1991 Starting from the core

Chapter 515 The CPU needs to be accelerated

After chatting, the two quickly got down to business.

Li Mingliu took out the information prepared for tomorrow's meeting and handed it to Su Yuanshan. He knew that although Su Yuanshan claimed to be "not as good as himself", Li Mingliu had absolute self-knowledge - just as he had said before by "flattering", geniuses like Su Yuanshan are now in the flash memory The field is not as good as one's own, it's just that the ambition is not there.

You know, the current direction of the flash memory department was almost all formulated by Su Yuanshan! And it turns out that this is the right direction.

As we all know, any direction Su Yuanshan sets is based on technology...

Therefore, to say that Su Yuanshan does not understand flash memory means that he really does not understand Su Yuanshan.

"Our main difficulty now is that our theory is enough." Li Mingliu pursed his lips. Once he talked about specific technical issues, he became serious: "Storing three information states in a cell - we temporarily put This kind of particle is called TLC, Triple Level Cell according to the naming convention - in theory, it is feasible. But in fact, there are too many difficulties to be solved, mainly the problems of electron escape and voltage state."

"According to theory, we need to detect 8 voltage states... This requires the simultaneous improvement of control chips and wafer fab processes, and as the number and density of transistors increase, the problem of electron escape becomes more and more serious. Our start this year Yes, we mainly discuss how to solve electron escape."

Su Yuanshan listened to Li Mingliu's introduction over and over while looking at the technical documents.

Frankly speaking, if he hadn't been paying attention to Li Mingliu's team, he might have had to work hard to understand it.

"Well, I still say what I said, in the field of flash memory, you are professionals now."

"Even an ordinary person can become an expert in this field after 10,000 hours of in-depth training - this is the 10,000-hour rule."

Li Mingliu was slightly startled: "Um? Who proposed it?"

Su Yuanshan was also startled, and he immediately thought that this statement would have to wait at least ten years before it was proposed...

"Just take what I say. Even ordinary people can become experts in 10,000 hours, let alone your team... You work more than twelve hours a day, do you really think I don't know?"

"I don't have any other hobbies." Li Mingliu laughed: "The stupid bird should fly first."

"Tsk, since when did you become so humble?"

"It's just in front of you." Li Mingliu said calmly.

"..." Su Yuanshan couldn't help but dumbfounded: "Okay, anyway, the flash memory industry mainly depends on you. I'll go back and think about it to see if I can think of something evil..."

Li Mingliu: "..."

The next day, Su Yuanshan really couldn't put forward any effective suggestions. On the contrary, Li Mingliu's brainstorming gave Su Yuanshan a lot of inspiration.

For example, some people suggested that since the problems of integration and electron escape cannot be well solved at the current stage under the existing conditions, then we should simply start with the control chip and use algorithms to program and optimize each cell (unit) to achieve better results. Minimize data interference caused by electron escape.

Others have also suggested that regardless of the process and integration level, we can first develop the TCL architecture and then talk about it - because now everything is based on calculation and simulation.

And everyone knows that with the higher the level and the more advanced the process, the more metaphysics will appear.

That’s why there’s that sentence: “On paper, it’s easy to understand, but when you realize it, it’s going to be released.”

Sometimes, no matter how perfect your design is, no matter how smooth the simulation is, and you are a top-notch movie, you may find some faults.

After the meeting, Li Mingliu found Su Yuanshan and asked him what to do...

Only then did Su Yuanshan finally understand Li Mingliu's true intention of "inviting" him to this meeting.

It seems that Li Mingliu still prefers to figure out the architecture first and try the tape-out first.

In response to this... Su Yuanshan said that he should wait a little for a while.

"This guy Lao Li... ah, is a bit too much. Senior brother, please don't imitate him."

Su Yuanshan said with a smile.

After attending the meeting of Li Mingliu's team, Su Yuanshan came to Gao Xiaodi, and the two inevitably talked about the content of Su Yuanshan's previous meeting.

"What's wrong with Old Li?" Gao Xiaodi was Su Yuanshan's senior brother at the same school, so the meeting between the two was quite casual. He sat on his desk with his butt tilted and looked at Su Yuanshan in surprise.

He has worked with Li Mingliu for a long time, so he naturally knows the relationship between Li Mingliu and Su Yuanshan - they are closer than he is with him! Moreover, Li Mingliu didn't accept anyone but Mr. Xi and Su Yuanshan.

Of course, although Li Mingliu is now his boss in name, they actually belong to different departments and are both leaders. In addition, this is a scientific research center, so everyone usually has no conflicts, but instead talks about them. Because we are both in the Shanghai stock market, our relationship is relatively good.

"He wanted me to support his tapeout of the TLC architecture."

Gao Xiao was startled for the first time, then immediately burst out laughing: "It's too much, it's really too much! Are you planning to withdraw from the film without even mentioning your character?"

"You said... yes, it's true, but the time is really not ripe. Let's save some money and time." Su Yuanshan said seriously: "Brother, where are you?"

On Gao Xiaodi's side, although Su Yuanshan said that he was interested in cultivating the independent management capabilities of the CPU department, it did not mean that he really did not pay attention at all.

At the very least, in his opinion, Gao Xiaodi's performance in the CPU department has been unqualified since the release of the CPU last year.

Therefore, this time he came to the flash memory conference at Li Mingliu's invitation, and at the same time he wanted to sound a wake-up call to the CPU department.

"Our side is pretty good, right?" Gao Xiaodi's eyes wandered, and he didn't know what Su Yuanshan was referring to.

If we talk about architecture, the architecture is currently developing and adjusting the micro-architecture step by step. If nothing unexpected happens, a new micro-architecture will be launched during the National Day this year, as well as a higher frequency.

Next year will be a year of process improvement, but also a year of architectural innovation. The CPU department already has enough plans to meet the overall challenges next year.

"Um...I mean, where is the performance report?"

Su Yuanshan looked at Gao Xiaodi.

Hearing Su Yuanshan talk about the performance report, Gao Xiaodi immediately felt a little embarrassed.

Since the CPU was released last year, Su Yuanshan had a long conversation with him. And actively supported him in establishing a sales network...

It is clear that this is to allow the CPU department to become an independent enterprise in the future.

But now several months have passed...except for the many orders received at the time of release, the performance has not improved much.

After all, Gao Xiaodi has always been engaged in technology.

Su Yuanshan sighed softly: "Brother, we need to speed up the pace of the market."

Tap the screen to use advanced tools Tip: You can use left and right keyboard keys to browse between chapters.

You'll Also Like