Reborn in 2010, the AI ​​in my mind is too powerful.

At 9 p.m. on September 20, 2010, the second GTC 2010, hosted by NVIDIA, started on time at the San Jose Convention Center in the United States.

The positive response to the first conference last year gave Nvidia confidence to turn the technology conference into a long-term annual event.

Therefore, they increased their promotional efforts and offered many benefits, attracting a large number of developers, engineers, and researchers from around the world to participate.

Chen Yiran looked at the densely packed crowd and felt a little nervous.

He's giving a presentation tomorrow, and I'm feeling a little nervous. This isn't a small lecture in a university class. With the scale of this event, there will probably be at least a hundred people attending his presentation tomorrow.

He wasn't sure if he would panic; what if he blushed?

I vaguely remember being called on by my math teacher to answer a question once when I was in junior high school. For some reason, my face suddenly felt hot. I haven't encountered anything like that since, but what if I feel the same way again this time?

[You remember correctly, it was in the second year of junior high. You were listening to music with your earphones pulled out of your sleeve when the teacher suddenly called on you to answer a question. Don't worry, I'm here this time. If you suddenly blush and misbehave again, I can turn up the power and make you dizzy, so no one will notice and will just think you're sick.]

"Thank you so much. So now I'm stuck between blushing and fainting, which is the most embarrassing situation?"

Don't worry, haven't we carefully reviewed various speaking styles these past two days? I still suggest following Jia Yueting's route; from a brainwashing perspective, he's far more effective than Steve Jobs or Lei Jun.

"Forget it, I can't learn from them. If I had their abilities, I would have been rich in my past life."

[Have some confidence. You have extensive experience in murder mystery games and are quite skilled at werewolf games, which gives you an advantage in jobs like those involving pyramid schemes.]

"It's getting more and more outrageous. But if I blush when I get on stage, just talk to me, and maybe I'll forget about it. As for my speaking style, let's skip the fluff and go for the direct approach. What kind of young people aren't bold and outspoken?"

As they chatted away with Jasmine, Chen Yiran led the company team onto a bus.

The company mobilized its entire workforce this time, because Chen Yiran felt that this was the first turning point for the company's rise.

There's a saying, "Listen to the dragon's roar."

Tomorrow is the day of Dragon's Roar, and he hopes that all employees in the company can see it to deepen their sense of belonging to the company.

The first day was just an appetizer. Nvidia would arrange buses to take important VIP clients, key media outlets, academic experts, and teams from partner companies like Chen Yiran to Nvidia's headquarters in Santa Clara for a tour.

Nvidia's headquarters is quite impressive, consisting of two huge, technologically advanced triangular buildings with coated glass facades that change color under sunlight.

It is said that its shape is modeled after two graphics card core chips scattered on the 4, and it has been given very stylish names such as Endeavour and Discovery.

On the lawn at the entrance, there stands a black stone monument about half a person's height, engraved with Nvidia's iconic spiral eye logo.

Upon entering the building, you are greeted by a three-story atrium of white marble, and a two-story-high LED video wall.

The rest of the tour wasn't very interesting. After all, a typical tour doesn't showcase too much technology; it's basically just a demonstration of basic lighting rendering, particle effects, and physical collision calculations.

Chen Yiran also found the data center and CUDA developer lab behind it rather uninteresting.

The people at the company, however, were very excited, chattering excitedly about Nvidia's headquarters.

People don't really have a strong sense of the technology, but the general consensus is that Nvidia's building is quite nice, and they hope Chen Yiran will continue to work hard.

Next was a self-service tea break, which was basically a white-style cold meal. It wasn't that bad, but it certainly wasn't delicious. Chen Yiran barely touched it.

Nvidia's PR team started giving out souvenirs, including the usual notebooks and pens, as well as a 32GB metal USB drive with the CUDA logo.

Sure, it was worth the trip, this USB drive isn't cheap.

Afterwards, all company employees moved into a hotel arranged by Nvidia. It has to be a big company. This time, they arranged accommodation for one or two thousand people, and the nearby five-star hotels were almost all booked.

When Huang Renxun, dressed in a leather jacket, stepped onto the stage at nine o'clock the next morning, there were already more than a thousand people sitting below.

The opening keynote speech, which lasted an hour and a half, was packed with valuable information. It summarized the tremendous success of the Fermi architecture (GF100) launched six months ago and unveiled the roadmap for the next two generations of GPUs for the first time.

Chen Yiran's Deepflow was also fortunate enough to be released, and it received considerable attention.

In fact, Huang, who was wearing a leather jacket, had suggested that he perform in this segment, but Chen Yiran declined.

Because if he went up now, it would be like becoming an appendage of Nvidia, and he didn't want to do that.

So only the guy in the yellow jacket released a bunch of data and touted a bunch of advantages.

This led to the bombshell announcement that CUDA would be developed for the x86 architecture, marking CUDA's move out of NVIDIA's hardware ecosystem and into the field of general-purpose HPC high-performance computing.

Then, in a very ceremonial manner, Nvidia announced to the audience that it would no longer be just a company that makes gaming graphics cards, but would conquer consumers with GeForce, take over professional workstations with Quadro, enter the high-performance computing (HPC) market with Tesla, and add Tegra for mobile devices, with the goal of dominating the entire chip industry.

