my return to life

In the near future, there is a plan to cooperate with Bifang to upgrade to a cloud computing platform to save costs.

Simulation works almost throughout the design.

After all, it is not possible to do a bit of design and go to tape-out without confirming anything. It is always necessary to carry out various verifications and tape-out after confirming that there is basically no problem.

The investment in simulation can be called massive, and tape-out is not cheap.

Anyway, these are the reasons why it is so expensive to be better in the field of chip design.

Chip design is not to use a computer to draw two pictures, but involves a massive project. Even if the ARM public architecture is used, there are many other technical tasks.

It's just that from a more fair point of view, the technical content has a priority...

The design progress of Shenlong 2 is quite good.

Fang Nian didn't talk too much nonsense, and said with a smile: "Academician, there is some good news, tell me quickly."

After all, Bai Ze Laboratory now only has two major tasks: the research of Shenlong 2 and the confidential laboratory.

So Fang Nian is quite looking forward to it.

Sure enough, Chen Jianye said: "The confidentiality laboratory has made some breakthroughs, but..."

Having said that, Chen Jianye hesitated.

"It can't be regarded as a breakthrough in the independent design of mobile phone processing chips. To be precise, it is the instruction set architecture that has been preliminarily completed before... There is a problem."

Chen Jianye was a little embarrassed and said: "We have carried out a lot of verification and revision work, and even completed the preliminary design of the micro-architecture based on this instruction set architecture, and even a certain amount of simulation work;

Turns out...it's really not suitable for small mobile devices,"

"Especially the cooperation with Godson started last month, and after some rationalization and adjustments, it has no chance to be used with mobile phones, it can only be used for computer CPU..."

Fang Nian: "!!!"

-

①: The difference between instruction set architecture and microarchitecture has been explained before, the simplest is that the former is a specification, and the latter is a method;

If the former changes, the whole system will change, which is why Wintel is so invincible;

The latter change mainly affects IPC (referred to as architecture performance);

Another formula: CPU performance = clock frequency (MHz or GHz) × IPC (how many instructions are executed in each clock cycle of the CPU);

The clock frequency mainly depends on the production process, that is, how many nanometers; therefore, whether the IPC performance is excellent or not becomes the difference in the performance of CPUs from different manufacturers with the same process. It is the CPU core, that is, the microarchitecture, that affects the performance of the IPC.

======

PS: The extra upgrades for ascension to stars mentioned earlier have all been completed, and next... anyway, we still need to ensure two updates, and we can also free up our hands to add updates. Bosses, come here~

Chapter 733 Mr. Guan, your vacation is ruined, and my mother quit!

"..."

"..."

After Chen Jianye reported one by one, he heard silence on the other end of the receiver, and hurriedly continued.

"Mr. Fang, the result of such a major mistake is mainly my responsibility."

"The project of the confidential laboratory has not been disclosed to the public, but some research results have been collected to me, and then some technical results that can be used internally are released. After communicating with Godson, it has become the current situation."

Speaking of this, Chen Jianye sighed: "I don't want to shirk responsibility. There is a fundamental mistake in the technical direction. It just happens that Godson cooperates, so I boldly let it continue."

"This unnamed instruction set architecture is 100% self-developed in accordance with the reduced instruction set standard. Because of Godson, a certain compatibility design has also been carried out, and the simulation test results are also very good."

Hearing what Chen Jianye said, Fang Nian suddenly interjected: "That is to say, the completion of this instruction set architecture is already very high?"

"Essentially only the actual tapeout verification is missing?"

Before Chen Jianye could speak, Fang Nian quickly asked a few more questions: "How about the various performance indicators of the simulation test? What about the R&D investment?"

Finally came to his senses, and said with a peaceful smile: "Academician, don't rush to say mistakes, and don't rush to take responsibility. At least what I've heard now is already great news!"

Hearing Fang Nian's laughing voice and excited tone, Chen Jianye almost didn't react.

But after all, he is an outstanding academician in the prime of life, with a clear mind and quick response.

