Reinventing the Millennium

The news that Microsoft is interested in acquiring Nokia came out of nowhere, but given the moves of similar software giants, many people believe that this possibility is gradually increasing as Nokia's mobile phone business declines.

Just like Google's acquisition of Motorola.

In theory, Microsoft can gain an in-depth understanding of the hardware research and development process and the channels for producing and selling hardware devices, and it is possible to save its mobile strategy.

This is of course very meaningful.

As for Nokia…

Although it is difficult to admit it, as the smartphone giants have consolidated their position in the global market, the once giant mobile phone has become a waste.

It's easy to make Nokia realize this cruel reality. Sales and reports can clearly reveal everything, but it will take time for Nokia's shareholders and executives to accept it.

As an opponent who once fought with Nokia over patents, Fang Zhuo heard the emotion in Yu Hong's voice.

He asked: "Do you feel a sense of crisis when you see a giant like Nokia falling?"

"No, I just feel honored." Yu Hong was still very emotional, "Such a big giant fell like this, and I was also involved in it. I feel a deep sense of satisfaction. If I can share Nokia's patents, then Even better."

Fang Zhuo found that Mr. Yu's eyes were shining when he said this.

He smiled and shook his head: "I'm afraid this is a bit difficult. Even if Nokia wants to sell, it may only sell long-term patent licenses. Maybe we can talk to Google about motorcycle patents in the future."

Google's acquisition of Motorola has caused some changes in the market. At the very least, the former has transformed into a part of the hardware manufacturer, which is always easy for other brands to have doubts. The cooperation between Samsung and Google has cooled down.

Even though, Google has always claimed that the acquisition of Moto can help Android system enrich its patent moat.

"It was strange after Motorola was acquired by Google. Its products were delayed again and again." Yu Hong was a little confused about Motorola, which is currently under the control of Google. "After the acquisition was completed in the fourth quarter of last year, it said that it would devote all its efforts to building an Android phone. , this half year has passed, I can see the news every month, and then I am let go every month."

This situation is not normal.

Motorola must already have products internally, and it should not take too long to adapt to the native Android claimed by the new CEO.

However, with this delay, even if it is released immediately tomorrow, the so-called flagship will soon become obsolete, because this year’s mobile phones equipped with new chips and new devices have begun to become popular.

According to the current time point, we will either have a head-on confrontation with Yusanjia later or directly, and the process may require new process adaptation.

unless……

Now unless Motorola really comes up with a super shiny flagship phone, it can turn the situation around.

"This only proves that everything is not going to be alright with the acquisition, and Motorcycle is currently facing serious internal operational problems." Fang Zhuo said with a smile.

Yu Hong nodded and suddenly asked: "Will Ice Core be able to launch 28nm by the end of the year?"

"It gets more and more difficult as we go forward." Fang Zhuo frowned when he heard this question, "After Dr. Liang solved the technical solution, the progress in Luzhou was pretty smooth. There will be a difficulty next. One is to go further. in terms of craftsmanship, and the second is in terms of our publicity and customers.”

Ice core's progress in 28nm cannot be hyped, and the signals sent to the outside world may also cause unrest among customers. This cannot have both sides.

"It's hard to pass the test." Yu Hong breathed a sigh of relief, "Let's let Mr. Fang suffer the first loss for 28nm."

She made a downward gesture.

There have been previous exchanges about the announcement of new processes, which must be the product of Yike forcing it, possibly overturning, insufficient yield, and an overall loss.

Of course, this will be a third-party source, but the performance of Yike and Bingxin will also be ambiguous.

Fang Zhuo sighed and wanted to talk about Ice Core's current R\u0026D work in three shifts, but felt that there was no need to exaggerate. He shook his head slightly and asked, "Tomorrow I will meet with NVIDIA's Huang Renxun, will you come with me?"

"Don't go, so as not to cause trouble." Yu Hong refused flatly, and then said, "I will ask the secretary to give you two technical studies later. They were done by Siri's team. Oh, yes, we also hired a small team from Google. , Well, they should be considered as the ones who tricked and brought Director Ng of ‘Google Brain’ here.”

Andrew Ng is also an associate professor at Stanford and is very familiar with the Siri team.

He recently made a breakthrough that shocked the industry. He developed an artificial neural network that learned to identify which videos were about cats by watching YouTube for a week.

Some media praised Andrew Ng for turning a new page in the field of artificial intelligence.