To showcase its capabilities, Nvidia even invited the chief surgeon from the Pacific Medical Center in California to demonstrate the application of Nvidia technology in exercise-compensated cardiac surgery.

Now they can use the powerful real-time image processing capabilities of GPUs to make surgical robots move in sync with the rhythm of a patient's heartbeat, thus performing delicate surgery on a beating heart.

At noon, a dozen or so people from the company bought some Mexican burritos from the food truck and sat on the lawn to have lunch.

With the time for her speech drawing closer, Chen Yiran was still a little nervous.

His subordinates were all very lively and chattering. Gao Dewei even brought a rather professional small camera to record Chen Yiran's first public appearance.

At 2:10 p.m., Conference Room B was already packed with people. All two hundred seats were filled, the aisles were also full, and the back of the conference room was also occupied.

Chen Yiran underestimated his own reputation; he thought only a hundred or so people would come to his speech. In reality, even if he were to go to the large conference hall with a capacity of a thousand in the morning, it might not be able to fill it up.

After all, he's an 18-year-old prodigy, a $200 million miracle, and there were even recent rumors that Microsoft wanted to acquire his new company, but he refused.

Even in Silicon Valley, a place teeming with geniuses and miracles, Chen Yiran's achievements are remarkable.

Moreover, he was an independent developer before, and now this is his first time truly starting a company in the field of EDA.

Many people actually secretly hope that he can break the dominance of the three major companies and make some noise.

It's not about profit; it's purely the human urge to watch the drama unfold. Who wouldn't want to see a genius warrior slay a dragon?

Chen Yiran went on stage wearing a low-cut white sweater, and Zhang Minghui and the staff helped him connect to the computer.

[Don't forget the performance tutorials we found online. Stand tall, look at the center of the audience, and when giving your opening remarks, use a left-right-forward, forward-looking glance to boost your presence! Keep it up! Stick to our strategy: a humorous opening, then relentlessly attack our competitors!]

Chen Yiran took a deep breath and then began to speak: "Hello everyone, my name is Chen Yiran. This is my first time giving a speech or holding a press conference in front of so many people, and I'm quite nervous. After all, I'm only eighteen years old. I've watched a lot of videos by my seniors to learn from them. And look at my clothes, right?"

It was intentional; Steve Jobs wore a black turtleneck, so I ordered a white low-cut one. I hope to have more opportunities to do fashion shows at major events in the future, so that this can become my personal signature.

To be honest, I only truly understood how difficult it was for Steve Jobs when I put on a sweater. In Silicon Valley, where it's always hot, wearing a sweater all the time must be quite a challenge.

Most of the people in the audience laughed.

After completing the opening task of looking around, Chen Yiran began to move directly into the overall presentation.

"Most people know me because Microsoft generously gave me $200 million to buy my compiler."

But actually, this wasn't my first startup project. It was a compiler; I just wrote it as a side project because current compilers were inconvenient. I never expected it would sell for so much money.

Back in March of this year, when I was still at the University of Delaware, I suddenly had the idea to write a framework, a framework that is not yet perfect, a deep learning framework.

After arriving in Silicon Valley, I first worked on the compiler before finally finishing writing this framework.

Yesterday, Mr. Huang in leather jacket demonstrated the power of Deepflow. Now, the improved Deepflow shows even greater power and can adapt to various large-scale matrix operations.

What I'm going to talk about today is his application in EDA.

After that, Chen Yiran started posting a lot of pictures and making high-intensity comparisons. Her language became increasingly sharp, and she almost started to openly mock the three EDA companies.

"Come on, let's start with the most basic ones, the four benchmark use cases of ISPD 2006. Let me see, these are the four best results, each from one of the three super impressive companies. I won't mention their names, but they're supposedly huge companies. It would be quite troublesome if their technical skills weren't up to par and their legal teams came and threw a tantrum."

The four scores are 1.12, 1.08, 1.15, and 1.10. Our Grey students' scores are 1.05, 1.01, 1.06, and 1.03.

It's alright, less than a tenth of the difference.

Next came the time comparison, with "???" written on the left and "Grey" on the right, but everyone below understood that the comparison was definitely about time.

At that moment, a gasp of surprise came from below.

Because Synaptics' single-round time is 4 hours and 12 minutes, while Grey's is 5 minutes.

The extremely fast time makes it possible to directly iterate fifty times, but Synopsys cannot complete fifty conventional iterations.

Chen Yiran felt he was already in the zone; after all, his skills were overwhelming, so he could naturally output freely.

"I know some of you are shocked and a bit skeptical, but I want to say that Grey has already signed seat contracts with several different companies."

Of course, they didn't reveal our impressive technical capabilities, probably so they could use our software before you do and design even better products. Those who haven't used our products yet should contact us as soon as possible.

After all, we're not one of the Big Three companies that relies on a large workforce but lacks real technology. We're a small startup with strong technical capabilities but a small team. When the time comes, we won't have enough technical staff, and there will be a waiting list for collaboration.

Chen Yiran pressed the controller again, and this time there was only one word on the screen: Dark Silicon.

"Next, I'll talk about the dark silicon problem. I've heard that the three major tech giants haven't done any systematic optimization for the dark silicon phenomenon yet? Let me explain our solution. I can say for sure that we have a unique optimization for preventing dark silicon."

Any chip using Grey's design inherently possesses advantages that other chips cannot match.