"This set of instruction set architecture has been determined to be used only for computer CPU design, and not for mobile phone CPUs. Basically, it can be said that the research and development investment of the confidential laboratory is on it. Roughly, the research and development investment is 11.2 billion."

"Of course, the entire project's research and development process will also save the research and development investment applicable to mobile phone CPUs. After all, they are all streamlined instruction set systems."

Chen Jianye first answered the funding question that most bosses are most concerned about, and then introduced.

"The pre-silicon completion rate can be regarded as 95%, because of the technology sharing of Godson, the follow-up basically only needs tape-out verification;

However, the overall CPU design currently completed is based on SMIC's 65nm process after sharing some technologies with Godson. "

After a short pause, Chen Jianye introduced the simulation test performance one by one.

"Mr. Fang, let me first introduce the content of this brand new instruction set architecture:

Contains 233 basic instructions, 20 virtual machine extensions, 156 binary translation extensions, ... a total of 1991 native instructions;

The instruction format has been redesigned, and there are 6 available formats;

……

It is mainly developed based on Nuwa DeskOS, which can run native programs based on this instruction set architecture, and can also be compatible with DeskOS programs of other instruction set architectures through translation, including MIPS, X86, and arm;

Because of Godson's technology sharing, the translation overhead for the MIPS instruction set is extremely small;

The MIPS target for translation execution is 100% native performance, the target for translation execution arm is 95% native performance, and the target for translation execution X86 is 60% native performance;

Can support Windows system, but the overall translation execution target for Windows is only 30% of the native performance;

Because MIPS can be 100% translated and executed, it is fully compatible with the MIPS ecosystem. "

"The simulation performance of this CPU is approximately equal to the Pentium E07 released by Intel in 2140;

Main frequency 1.6Ghz, dual core, bus frequency 800Mhz, share 1MB L200 cache, FSB 8MHz, multiplier 65, design power consumption [-]W;

The interface design will support the chipsets of mainstream motherboards currently on the market;

For now it is only available for desktop use. "

Listening to Chen Jianye's detailed introduction on the phone, Fang Nian frowned slightly: "Isn't that some kind of exaggeration?"

Fang Nian even wanted to talk about Da Yuejin.

Then he explained: "Although I have no impression of the Pentium model you mentioned, but before Intel launched the Core series in the past two years, Pentium was its high-end series, and the U in 07 had such performance parameters..."

"Godson's previous design and development capabilities were not strong."

Hearing what Fang Nian said, Chen Jianye scratched his head a little bit, unable to laugh or cry: "Mr. Fang, you may not know that I am a relatively cautious person, and usually only talk about 80% of the actual simulation performance."

"Godson's design and development capabilities are not strong because of lack of money;

The design of an ordinary-grade CPU for a computer that is several years behind, the laboratory still invested more than 11 billion, not counting the massive technology accumulation before;

If you don't even have this performance, then I think it really failed your expectations of Bai Ze Lab. "

After listening to Chen Jianye's explanation, Fang Nian breathed a sigh of relief: "The more than 11 billion yuan is very worthwhile."

"It's at least a few billion cheaper than Shenlong 1, and it won't matter if it's a few years behind."

After a short pause, Fang Nian pondered and said: "The research and development of the instruction set architecture has achieved such a great result, and the work of the security laboratory may be adjusted to a certain extent;

Don’t rush to stream the film first, I hope to go through the solution of integrated nuclear display, and try to see if it can be done in one step. In this regard, the company will promote Jingjiawei of Xiangchu to join a large-scale technical alliance. "

"Secondly, I hope academicians can help in the technical field. I want to promote the establishment of an independent instruction set architecture alliance. The members of the alliance must not only be Godson;

Of course, the first point is that there should be no lawsuits in the command system within the alliance. "

"This is the rough content. I will let the frontier list the specific plan as soon as possible."

Then, Fang Nian said: "However, academicians still have to work hard on the originally planned development direction. Shenlong 2 has completed the basic design verification, and our mobile phone CPU independent instruction set architecture has not yet appeared."

