Font
Large
Medium
Small
Night
Prev Index    Favorite Next

Chapter 148 MIPS Patents

Liang Yuan looked at the thin silicon crystal in his hand. There was a twisted Maibius ring printed on the gray-black surface on the back of the processor. It looked very similar to the common ones on computers in later generations, similar to the triangle loop icon on the garbage recycling bin. This thing should be the latest trademark designed by Su Liangyu and others for chip laboratories.

After staring at Maibius ring for a few glances, Liang Yuan felt that this pattern with eternal, infinite development, and even involving the fourth-dimensional space was in line with the great significance of microprocessors to humans. When the information age came, in a sense, the earth's civilization should be a history of carbon-based compounds based on silicon.

The right side of the product logo is printed with capital art form c2, and the words risc are hanging on the back of c2, indicating that this is a processor that uses a streamlined instruction set, but the lower side is a row of production numbers like nh1001127, and the working voltage of this CPU is marked below the numbered numbers is marked with 3.2 volts.

Liang Yuan tossed and turned the CPU in his hand and played with it for a long time. Although there was a lazy smile on his face and it looked calm, the emotions in his heart had already rolled like huge waves. Except for a few people, no one would realize how much impact this small silicon crystal would have on the entire human society.

In the 1980s, it was the critical era when the industrial system of the traditional Western industrial power went from combined industrial automation to intelligent automation with chips as the core. It used the mainstream CPU main frequency to break through 100mhz on a large scale as a sign of the maturity of intelligent automation. This technology was only five or six years left. The gentle wind of the information age has come.

Now, Liang Yuan finally opened the door and squeezed in before the semiconductor train was on the highway track.

"Old Su, how much does this CPU frequency work?" Liang Yuan finally raised his head and looked away from the chip in his hand.

Tang Wan has been famous for her high-tech electronic computers for a long time. Listening to the conversation between Liang Yuan and Su Liangyu, the small black film in Liang Yuan’s hand is the brain of the electronic computer. Not only did she become interested, she took the chip from Liang Yuan’s hand and looked carefully.

"Far from 16mhz and 22mhz, it can be said that it has the highest main frequency among the 80286 chips in the world. It is 2mhz higher than the am286pro launched by the American Amd Company in 1984." Su Liangyu said with a smile.

It seems that there is only four years of gap. It seems to be within reach, but Liang Yuan clearly realized that this C2 processor is the ultimate that the chip laboratory can do at present, and the 16-bit processor has entered the phase of elimination. 32-bit is the enduring mainstream in the future processor development history. This newly born C2 processor is destined to be a flash in the pan.

"I heard from Lao Su that we have grasped the tail of the international mainstream level. Are there any shortcomings in the patent?" Liang Yuan asked with a smile.

"Lao Fang's idea is actually similar to Dr. John Hennessy, who founded MiPS Company in Stanford. However, Wang An Computer was not optimistic about the C2 processor at the time. When Lao Fang developed the C2, he had limited resources and had to start with the 16-bit processor. John Hennessy started with 32-bit. In terms of patents, the bus and instruction sets of the two have a lot of similarities. Before Lao Fang left, Wang An Computer was also negotiating patent issues with MiPS."

"What are the differences now?" Liang Yuan asked curiously.

"Mips ask for a price is too high. In terms of patent layout, there are more than 20 basic patents involving Risc architecture and Mips instruction sets. Although some patents of C2 processors may shake the basis of Mips patents, the path to upward development has been blocked by Mips. Faced with this situation, Mips has issued a price of $100,000 per basic patent authorization, and at the same time, all patents of C2 processors belong to Mips Company..." Su Liangyu and Liang Yuan explained in detail the extremely complex patent relationship between C2 and Mips.

A large number of proprietary computer words made Liang Yuan dizzy when he heard it. When Su Liangyu finished speaking with an unsatisfactory statement, Liang Yuan hurriedly asked: "Old Su, how likely would we win a lawsuit with Mips Company? What is the authorization level that Mips Company is talking about?"

"If you go to a lawsuit, you will have half the chance of winning or losing. According to the company's legal counsel in the United States, even if you enter the judicial process, the possibility of reconciliation with MiPS will still be the greatest in the future. However, we will have a high chance of obtaining the right to use miPS technology and part of the patent ownership for free. If you pay the patent payment and get authorization, we will only obtain the permanent free use authorization of miPS, which means we do not have the right to re-authorize the technology in our hands." Su Liangyu explained.

"In fact, the price of MiPS is 2 million, which is also related to the r300 chip it is currently developing. According to my old colleague in the United States, MiPS's financial situation is very tense. If this factor is not available, MiPS's attitude towards patent authorization is still very open." Fang Dongxing added.

"Lao Su, the technology derived from the basic patent should belong to us, right?" Liang Yuan asked.

"Well, we have complete and exclusive patent ownership of all patents derived from twenty basic patents." Su Liangyu nodded and said.

Liang Yuan tapped the surface of the coffee table with his fingers, and his head was running rapidly. On one hand, there was a long lawsuit and ownership of some mips patents, and on the other hand, there was a two million US dollars and the right to use the mips patents. After gritting his teeth, Liang Yuan decided to put the bag in order.

Although part of the ownership of the mips patent looks beautiful, this kind of large patent lawsuit will not have any results after three or five years. In addition, there are too many uncertain factors that are about to change the international situation. Two million is about two million, and the right to use is right. In terms of architecture, based on mips technology, the multi-threading technology of developing processors and high-performance embedded processor power consumption, etc., has a great advantage compared to the dominant arm in the field of smartphones in the future.

