King of Technology: A Fresh Start in 1983

In the capital, the late autumn wind carries a chilling whistle.

However, the scene in front of the newly completed Joint Laboratory for Advanced Computing Technology (ACL) building in Tsinghua University is bustling with activity.

Huge ACL logo—

"Gears mesh with abstract chips, with a cluster of flames leaping at the center."

Chen Jianghe stood at the front of the crowd, his dark blue suit perfectly pressed.

Su Wanqing, standing beside him, was dressed in a sharp dark gray suit skirt, her long hair neatly tied up, her eyes still carrying a cold and aloof air.

Zhou Yuguang, Li Mingyuan, Zhao Xiaobo, Zhang Wei, Zhao Lin, Xiao Wang and others stood on the left and right, their faces beaming with excitement and pride.

"...Comrade Chen Jianghe led the Tiangong Technology team to achieve a major breakthrough in the research and development of the [Galaxy] supercomputing core coprocessor."

"Their independently developed and designed 'Xuanyuan-M1' chip boasts outstanding performance and has made significant contributions to major national science and technology projects! This is hereby commended!"

The leader of the Ministry of Electronics Industry spoke loudly and powerfully, his voice carrying throughout the venue through the microphone.

Thunderous applause!

The flashbulbs were focused on Chen Jianghe.

He gave a slight bow, his gaze sweeping over the familiar faces in the audience…

Representatives from Tiangong, key technical personnel from Huaxin Precision, and undergraduate and graduate students from Tsinghua University's Institute of Microelectronics.

"Thanks!"

Chen Jianghe's voice was calm and steady, "This honor belongs to the entire Tiangong team!"

"This belongs to every brother and sister who toiled away in makeshift sheds, their eyes bloodshot from working in front of oscilloscopes!"

"More importantly, it belongs to mentors like Engineer Su, Engineer Zhou, and Professor Li, who guide us and help us break through barriers!"

He turned to the side and bowed deeply to Su Wanqing, Zhou Yuguang, and Li Mingyuan!

Su Wanqing looked at Chen Jianghe with a calm and unwavering gaze.

Immediately afterwards, Chen Jianghe stepped forward, his voice suddenly rising in pitch, carrying a sharp edge.

"Today is not only the listing day of ACL, but also the launch day of Tiangong Technology (Group) Co., Ltd."

The audience fell silent for a moment, then erupted in even more enthusiastic applause!

This means that Tiangong has officially transformed from a grassroots workshop into a modern technology enterprise with a complete industrial chain, including chip design, board processing, complete machine assembly, and technology research and development platform!

"Tiangong Group's goal!"

Chen Jianghe's gaze swept across the entire room. "To create world-class core components for computing technology that belong to China!"

"Let the name [Xuanyuan] resound throughout the world!"

With a wave of his hand, he announced, "The first project since the establishment of Tiangong Technology, the development of the [Pangu II] high-performance multimedia coprocessor chip, is officially launched today!"

"The goal is to seize the commanding heights of the PC multimedia wave!"

"[Pangu II]!"

The name, like a pebble thrown into a lake, stirred up ripples in the hearts of the attendees!

Media reporters frantically took notes, while their colleagues displayed a variety of expressions, ranging from amazement and skepticism to a eager desire to take on the challenge.

The ceremony ended and the crowd dispersed.

Chen Jianghe, Su Wanqing, and others stepped into the brand-new ACL building.

In the spacious and bright laboratory, rows of brand-new 386 computers and testing instruments gleamed with a cold luster.

"Jiang He, aren't you taking too big a step?"

Looking at the spirited Chen Jianghe, Zhou Yuguang was not without worry. "[Pangu II] is positioned as a multimedia project, and floating-point arithmetic (FPU) is the core challenge."

"Intel's 387 coprocessor patents are extremely stringent, and circumventing them is not so easy..."

"Mr. Zhou, the market waits for no one!"

