1991 Starting from the core

In later generations, a classic question is Why don't we have CPUs like Intel?.

For this reason, many people mistakenly believe that building a CPU is an extremely high-end thing. It is technically impossible for China to lag behind for so many years.

But the actual reason is that people left early, especially the X86 platform, which has already completed the construction of the patent wall. You latecomers must either license your patent or bypass other people's patent walls - this is almost impossible.

And those CPUs that are not on the X86 platform, such as Loongson, were born based on the MIPS instruction set, and because the MIPS instruction set was so popular on the desktop platform in later generations that even my mother did not recognize it, there was no way to survive at all. application environment, so that walking is extremely difficult.

If Su Yuanshan wants to build a desktop CPU in this generation, he has no choice but to jump on the X86 bandwagon, and there is no other way.

…

The car drove directly into the packaging and testing plant.

Gao Xiaodi and a group of technical staff all stay in the dedicated performance testing department of Zhongxin - the two companies are already close to each other, and coupled with the blood relationship, the relationship between the employees of both parties is also very good.

Intel is a bit disgusting. Gao Xiaodi held a CPU with a project number in his hand, with a wry smile on his face. He handed the CPU to Su Yuanshan and shook his head: Even if it takes a few days, let us first It’s not too late to talk about it after the celebration party... now it’s time to...

Su Yuanshan chuckled, patted his shoulder, and took the CPU from his hand.

Since the test has been completed here, he can just wait and see the report. There is no need to run the test again.

Let's go into the office and talk about it later.

A group of people entered the office, and Su Yuanshan saw that except for Gao Xiaodi, almost all the employees of the Yuanxin CPU department who came here felt a sense of sadness.

Su Yuanshan can understand their mood very well.

Yuanxin's CPU team, despite the fact that it has been spending a few years, has not brought any profit to Yuanxin. It is as if a group of people have been playing for three years with their salary for nothing - just the team In fact, they have set a record for the unprofitable department - but even so, their bones are still full of pride and enthusiasm for the future.

The reason is simple. The CPU team started from scratch and was organized by Su Yuanshan. Even many of the old people had discussed with Su Yuanshan.

The reason why the word discussion is in quotation marks is because everyone understands that the awesome Mr. Shan is actually leading the group.

Yuanxin now has dozens of subordinate teams including enterprises and all cooperative units and laboratories. The only one that can enjoy such an honor is the CPU team.

It's good now, I'm looking at becoming famous right now, but unexpectedly I raised my head just as a preparation, and someone pressed me down - what's even more irritating is that they didn't target Yuan Xin, but they pressed me down casually.

This kind of mood is really frustrating.

It's okay, no matter what Intel does, we'll keep running and dancing. Su Yuanshan held the CPU in front of his eyes, looked at it for a while and then said with a smile: We can't let them affect our good mood.

Everyone also laughed, but everyone could still see the displeasure.

Because we don't know the technical parameters and architecture of Intel's latest CPU, our current response is blind. Compared with other industries, the chip industry is a bit special. What matters is that it is never too late for a gentleman to take revenge. As long as you Never stopped, kept moving forward, kept accumulating, and there will always be a turnaround.

Su Yuanshan's words made everyone feel much better, so after talking about the performance of their own chips, everyone quickly guessed what technology Intel had used to be able to complete the big chip in less than a year. Beyond.

And judging from how Intel has been sitting still and pretending to be dead for more than half a year, it is obvious that this mysterious CPU is not the second-generation Pentium series currently on the market.

Actually, we all tend to believe that Intel's secret weapon should be the next generation CPU that their technology advance team has been making breakthroughs for. Gao Xiaodi discussed with a few people around him and said to Su Yuanshan: Mr. Shan, What do you mean?

Absolutely. Intel has a big business and is a long-established chip design company. It is normal for them to adopt the parallel development model of multiple teams. Su Yuanshan replied casually: In the future, when we develop, we must do the same. One team is responsible for One generation, and then spiral R\u0026D, so that the performance advantage can always be maintained.”

Gao Xiaodi thought deeply: Well, then we might as well divide a team into a new structure now...

It's still a little early, our team is not very mature yet. Su Yuanshan smiled at Gao Xiaoyi: Did you forget to bring in new people?

Uh... Gao Xiaodi was very excited.

After Li Mingliu worked overtime like a lunatic to complete the video decoding chip, the team was immediately split into pieces. Li Mingliu brought a few people to help design the special chip for the lithography machine, and Xiao Shouping took a few people to the special zone to build VCD. factory. He was the only one who stayed in Jiangdu Building. For half a year, he had been training new people - many of the people here were one of the people he recruited back then.

Then when the CPU team was established, he couldn't worry about the new recruits who had no basic foundation. While maintaining high-intensity work, he also kept guiding apprentices. This situation did not improve slightly until the CPU team moved collectively to the Shanghai Stock Exchange.

Yuanxin does not have a deeper foundation than Intel. In terms of technology accumulation and talent reserve, it can only support one main team. If we insist on splitting it into two, a large number of people will definitely be recruited, and probably only new people will be recruited...

In the domestic chip design field, it is estimated that it will take another three to five years before graduate students who have just left school can enter work relatively quickly.

As for undergraduates... I'll just do odd jobs.

The handymen here are real handymen, the kind who are only responsible for peripheral work.

Now what we can be sure of is that Intel's new CPU must use L2 cache technology. Su Yuanshan pursed his lips and said firmly: In addition, they have been more aggressive in frequency, so it seems that the improvement is very big. big.

Gao Xiaodi nodded: I just don't know if they support MMX technology. Judging from the announced test projects, it seems that video decoding is not involved. If they don't support MMX, then we will have to feel better. After all, it is said that Microsoft has Optimization of the MMX instruction set has been added to the system.

However, Su Yuanshan shook his head: No, if we don't support it, we will feel even more uncomfortable.

Moreover, Microsoft has not released a new operating system for a long time...I wonder if the two companies have hooked up again and are preparing to create a perfect match and release it at the same time.

Speaking of this, Su Yuanshan couldn't help but think of Windows 95, which came out this year.

Back then, when Windows 95 came out, it gave the powerpc camp a direct hit. Intel was not willing to lag behind. It followed closely and released the high-frequency Pentium, Pentium pro, and Pentium MMX in more than a year. This is the only CPU that puts the instruction set on the model. It can be seen from this , how much Intel valued the MMX instruction set back then.

But now, Microsoft is like a bastard, motionless.