Zhang Ran heard Li Liyou directly say the RISC-V instruction set, and said in surprise: "Do you know this instruction set?"

As the CEO of a chip company, Li Liyou is well aware of the dynamics of the chip industry, not only knows the RISC-V instruction set, but also knows its ins and outs.

In 2010, a team of researchers at the University of California, Berkeley, was ready to launch a new project that required the selection of a processor instruction set. They analyzed multiple instruction sets such as ARM, MIPS, and X86, and found that these instruction sets were not only becoming more and more complex in design, but also had intellectual property issues and could not be used casually. So Berkeley's research team formed an improvised team of four to design a new instruction set from scratch!

The RISC-V instruction set is designed to be very simple, with less than 50 instructions in the base instruction set, but it can already be used to implement a processor with basic functions such as fixed-point arithmetic and privileged mode. Users can also customize new instructions according to their own needs if they want. In this way, the set of instructions is both simple and flexible. What's more, the Berkeley research team has completely opened up the RISC-V instruction set and made it open source, so that RISC-V is not subject to licensing fees like other instruction sets.

Because RISC-V is open source and free, numerous academic institutions can develop processors that are compatible with the RISC-V instruction set without paying a dime. As a result, academic institutions such as the University of California, Berkeley, Massachusetts Institute of Technology, Princeton University, and others have joined RISC-V.

Li Liyou explained: "Spreadtrum's CPU is the same as Huawei's HiSilicon, which uses the ARM public architecture. We spent $5 million on licensing fees alone, which can only be used for five years, and we have to pay ARM for each chip. We were reluctant, but there was no way. After learning about RISC-V, an open-source instruction set, I researched it and found that this instruction set is very good, summing up the lessons of predecessors, avoiding a lot of historical baggage, and achieving a simple but not simple effect, and it is free. I've thought about whether Spreadtrum will be able to develop a new mobile phone chip based on RISC-V? But I just thought about it, Spreadtrum's capital and technology are limited, and it is too difficult to develop a new chip with a new instruction set! In fact, I think this can be done by the country, the country has always emphasized independent research and development, but there is no market for its own research and development architecture, and the use of ARM and MIPS and other architectures must be controlled by others, if RISC-V is adopted, these problems can be well solved. ”

Zhang Ran stared into Li Liyou's eyes and asked, "Do you think it is feasible for Century Summit to develop VR chips based on the RISC-V instruction set, or to bet on RISC-V?"

Li Liyou understands why Zhang Ran wants to bet on RISC-V, the real giant does not want to be controlled by others. Others can get stuck in the neck, and others in the instruction set can also get stuck in your neck. Take Huawei as an example, if the United States says that we will not sell CPUs to Huawei, Huawei is not afraid at all, and it has HiSilicon in its hands. But if the United States puts pressure on ARM and ARM's authorization is not used by Huawei, then Huawei will be really dumbfounded.

In the early years of Android's launch, Google required applications to be written in Java and run on the Dalvik virtual machine. Many low-end processors struggle to run virtual machines, and virtual machines take up a lot of memory, which increases costs. But Google would rather run the risk of putting Android at a competitive disadvantage than promote the Dalvik virtual machine. Because Google doesn't want to see ARM dominate in the field of mobile phones, it hopes that MIPS, x86 and other CPUs can also run Android.

Zhang Ran hopes that VR will adopt the RISC-V instruction set, obviously because he does not want to be controlled by others, and truly takes his fate into his own hands. Li Liyou was greatly impressed by this, this person's vision and courage are amazing!

Li Liyou pondered for a few seconds and said: "It's feasible, because VR is a new field, it's a blank field, everyone is starting from scratch, RISC-is charged, and the cost is very high." For example, our Spreadtrum processor is the ARM architecture that we bought, which cost almost 5 million, and in addition to paying them for each chip. A lot of big companies are very unhappy about this, but they don't have any other choice. Now that ARM dominates the mobile phone market, RISC-V has no chance at all, but it is different in the field of VR, ARM is at the same starting line as RISC-V, RISC-V is open source and free. I believe that many companies will abandon ARM and switch to RISC-V. With these large companies involved, it is entirely possible for RISC-V to build its own ecosystem. ”

Li Liyou paused and said: "But the risk is also very large, the RISC-V instruction set is a new instruction set, a blank sheet of paper, and everything has to start from scratch." I think it will take at least five years from chip design, to tape-out manufacturing, to packaging and testing, to the final market. And it takes longer to build and build an ecosystem. In this process, it is a strong alliance, and many manufacturers are involved. When they move into VR, they quickly build their own ecosystem. If we build an ecosystem slower than theirs, we're a complete loser!"

In fact, Zhang Ran is not good at the instruction set at all, and he can even say that he doesn't understand it, but in his opinion, the competition between chips is similar to the competition of mobile phone systems, and it is not exactly a technical competition. Many people are asking, why can't China make its own chips? In fact, it's not that it can't make it, but that there is no ecosystem, and if it is made, there is no way to promote it in the market.

As an expert in the Internet, Zhang Ran knows that the key to Android's ability to defeat other competitors and dominate smartphones is that Android is open source, and there is a steady stream of developers developing apps for Android, allowing Android to quickly build an ecosystem.

This is also the case with instruction sets, other instruction sets are either like X86 and don't let you use them, or like ARM, you need to pay a high license fee, so to speak, all of them are closed-source.