The team he leads joining Yike is undoubtedly another complement to its technological R\u0026D capabilities.

After Fang Zhuo got the "technical research" mentioned by Yu Hong, he found that it was not only a description of Yike's related research and development, but also a considerable part related to tomorrow's trip.

According to Ng's research, GPU parallel computing is very consistent with neural networks. In order for AI to recognize a cat, he used 16,000 CPU processors, but after using GPU parallel computing, only 12 were needed.

Simply put, neural networks are "highly" parallel, and the GPU architecture uses more transistors for numerical calculations than for CPUs, rather than for caching and flow control.

In today's industry, only NVIDIA is investing heavily in the research and development of GPUs and related platforms named CUDA.

Fang Zhuo understood what Yu Hong meant. This was also a small bargaining chip that could be discussed with Huang Renxun. Nvidia was also looking for new strategic growth points.

In fact, in addition to GPUs, Nvidia launched Tegra, a chip based on the ARM architecture, in 2008 in an attempt to enter the mobile field, but it did not produce any splash.

Then in 2010, dual-core processors were iterated, and in November last year, NVIDIA cooperated with Taiwanese to launch a quad-core processor, with the intention of accumulating parameters in the multi-core field.

NVIDIA Tegra's "front-running" behavior was recognized by the secondary market. The paper parameters of the quad-core processor were very strong, which once caused the stock price to rise by 30%.

However, when Tegra was installed on HTC, the phone overturned, overheated, and froze, which naturally led to technical doubts and criticism.

In addition to the poor performance of quad-core Tegra, the main reason is that Nvidia is very lacking in baseband technology and does not use Qualcomm's baseband chips.

When Fang Zhuo came to visit this time, he naturally could not only talk about the development of the ice core and semiconductor industries, but also talked about the mutual cooperation space.

In the evening, he carefully reviewed the information again and concentrated on his meeting with Huang Renxun and others.

The next morning, Fang Zhuo drove to NVIDIA's headquarters in Santa Clara and met the Baodao man who founded NVIDIA.

It has to be said that many Chinese from Baodao have really made outstanding achievements in the semiconductor field that have impacted the global industry.

Compared with Qiu Ciyun's previous visit, NVIDIA head Jen-Hsun Huang went to the door in person to greet him this time, and invited the head of the Yike Department into his office with pleasantries.

Just as Fang Zhuo directly praised NVIDIA driven by Huang Renxun, Huang Renxun also admired Mr. Fang who founded the Yike Department in his heart.

Objectively speaking, Mr. Fang is perhaps one of the most well-known Chinese business figures in the world...

Nvidia and Yike only have a shallow and barely competitive relationship. More importantly, they are important overseas customers of Ice Core.

Therefore, Huang Renxun actually didn't have much dealings with Mr. Fang. The last time they met was at an exhibition in Barcelona, ​​and they only had a few words about the dual-core mobile phone processor launched by NVIDIA.

However, he was very familiar with the other party's general actions and business deeds.

Huang Renxun knew what Mr. Fang was here for this time. After pouring the tea, he heard something unexpected said by the other person before he opened his mouth.

"Mr. Huang, I'm here to investigate." Fang Zhuo said with a serious face, "Nvidia's recent stock price is not doing well. You, the CEO, have to explain it to me."

These words were not very polite, and the relationship between the two was not that familiar.

Huang Renxun is a little confused. The stock price has indeed suffered due to chip overturn recently, but... did the person who won Yike this round criticize it when they first met?

He held the tea cup and blinked, with some guessing.

Fang Zhuo revealed the answer that was not too unexpected in an instant: "My MIGA fund holds some NVIDIA stocks. Yesterday, I had a call with the fund manager and heard him mention NVIDIA."

Huang Renxun smiled slightly and wanted to explain the situation of the world's first quad-core processor to the head of one of the Yusan family. Then, he heard a sudden and unexpected supplementary answer.

"However, we are still very optimistic about Nvidia's strategy and prospects, and if we continue to buy, we will almost reach the sign-raising line." Fang Zhuo said with a smile.

Huang Renxun was surprised. The quotation line was 5%. According to Mr. Fang’s tone, it was not far away. This was already considered a lot among Nvidia shareholders.

He was just about to ask about the relevant shareholdings, but then he thought about it. A rich man like Mr. Fang has used it from countless institutions behind the scenes. This statement is credible if put here in person.

