guy is only a junior this year, and yet he can so calmly comment on the level and strength of an outstanding doctoral graduate from Berkeley, California, and say that he has learned a lot from frequent exchanges of academic experiences?
grass!
What a genius~
"It seems that your research directions match well." Wen Liang said with a smile, "In this case, the annual salary will be set at 600,000, and a human resources specialist will coordinate with you about other benefits and benefits."
"I can't destroy our Bolang sign that we value talents, so I won't wrong Dr. Lin."
Lin Tong: "..."
I am a science and engineering man and my thinking is correct, but I am not stupid!
Where do you leave it here and here?
He never expected that the process of joining Bolang would be so tortuous, and that he would be used to show off~
Of course, for the sake of money, Lin Tong said he could remain silent.
An extra 200,000 RMB a year is still very comfortable.
Because Lin Tong was responsible for researching GPUs, Wen Liang became more talkative.
"I wonder if Dr. Lin is interested in the GPGPU (General Graphics Processing Unit) direction?"
Hearing this, Lin Tong said straightforwardly: "GPGPU must be the future trend. Using GPU only to process graphics is a huge waste of computing resources;
From the introduction of the concept of Shader, to IDIA's launch of the CUDA computing platform, Apple's launch of the openCL framework standard, and the realization of GPU-oriented programming, GPGPU has become an inevitable development trend."
At this point, Lin Tong paused briefly and further said: "In fact, GPGPU is the direction I focus more on research. I hope to gradually transition to the research and development of mobile GPUs by completing the research and development of mobile phone GPUs."
GPGPU is not actually hardware, it can only be regarded as a concept.
In a broad sense, any GPU in the IDIA product series that supports the CUDA computing platform, or other GPUs that support the openCL standard framework, is considered a GPGPU.
It can be understood that after supporting these standard frameworks, the GPU can be abstracted into a stream processor by writing programs, and its massive arithmetic operation units can be used to perform general calculations such as scientific operations and data analysis.
Wen Liang listened happily and nodded in praise: "Not bad, not bad. I very much look forward to your in-depth breakthroughs in these directions."
"Bolang also needs to complete its technology accumulation in these necessary blank areas."
Lin Tong did not shout slogans, but pushed up his glasses and said calmly: "I will work hard."
Wen Liang nodded lightly and did not delve into more specific technical directions.
He doesn't know much either.
I just paid a little attention to it as a man in science and engineering in my previous life, and I have seen Bolang's integration of technical information on related aspects.
…………
Lin Tong went through the entry procedures under the personal guidance of Yuan Shiya.
Tang Pu, Li Bowen, Wang Zhiyi, and Zhao Mingjie, who were in the company today, then came to the general manager's office.
Chen Jiaxin...Jiaxin has been quite busy recently.
As the legal representative and external spokesperson of Bolang Group, Chen Jiaxin now has quite a lot of things to deal with.
In the recent stage, I have mainly attended some necessary business summits and the like.
There was a lively chat in the office at first.
Then everyone sat down separately. Wen Liang picked up a pot to make tea and said with a smile: "Xiao Qin gained a lot from this study tour and brought back an outstanding talent to the company."
"Can't you tell~ Xiao Qin can actually pull people?" Zhao Mingjie immediately became angry.
Wang Zhiyi pushed up the glasses frame on the bridge of his nose, pursed his lips and said, "Xiao Zhao, this is because you are ignorant. Our Xiao Qin is a genius, and geniuses naturally attract talents, unlike you, who is a jerk."
Li Bowen said in a good-natured tone: "Why are you attacking me personally?"
I actually watched the show with my arms crossed.
Tang Pu next to him also made the same move, he didn't even bother to say anything, and just showed off to watch the show.
The 'grievances' between Wang Zhiyi and Zhao Mingjie are inextricable.
These two things are either attacking each other or showing off to each other.
After a while of fussing, Tang Pu asked: "Can we, Mr. Wen, tell us that Xiao Qin has brought back a very good talent?"
Wen Liang nodded slightly: "Doctor from Berkeley, California, will be responsible for self-research of GPU for mobile phone SoC."
"What the hell?"
"Fuck!"
Suddenly there was a lot of noise.
Although GPU is slightly less difficult to develop than CPU in terms of composition structure, after all, under normal circumstances, GPU resources are allocated by the CPU.
