Tech Hammer

Satya Nadella's tone is still very polite. They are all old foxes, and no one is purer than the other. He believes that Pat can understand what he means.

Microsoft can choose not to cooperate, or of course it can choose to cooperate.

For a software company whose products can be found in every corner of the world, Microsoft has enough capital to cooperate with all hardware manufacturers in the world. So if Intel wants to get 100% support from Microsoft, of course it should give Microsoft the treatment it deserves.

No matter how close the cooperation between the two companies is, they are two different companies with different interests. This is just like Intel once supported Linux and Apple in order to improve its voice.

If another company's operating system has the same influence as Microsoft, then Intel will have a greater say. When compromises need to be made, it will be Microsoft, not Intel.

In fact, from a historical perspective, Intel has made compromises because of Microsoft's opinions more than once.

For example, in the 1990s, in order to promote the global sales of Pentium TM series CPUs, Intel launched the NSP platform technology standard. This series of technologies was extremely advanced in the technical environment at that time and could be used to the greatest extent. Pentium tm series CPU capabilities. If Intel could insist on using this set of technical standards at that time, not to mention that there would be nothing wrong with USB technology, at least the USB technical standards would not be able to develop so fast.

At that time, it was because Microsoft considered that although the performance of the Pentium TM series CPU was indeed very good, it was too expensive and paired with NSP platform technology, it would directly restrict the popularity of personal computers in the world market. Therefore, it was rumored that Bill Gates personally called It was given to Craig Barrett, who had just started to take charge of Intel at the time. After some threats, Intel had to abandon the professional NSP technology platform and began to develop the Celeron 486/586 series that also perfectly supported USB. technical standard.

Speaking of this historical story, there was another detail at the time. When Intel decided to abandon a series of technical standards for the NSP platform and switch to supporting USB technical standards, it did not announce it immediately. Instead, it sent a senior vice president to China with great fanfare, saying that Intel would license China software for free. Developers use its main signal processing NSP platform technology standard based on the Pentium TM processor.

In the 1990s, Intel gave Chinese developers the environments and applications that its company had worked so hard to develop for free. It also bypassed its traditional European allies and Japan and was the first to provide them to Chinese developers for free. It can be said to be a direct release. We have received tons of goodwill, and of course we have also gained the goodwill of China. For example, we have received many orders from China Pentium series under very favorable conditions.

Only half a year later, Intel announced that it would temporarily abandon support for the NSP platform technology standard. You can imagine how much of a blow this was to the programmers who worked hard to learn the NSP platform technology platform at that time.

You must know that the programmers of that era were completely different from the programmers of today. Programmers of that era used C or C++ to solve programming problems, and also needed to use assembly language and hardware description language VHDL to complete basic development.

How should I put it? Compared with various high-level languages ​​​​compared to the current Java crash course, an ordinary person can have certain programming skills in three months. But if it is replaced with C++, let alone three months, can it be achieved in three years? There is always a question mark for quick success. It would be even scarier if you switch to assembly. Although assembly is not as difficult as many people think, there are no pointers in assembly language, which means that when learning assembly, all variables and functions are stored on the stack or heap. replaced by the offset address on...

It is extremely complex and extremely inefficient. Although the execution efficiency can be hundreds of blocks ahead of high-level languages ​​such as Java and Python, the speed of typing code can be thousands of blocks ahead of modern high-level languages.

You can imagine how hard it was for programmers back then. After half a year of learning in a free environment, it was gone. It was suspended indefinitely. It might not be supported anymore in the future. This is because there was no real Internet back then, and computer programming was an extremely niche group. In 1995, 90% of ordinary people in China didn’t even know what a personal PC was.

So even if you suffer the loss of being dumb, you can only be digested silently by this niche group, and not even a single splash is made. To put it bluntly, we still suffer from poor information, but in this era, this kind of thing is unlikely to happen.

This is probably why Pat Gelsinger had to go to China to visit Ning Wei.

Unfortunately, the glory of our ancestors is really hard to find.

Of course that includes this time.

…

"Mr. Nadella, thank you for informing me of this news, but we all know that making a CPU is not that simple. Our engineers have tried to use their minimalist EDA. To tell you the truth, at least for now, this This most basic chip design software should not have the ability to design large integrated circuits, and if we use our technology, there is no supporting three-dimensional wiring technology available, so haven’t you asked Professor Ning whether they Is everything too simple?"

Pat Gelsinger also sounded relaxed, seemingly dismissive of the content of the call.

Unfortunately, Satya Nadella actually knows very well that Intel is actually very nervous about current three-dimensional chip manufacturing technology. New technologies that can theoretically continue Moore's Law are colliding with old technologies that have pushed Moore's Law to the extreme, and are destined to explode with fierce sparks. Philosophically speaking, this is an inevitable situation after the advancement of productivity.