Therefore, Zhang Ran has reason to believe that when an excellent open-source instruction set appears on the market, this instruction set is likely to be sought after by countless developers like Android and become the mainstream of the market.

Zhang Ran was satisfied with Li Liyou's answer, this person did understand the RISC-V instruction set, so he said: "This is a gamble." Now everyone is betting, Google is betting on artificial intelligence, Microsoft is betting on cloud computing, and other giants are also laying out, everyone is betting on the future, and they are betting on the future, but we are betting on VR. Since we want to bet, then let's just bet big and place a bet on RISC-V. If we lose, we will maintain today's pattern at most, but if we win, we will no longer be controlled by others, and we will surpass Apple and Google and become the hegemon of the network!" He stared into Li Liyou's eyes and said slowly: "Is Spreadtrum willing to take over the development of VR chips based on RISC-V architecture?"

Li Liyou swallowed his saliva and said in a trembling voice: "Spreadtrum lacks funds and technical talents, and it is very difficult to do such research and development with the existing strength, and the progress will be very slow. If we have enough money, if we have enough talent, no, as long as we have enough money, we can poach people if we have enough money. If there is enough financial support, I am willing to take over!"

Zhang Ran said with a smile: "Money is not a problem, we will invest $10 billion in Spreadtrum in the five years after the completion of the acquisition, and we can increase the investment if necessary." After five years, the investment will continue to increase. Spreadtrum can take money to set up labs in the United States and Europe, or it can poach top talent from other companies to develop VR and mobile phone chips. My only requirement is to come up with excellent chips as soon as possible, of course, I don't set a time, how long do you want to come up with chips, it's a new field, and the first-mover advantage is very important!"

When Li Liyou heard that he had invested $10 billion in five years, he was overjoyed, and he was really a big deal, and Spreadtrum was finally going to get rid of the poor days of buckle and rope! It was indeed right to accept the acquisition of Century Peak! He said excitedly: "Mr. Zhang, with the support of Century Peak, we will definitely be able to develop excellent VR chips! I assure you that within five years, if Spreadtrum cannot develop a VR chip based on the RISC-V architecture, I will come to see you!"

Zhang Ran very much hopes to make a breakthrough in VR chips in a short period of time, and if Spreadtrum can come up with VR chips as soon as possible, it will be able to form a cycle of R&D and expansion, and quickly seize the market. Even if Qualcomm and Intel want to catch up, it will not be easy. However, Zhang also understands that chip research and development is not an overnight thing, and waved his hand with a smile: "What do I want your head to do? I want chips, and I said, I don't have a timetable, you just do it with your heart!" He added: "VR is the future, chips are important, but mobile phone chips are equally important." On mobile phone chips, I hope to catch up with MediaTek and even Qualcomm as soon as possible!"

Li Liyou knew that Century Summit acquired Spreadtrum because Taiwan Bend banned MediaTek from selling chips to Century Peak, and Zhang Ran and the management of Century Summit were very dissatisfied with this and held their breath. He is very confident in this: "The chip thing is fed by money. In 2008, Apple bought a small fabless semiconductor for $278 million, and then acquired several chip companies, and recruited various chip designers through high salaries, which led to Apple's chip all the way. Although we can't buy and poach people like Apple, and it is difficult to come up with top-of-the-line chips like the A series in a short period of time, it is not a problem to surpass MediaTek. ”

Zhang Ran laughed: "I look forward to the good news from Spreadtrum. ”

After Li Liyou left, Zhang Ran chatted with Wu Ruide for a while. Although the peak of the century has pressed the treasure on the VR and RISC-V instruction set, it must not be hanged from a tree. Next, Century Summit will also acquire or invest in other chip companies, ARM and other structures will bet. In this way, even if RISC-V fails, Century Summit can quickly move to other architectures.

On September 11, Spreadtrum Communications announced that it had received a $1.38 billion takeover offer from Century Summit. Under the terms of the offer, Summit will pay $28.50 per U.S. Depository Share for Spreadtrum Communications, representing a 28% premium to the previous day's closing price of the stock. The price is close to Spreadtrum's all-time high of $29.29, and the total value of the stock acquisition transaction is $1.78 billion on a fully diluted basis. It is also the largest takeover offer for a semiconductor manufacturer in about a year.

Following the announcement, Spreadtrum shares rose $3.91, or 17.54%, to $26.20 during regular trading on the Nasdaq, after hitting a new 52-week high of $26.69. Over the past 52 weeks, Spreadtrum Communications has traded at a high of $26.69 and a low of $14.50.

In the eyes of many people, the impression of Century Summit is that it is rich and capricious, constantly buying, buying, buying, and even giving Century Peak a nickname, called "Buy, Buy, Buy, Buy Company". The 1.8 billion acquisition of Spreadtrum is not a big deal in the acquisition of the peak of the century, but only a medium one. But after this news came out, it still attracted huge attention, after all, chips have always been a pain in China's throat.

Overseas media believe that after the acquisition of Spreadtrum, Century Summit will challenge MediaTek and Qualcomm, and MediaTek will be particularly impacted. The domestic media also reported the news for the first time, and many media are asking, what does Zhang Ran want to do? What impact will this have on China's chip market? Some media even put forward an exaggerated point of view, China's chip problem must be solved by the peak of the century, Alibaba, Tencent, and Baidu!