Of course, Liang Yuan is willing to spend 2 million US dollars. It is still the condition of using all the patents of mips for free. According to Liang Yuan's estimate of the future, it is not uncommon for the production of risc chips to hundreds of millions of chips in the future. If each chip pays a certain licensing fee for mips, even if the lawsuit is transferred for 10 years, Liang Yuan will continue.

After having all the authorization of mips, the chip laboratory can develop various high-performance chips based on mips patents at will, without having to design chips across industries and deeply change the CPU architecture, and then regain the architecture license from mips.

To put it bluntly, in addition to the American mips company, the chip laboratory is equivalent to a Chinese mips company. Apart from the right to have no external authorization of processor architecture, the remaining rights are no different from American mips, and they are much more free than ordinary processors in-depth authorization.

"King Fang, I read the follow-up research and development report of the chip laboratory a while ago. You have put the C3 processor on the agenda, right?" Liang Yuan asked Fang Dongxing.

"Well.c2 is a mature processor. The laboratory has not invested much energy in the architecture. It just concentrates on improving the process of mass production. The focus of laboratory research and development is to always be on C3. At present, the entire project progress has been completed by fifteen percent." Fang Dongxing briefly said the development progress of the C3 processor.

Liang Yuan estimated the time. According to the current research and development progress of the chip laboratory, it will take at least 20 months to complete the development of the C3 processor from 16-bit to 32-bit. If we include the steps of small trials, pilots, and dies, it will take at least two years to go offline. For Liang Yuan, who is in urgent need of high-performance chips, it is really unbearable.

"Fang Gong, you just mentioned that MiPS is very open-minded in the authorization of processors. In addition, the company is short of funds. Do you think we can buy back the full authorization of MiPS, the first 32-bit Risc processor r200 developed by MiPS in 1985?" Liang Yuan asked in a pondering manner.

"Xiaoyuan is planning to integrate the C3 project and R200 technology together?" Fang Dongxing asked.

Liang Yuan nodded and said, "Fang Gong, after the chip laboratory is successful, there are two main application directions, one is various control chips based on c2, and the other is a supercomputer based on c2 processor."

"For various control chips, performance is not the most important. C2 can handle it completely, but it is different for supercomputers. When using the same number of CPUs, the processing capacity of C3 can be at least twice as high as that of C2."

"I think since the C3 processor and the R200 of MiPS are developed based on a technical system, integrating them together should greatly accelerate the speed of the C3 processor being offline. Of course, these are my wishful thinking now. Whether it can be realized depends on Fang Gong and Lao Su's intentions. After all, you are professionals." Liang Yuan said sincerely.

"When using supercomputers, of course, the 32-bit processor has a huge advantage. Xiaoyuan's idea is good, but it is hard to say whether he can obtain full authorization for the r200 chip. If we put this condition in a package agreement for patent negotiation and then pay a certain price, it is still possible to succeed. Mips does not want major loopholes in its own patent system. It is reasonable to use the previous generation of technology to make appropriate compensation." Fang Dongxing said while thinking.

"Has Xiaoyuan decided to take the supercomputer project?" After Fang Dongxing finished speaking, Su Liangyu asked in surprise.

Liang Yuan nodded and said, "My mother has been talking to the British Rolls-Royce Company for more than a month. I guess the entire technology introduction agreement can be signed before New Year's Day. The gas engine laboratory is in the same yard as you. Lao Su will just grab someone and ask."

"After having a certain foundation, no one can count on subsequent research and development, and it is even more unrealistic to achieve rapid breakthroughs in metallurgy and materials. My mother put her idea on design. Anyway, I heard from her that it takes at least 4 billion supercomputers to run at least 4 billion times per second to solve the new problems encountered in design."

Su Liangyu knew very well about the overall architecture and development focus of Yuanjia, so there was no big surprise that Liang Yuan decided to develop supercomputers. From a domestic perspective, R&D supercomputers have little risk and high profits, which is much more cost-effective than building personal computers. For the Republic in the 1980s, the annual consumption of computers would not exceed 100,000 units.

"4 billion times per second, similar to the Livermoa National Laboratory in Rolands, California, USA, the fastest-running Clay 28 supercomputer equipped in the United States the year before yesterday. Xiaoyuan, your requirements are really not low." Su Liangyu said smacking his lips.

"Old Su, is this thing difficult? It's almost done if you use 20,00030,000 c2 processors to combine them together." Liang Yuan asked in confusion.

"Xiaoyuan, the difficulty of parallel computing of 20,000 chips is not mentioned. The server cabinets used by 20,000 processors can be at least four or five kilometers long. Even if they are installed in a high-density manner, they still need a football field size to be placed."

"No way? Why do I feel like this room can be installed?" Liang Yuan looked disbelief and casually expressed his cognition from his previous life.

In Liang Yuan's memory, the supercomputers in later generations all used multi-core blade servers. Supercomputers, like Tianhe Supercomputers, which are composed of million-level chips, are not as big as a football field.

"How is that possible?" Su Liangyu and Fang Dongxing said with a smile.

"The supercomputer in Xiaoyuan's mouth requires at least 20,000 CPUs. Even if it uses a high-density rack server, a cabinet is calculated at 7u, which requires nearly 3,000 cabinets, and the width of one cabinet is nearly half a meter..." Fang Dongxing calculated a detailed account for Liang Yuan to see how this big place in the football field came from.

After hearing Fang Dongxing's words, Liang Yuan felt intuitively wrong. Even if there are more cores of processors in later generations, one processor with 8 cores will be the best. Supercomputers like Tianhe still have to face 100,000-level processors. If Fang Dongxing's arrangement is performed, they may not be able to let go of a football field.

Something must have been wrong. Liang Yuan bent his finger and tapped his forehead, thinking about Fang Dongxing's words, trying to find the root cause of the problem. (To be continued. Mobile phone users please read it.)
Chapter completed!
Prev Index    Favorite Next