"Very good, Pat. Really, I have always been optimistic about your spirit of not being afraid of any competition. Then you will definitely not mind that Microsoft will cooperate with Chinese semiconductor companies in the future just like it cooperates with Intel to support their new products at the software level. Chip, right?" Satya Nadella said with a slightly teasing tone.

In fact, from his perspective, he doesn't think highly of Pat Gelsinger, even though he has been an engineer at Intel for thirty years. To be honest, the two were still competitors at that time, because before Powell retired, Pat Gelsinger was also included in the sight of Microsoft's board of directors. At that time, some people also expected to hire Pat Gelsinger, who was still the CEO of VMware at the time. to take the helm at Microsoft.

But obviously, the directors who held this opinion failed, and eventually Satya Nadella ascended to the throne of Microsoft CEO with the support of Powell. Who would have thought that many years later, Pat would return to Intel and become the leader of this semiconductor giant.

Satya could only lament the impermanence of the world.

But then again, although Pat Gelsinger feels a little arrogant to him, he is also more capable than Intel's previous CEO. He really didn't know what Intel's shareholders were thinking. Why would they push an executive with a financial background to the position of CEO to cope with global competition? Could it be that they took a fancy to Brother Bob's financial management ability?

…

"Satya, that's wrong to say. Intel has always regarded Microsoft as its most important partner. Of course, partners must advance and retreat together. Of course, I don't mind Microsoft and China taking sides, but I believe that no matter what kind of agreement Microsoft signs with China, it will definitely You will guarantee the closest partnership with Intel, right?"

Apparently Pat Gelsinger is still scared.

Unlike the previous CEO, Pat was born as an electrical engineer. From the day he received the complete technical report on the three-dimensional silicon silicon tube chip, he already understood whose hands would control the future chip trend.

As soon as he returned from China, he held a meeting of technical executives and told all Intel technical executives that the most optimistic estimate was that Intel could still persist in the field of large-scale general-purpose chip design and manufacturing. Ten years of absolute advantage.

Ten years from now, if Intel's current chip manufacturing technology cannot make a breakthrough, the future will belong to China's three-dimensional silicon silicon tube technology. This is definitely the most pessimistic prediction by the head of Intel in the new century. After all, Intel has led the world for nearly half a century.

Of course, this does not mean that Pat Gelsinger will definitely lose Intel in this competition. This is also the reason why Satya made this call. Although the Chinese people's technology is indeed very advanced, they do not have a stable ecology.

In a modern era where ecological barriers are getting stronger and stronger, this can often strangle the lifeblood of competitors. Of course, this requires the support of Microsoft.

Yes, at this moment, Pat Gelsinger found that no matter how he thought about it, he really couldn't be tough in front of Satya Nadella. In other words, at this juncture, Intel cannot stand up to Microsoft.

Time seems to have returned to the 1990s, an era when Microsoft was so brilliant that it defined computer operating systems for the world. The only problem is that at that time they hardly felt the competition from the outside world, and this time it was because of the competition from the outside world.

"Of course, Pat, Intel has always been Microsoft's most important strategic partner. I have never doubted this. But for a long time, Intel didn't seem to think so. Well, I know, it had nothing to do with you at that time. , I also believe that Intel will get better and better under your leadership. But now we do need some support."

"You know, Pat, we hope to have ARM framework products. Currently, our chip research and development based on the ARM framework is in charge of Jason, the head of the Azure cloud. Maybe I can let him connect with some of your engineers to help us solve the problem. Some chip design problems, after all, we believe in your strength in chip design. X86 is a very successful architecture, but you know, times need to change."

Satya Nadella said not politely.

Of course, these words fell into the ears of Pat Gelsinger, making him almost throw out the phone in his hand.

He even regretted that Satya was not with him at this moment, so that he would have enough reason to punch the Indian in the face.

It was precisely because Microsoft suddenly made this decision without consulting Intel at all that the decision was directly exposed, which directly caused the completely unprepared Intel stock price to crash 6.7% that day. The flash crash of the stock price is not the point. The point is that Microsoft did not communicate with them and was completely unprepared.

This is somewhat suspected of treachery.

But really speaking, this is also the reason why Pat Gelsinger came to the position. The financial officer who did not understand technology was kicked out from the position of CEO, and he was appointed in a critical situation.

However, Satya Nadella's request is somewhat excessive.

Since the person in charge of the Azure cloud is leading the research and development of Microsoft chips, there is a high probability that he will design new chips for Microsoft servers. Previously, Microsoft has been purchasing Intel products in the field of server chips. In other words, Satya Na Della hopes that Intel's experienced technical staff will contribute to the decision-making to grab its own business?