"Okay, okay." Chen Jianye agreed with a smile, "I support what Mr. Fang said unconditionally."

"I will urge the confidentiality laboratory to be subdivided into more project teams to complete the company's strategic development requirements."

Fang Nian smiled slightly: "Academicians have worked hard. Recently, someone from the frontier will go to Luzhou to reward the outstanding employees of the laboratory."

"Then thank Mr. Fang for my colleagues first, haha~" Chen Jianye said heartily.

"..."

…………

After finishing the call, Fang Nian calmed down and fell into thinking.

The confidential laboratory inside the Baize Laboratory worked out its own computer CPU design by mistake, and basically passed the simulation verification.

This is an unfavorable behavior for the planning of Bai Ze Laboratory.

But it is not bad news for Frontier, or for Nuwa Lab.

Because the reason why Fang Nian took the initiative to spend money and promote Godson to form a technology alliance was mainly for the development of Nuwa DeskOS.

To a certain extent, a DeskSon alliance similar to wintel has been established.

From the root cause, it affects some major foreign hardware manufacturers, such as NVIDIA and other underlying driver support for DeskOS.

After all, Fang Nian knows very well that in the field of GPUs, only Jingjiawei is affordable in China, not the current Jingjiawei, at least Jingjiawei from 14 or 15 years ago.

However, seeing that Jingjiawei is a Xiangchu enterprise, Fang Nian is still very happy to join forces with it now.

Thinking about all this mess, Fang Nian suddenly muttered: "Ecology..."

Then he looked at Lu Weiyu and Guan Qiuhe, who were almost staring at him with longing eyes: "Bai Ze basically finished the design of Shenlong 2, and by accident, they came up with the instruction set architecture and microarchitecture of the computer CPU. It's a good thing, but..."

As he said that, Fang Nian spread his hands together: "Some of us are busy, especially Mr. Guan. I'm afraid your vacation will be wasted."

Fang Nian obviously said the last sentence on purpose.

Sure enough, Guan Qiuhe's eyes widened twice: "Why... it's a mess?"

"Isn't it good, what does it have to do with me?"

Fang Nian said seriously: "Because I saw the possibility of forming the 'ecology' that Frontier wanted from the very beginning."

"So, you have to get busy."

The boss Guan Qiuhe was not happy: "I don't understand ecology or something, anyway, I don't think it has anything to do with me."

"I'm going on vacation!"

"My old lady quit!"

Guan Qiuhe was about to jump up.

Lu Weiyu next to him thought about it and said, "How about...Mr. Fang, do you think we can help, Mr. Guan has been crazy busy these past few months."

"No, you can't help, only the head office." Fang Nian said calmly.

Guan Qiuhe: "..."

Seeing that there is no way to escape, Guan Qiuhe pursed her lips and said softly, "Then... that's fine, I'll just do it."

Seeing this, Fang Nian finally couldn't help but burst out laughing: "Mr. Guan, indeed only you can do it, but you'll be going in September at the earliest, so...it won't affect your vacation."

"Don't be like an angry little daughter-in-law."

Hearing this, Guan Qiuhe stared straight at Fang Nian, her chest fluctuated, her face showed fierceness, as if she was about to bark her teeth and claws to eat Fang Nian raw.

"You wait for my old lady! I'll kill you sooner or later!"

Fang Nian was not afraid at all, and shook his hands: "Okay, I won't quarrel with you anymore, it's time for the holidays, everyone is here."

"..."

…………

…………

After all, today is the weekend and Qixi Festival, so there is still time to do things after the festival.

It is not an urgent project in itself, even if it has been taped out successfully, it will be difficult to commercialize it.

Several cars drove into the yard in a row.

Wu Fucheng and Bai Congee are usually relatively frugal and simple, and they shared a Ferrari—Bai Congee still didn't bring his partner, similar to Wu Fucheng, and he didn't bring it out unless he finally determined the relationship.