Zhao Xiaobo interjected. He is now the group's vice president of marketing, dressed in a suit and tie, but his eyes still carry a rough and sharp edge.

"I'm out in the market all day, so I see it best! PCs with built-in sound cards and the ability to play VCDs are the future trend!"

"Without powerful multimedia processing capabilities, how can the [Xuanyuan] PC compete with other brands?"

"Technology is the foundation."

Su Wanqing's cool voice rang out. She walked to a whiteboard, picked up a pen, and wrote, "The core challenge of [Pangu II] is indeed FPU."

"Intel's x87 instruction set and IEEE 754 floating-point standard implementations are covered by a dense network of patents. A head-on confrontation is a dead end."

She paused, then drew an arrow. "We must find a different path! Very Long Instruction Word (VLIW) architecture!"

"Sacrificing some versatility to optimize the parallel efficiency of multimedia-intensive computations, such as audio decoding and image transformation! This is our breakthrough!"

"VLIW?"

Zhang Wei's eyes lit up. "This is a novel idea! But architecture design and compiler support are major challenges!"

After graduating in 1988, he officially joined Tiangong and is now the head of the chip design team at Tiangong Technology.

"Difficult problems are meant to be overcome."

Chen Jianghe looked at Zhang Wei with unwavering determination: "Zhang Wei, you lead the team and immediately begin preliminary research! I need to see a feasibility report and a preliminary architecture within three months!"

"Yes!" Zhang Wei straightened his back.

...

Chen Jianghe and Su Wanqing walked side by side in the empty corridor.

Sunlight streamed in through the huge floor-to-ceiling windows, casting long shadows on the floor.

"Are you under a lot of pressure?"

Su Wanqing looked straight ahead, her voice calm.

Chen Jianghe was startled again!

I couldn't help but recall what Su Wanqing said this morning...

Did Su Wanqing discover something?

Yes, with her exceptional intelligence, it was impossible for her not to have suspicions about him…

But the question is... why has she been helping him all these years?

"With you here, I feel much more at ease."

Chen Jianghe frankly admitted, then quickly changed the subject, "It's just... this [Pangu II] feels even harder to crack than [Xuanyuan-M1] back then."

"Only things that are tough to crack have value."

Su Wanqing stopped and looked out the window at the bleak autumn scenery. "The pinnacle of technology is built with blood, sweat, and wisdom. Are you scared?"

Chen Jianghe smiled, a smile that carried the stubbornness unique to men from the Loess Plateau.

"Having come this far from my humble beginnings in the tin shacks of Huilongguan, the word 'fear' doesn't exist in my dictionary, Chen Jianghe!"

Su Wanqing didn't say anything more and turned to walk towards her private office.

That aloof figure contained the strength that enabled Chen Jianghe to cut through all obstacles.

...

Late at night in the ACL lab.

Under the stark white light, the air was filled with the bitter smell of coffee and the faint ozone emanating from electronic components.

The whiteboard was covered with complex logic diagrams and timing formulas, and next to it were several simulation waveforms that were densely circled in red.

Zhang Wei's hair was disheveled, his eyes bloodshot, and he scratched his head in frustration: "No! It still won't work!"

"The conflict rate of this VLIW scheduler slot is too high! Parallel efficiency can't be improved at all! Compiler optimizations can't keep up either!"

He pointed to a dismal bubble graph on the screen, where the green fill rate, representing instruction execution efficiency, was pitifully low.

"Zhang Zu, Intel's x87 instruction set is Single Instruction Multiple Data (SIMD), which is inherently more efficient."

One team member wailed, "Writing the compiler for VLIW is pure hell!"

"How about...we consider partial licensing?"

Another team member whispered, "Intel's FPU IP license is expensive, but it's stable..."

A heavy silence filled the laboratory.

The VLIW route has encountered severe technical bottlenecks!

How to write a feasibility report with a three-month timeframe?