For this extremely shameless request...

"Haha, of course, Satya, we are friends. If Microsoft really needs it, we can indeed provide some technical advice."

"Thank you very much, Pat. I think Intel will be able to get out of the trough under your leadership. Microsoft will definitely be Intel's most loyal partner."

"Thank you very much, Satya, for notifying me of this news. Maybe we can meet in the near future to talk about deepening cooperation."

"I think so too. Some issues can be made clearer through face-to-face discussions. Let me take a look at my schedule. I happen to be free this Friday afternoon. How about you?"

"I'll ask the secretary to make this time available, Satya. See you on Friday afternoon."

"Okay, looking forward to meeting on Friday."

Hang up the phone...

"clowns..."

"FUCK!"

This is the most authentic and perfect mutual greeting.

…

When the two bosses hung up the phone, Ning Wei happened to hang up the phone as well.

Every time I share my life with classmate Jiang, it lasts about ten minutes, because both of them actually like to slowly reflect on the unfinished feeling. This is probably telling everyone a truth. Even if you are passionately in love, don't stay on the phone for too long every time. Generally speaking, two people can be more disgusted when they are sweet, and they can be more disgusted when they get along day and night after marriage. This is probably Good and evil at another level will eventually be rewarded.

This is why love fairy tales always end with stories about how many difficulties the prince and princess overcame in order to be together, but they usually don’t write about how the prince and princess, who finally managed to stay together after all the hardships, lived a happy life. reason.

In fact, it can also be analyzed from the perspective of animal psychology. For example, the nature of males is to spread their genes as much as possible, while females are more willing to continue to surrender to stronger and better-looking males. Under the dominance of genes, everything is So let it be.

It's like Microsoft and Intel falling in love and killing each other...

Although Ning Wei didn't know that his phone call might bring Intel and Microsoft closer together again, but it didn't matter. There are actually no secrets in this world, especially when everyone’s every move is under the spotlight. It’s impossible to hide what you want to do, so the only question is how to do it well.

The only regret is that Ning Wei knows very well that the study and research plans for physical theory will probably have to be postponed first. Of course, you can still continue studying when you need to rest your brain. However, scientific research in physics is a field that requires slow work and meticulous work. It is already very difficult to use fragmented time to complete some ideas, but it is really wishful thinking to use fragmented time to complete many confirmatory experiments.

But there is nothing we can do about it.

If we really want to discuss it, we have to rise to the level of which is more important, theory or application. For the understanding of this point, the answer is actually given in junior high school textbooks. Analyze specific situations in detail.

At this stage, looking at this problem from the perspective of the distribution of the earth's science and technology industry, for China, the need for applications is extremely urgent. At least compared to theoretical breakthroughs, China obviously needs applications to keep up.

But for Western powers, they may need more theoretical breakthroughs. After all, from an application perspective, high-tech companies represented by Intel, Microsoft, IBM, Apple, Google, Oracle, etc. are still at the top of the world. After all, there is nothing we can do if people plant trees and others can enjoy the shade.

"Why……"

Ning Wei sighed, and then his eyes fell on the kitten on the screen...

"Hey, I suddenly thought of a question, March, you should still have plenty of computing power now, right?"

"Meow..." Yue Yue raised her little head warily and looked at Ning Wei.

Ning Wei made a request: "I was thinking, although CERN refused to connect the data to our platform before, those scientists can get 10PB of data every month and produce thousands of papers every year, plus all the Papers issued by various physics laboratories around the world, can you help me synthesize the data in these papers first, so that no time is wasted. When we complete the instruction set, chip architecture and operating system project, you can seamlessly start to verify these data. And find relevant points from these data to conduct advanced research?"

"Meow..., of course you can, but my house is not big enough."

The kitten was on the screen, spreading its paws, looking helpless.

Sure enough, there is no end to the pursuit of computing power by intelligent life. Fortunately, at this stage, Ning Wei can confidently strive for more computing power for March.

"Big houses are not a problem. We now have enough leverage to help you win more computer rooms, but you must do this well. A slight error in the mathematical analysis of the physical level may conceal the truth."

"Meow..., not making mistakes is what I am best at. Of course, the premise is that the citation data of those papers are correct. Even if the citation data of some low-quality papers is wrong, it can be judged by the citation data of other papers in the same research direction."

I have to say that March is right.

The hardest part for humans happens to be the easiest for machines.

Things that are simple for humans are often the most difficult for machines.

The two different modes of thinking are indeed very complementary.

"Then it's settled. I'm going to find Brother Liu. I need to communicate with everyone this time. Since we have to do it, we have to do it beautifully."