"Mr. Fang." Huang Renxun thought for a moment and gave a very formal response, "We will soon launch shareholder feedback actions in the form of stock buybacks and dividends. The success of NVIDIA's overall strategy has given us very good cash flow. It also makes us very confident in our ability to continue to maintain this cash flow."

If nothing else happens, Nvidia will also provide a dividend plan as soon as the fourth quarter.

Fang Zhuo nodded deeply: "We believe in NVIDIA's development strategy. Others talk about it with words, but we support it with our feet."

There is nothing more sincere than handing over your money.

Even though Huang Renxun knew Mr. Fang’s style of accumulating old age, he was still somewhat impressed by what he said. He smiled and said: "Although the CPU chip has suffered a setback this time, our overall growth in the past few years has been okay. What is the basis of the MIGA fund?" Choose Nvidia as a stock allocation?”

"I have always believed that, perhaps because of my own personal outrage, I have always felt that companies that are willing to spend money on research and development are worthy of respect." Fang Zhuo said, "For example, the CUDA project for GPUs that you worked on was questioned at the time for spending money indiscriminately. Don’t you see the light now?”

This is the reason, although the reason why the MIGA fund started is the value dislocation after the financial crisis.

Huang Renxun was scratched, and he perked up and said, "Mr. Fang, indeed, our GPU..."

When he said this, he suddenly remembered the intelligent voice assistant Siri that Yike launched last Christmas, and said with a smile: "Mr. Fang should know that our GPU has good prospects in artificial intelligence."

Fang Zhuo was very concise and got to the point: "Well, I know that from Society of Mind to one algorithm, the development of machine learning will definitely be faster and faster."

The former is a view mentioned by Marvin, the father of artificial intelligence, in his early research. This idea is to build and combine thousands of independent computing modules, which play different roles respectively, and then connect all these modules through algorithms. , used to simulate the brain to generate thoughts.

This technical idea is extremely difficult to implement.

The one algorithm assumes that the units in the human brain do not have a clear division of labor. It is essentially a general-purpose processor that can be adjusted to handle different types of tasks.

This change in overall thinking triggered the rise of deep learning. It can be said that many people in Yike Siri were inspired by this.

Huang Renxun heard the summary of Mr. Fang's words and couldn't help but admire: "I heard that Yike has a strong R\u0026D team in the field of deep learning. Now looking at Mr. Fang, I know that the rumors are true and the relevant literacy is very high."

Fang Zhuo waved his hand: "I just listen to them talk when I have nothing to do. If you really ask me to talk, nothing can be said."

"Mr. Fang is willing to listen and insist on investing, which is remarkable." Huang Renxun said sincerely, "Yike's success is not without reason."

"Now that we're talking about it, we just happened to have a professor join the team, Andrew Ng from Stanford. He also mentioned that the parallel computing of NVIDIA GPU and deep learning are a natural match." Fang Zhuo made full use of the report he learned yesterday.

Huang Renxun was about to nod when he suddenly said in surprise: "Isn't Professor Wu at Google?"

"Yike can spark more sparks in the field of deep learning." Fang Zhuo said calmly, "According to the current technological development, Yike and NVIDIA have deep space for cooperation. We will continue to work on deep learning and hope that in the future It can make a difference.”

Huang Renxun nodded slightly and turned the topic back to today's main topic: "NVIDIA and Yike have already had a deep cooperation, and Ice Core is also from Yike."

Fang Zhuo smiled: "Yes, ice core is the cornerstone of our Yi Ke Department. This is also what I want to talk about when I come to visit Mr. Huang."

Huang Renxun looked at Mr. Fang: "Is Mr. Fang here as the president of Yike or the chairman of Bingxin?"

"None of them." Fang Zhuo shook his head, "I came here as the maintainer of the development order of the global semiconductor industry."

Huang Renxun laughed.

Fang Zhuo shrugged: "But before we talk about this, on behalf of Yike, I would like to invite Nvidia to join our baseband technology research and development alliance with Apple and Samsung."

Huang Renxun looked at the man in front of him and smiled: "Mr. Fang is so polite to come here with a gift like this."

"It's not a gift. These strategic prospects are what Yike attaches great importance to, and they are what I want to talk more about with Mr. Huang." Fang Zhuo said seriously, "As for the ice core, it will be launched soon, and we have already had extensive exchanges. But I still want to talk to our customers about the future.”