And because GPUs generally do not need market adaptation, even GPGPU uses a dedicated standard framework for adaptation programming, and each company has its own standards.
Therefore, under the current situation, Bolang is not so anxious about the need for self-research of GPU.
but!
However, Bolang's plan is to fully control the self-development of each module of the SoC.
Therefore, Tang Pu and the others were a little excited when they heard the news.
"It's also a coincidence." Qin Zheng said solemnly.
Hehehe joked a few words.
Wen Liang handed the brewed tea to several people, and then said: "Our mobile phone business is advancing step by step. In principle, the specific time planning is hoped to be combined with your research progress, Xiao Qin."
"In this way, the first and second generation products will have a better transition and connection."
Tang Pu also helped explain: "Because none of us had expected the actual difficulty of self-developed chips, the timeline at that time may be slightly different from the actual situation. From the company's perspective, of course we hope to have a clearer
The timetable will facilitate the coordination of resources in all aspects.”
Neither Wen Liang nor Tang Pu meant it euphemistically.
It doesn't mean anything else.
Of course Qin Zheng understood, he smiled and said: "I understand, Old Tang doesn't need to be so cautious, Mr. Wen is very frank."
"Yes, so we, Mr. Wen, are the core." Tang Pu was not embarrassed and was very calm.
Wen Liang just smiled softly and didn't say much.
Qin Zheng sorted it out and said: "I have some thoughts on this matter. Let's talk about the technical progress first. The research before September is invalid in terms of results, but the results are not invalid and have provided a lot of successful experience.
;
During this study visit, Academician Lai Cao helped with coordination, and I learned a lot about cutting-edge academics;
According to the current progress, with the resources of Huagong Integrated Circuit Design Laboratory and the support of the company, it can take as short as 4 weeks and as long as 6 weeks to complete the first version of the instruction set system;
We don’t need to design different specifications of ores (cores) under the same instruction set architecture like ARM does, because we use it for our own use and can just design one specification of ores.”
After a pause, Qin Zheng continued: "After I started my study tour, the company accelerated the design of a small amount of ARM architecture SoC and released the film. This means that the relevant teams of Huagong IC Laboratory and Bolang Laboratory have completed the running-in.
Work;
It also means that everyone has successful experience;
Based on these factors, the design of the first-stage instruction set should be completed by the end of this year at the latest, the design of the second-stage CPU ore should be completed by the end of the first quarter of next year, and the third-stage finished product design should be completed by the end of the second quarter of next year;
After that comes the technical coordination with the foundry and the initial tape-out work, which is expected to take a minimum of 45 days and a maximum of two months.”
"That is, the first chip will be rolled off the production line no later than the end of August next year."
"Of course, this time does not take into account resource and capital investment restrictions."
In fact, based on this time, the time cost of the first chip is already very high, a full 20 months.
Not to mention the capital investment.
You must know that now Qualcomm, Apple and other design manufacturers can produce at least one generation of products a year, and many can produce two or three types of products.
But after all, Bolang is a newbie, so it’s not surprising that after investing a few hundred million in the initial stage and then another few hundred million, it won’t work.
After listening to Qin Zheng's explanation, Wen Liang thought about it briefly and asked, "Can Dr. Lin finish the GPU part on time?"
"It should be much earlier." Qin Zheng replied, "The difficulty of GPU design will actually be a little lower. Optimizing this aspect will not be too complicated if we follow the previous violent mode."
After receiving the answer, Wen Liang made arrangements: "Old Tang, Xiao Li, you will coordinate in the middle. After Xiao Qin and Dr. Lin complete the design of the instruction set system, we will subdivide different R&D groups. In the initial stage, there will be no less than five groups. Normally,
We will equip three groups of people to develop at the same time, and I will solve the problem of capital consumption."
"clear."
"understood."
Tang Pu, Li Bowen and Qin Zheng nodded one after another.
Wen Liang's meaning has been very clear. Don't be stingy in investing in things that can be solved by spending money. The most troublesome things are those that cannot be solved by spending money.
For example, the process of creating a CPU from scratch relies on Qin Zheng's brain.
It's as if the system kernel relied on Zhang Yulin's brain from scratch.
In short, when Qin Zheng returned to China, the timetable for major research and development projects was set -