Harvest season.

Today (January 2nd), the domestic GPU unicorn Biren Technology officially listed on the Hong Kong Stock Exchange, creating the "first stock of GPU on the Hong Kong market." For this IPO, Biren Technology set an issue price of HK$19.60 per share, opened with a significant surge, and saw its market capitalization break through HK$100 billion.

Zhang Wen is the driving force behind Biren Technology. His career began on Wall Street; he later returned to China to pursue serial entrepreneurship and formally plunged into the domestic GPU wave in 2019, leading to the birth of Biren Technology. Subsequently, Zhang Wen led Biren Technology in igniting a financing frenzy within the domestic GPU sector, assembling an extensive list of investors.

With today's bell ringing, Songhe Capital has secured another super IPO. This marks the third major exit for Songhe Capital within a remarkably short span: from OnMicro's IPO on the STAR Market on December 16, 2025, to 51WORLD's listing on December 30, and now Biren Technology today. Songhe Capital has achieved a remarkable hat-trick in just 18 days.

Biren's bell ringing delivers a HK$100 billion IPO for Songhe Capital.

"Capable and efficient" was Songhe Capital Founding Partner Li Wei's first impression of Zhang Wen.

In 2018, Zhang Wen joined SenseTime as President, spearheading the establishment of its headquarters in Shanghai. Songhe Capital's investment team first became acquainted with Zhang Wen through SenseTime.

It turns out that as early as 2017, Songhe Capital had already invested in SenseTime, even dedicating post-investment management personnel specifically to the company, building strong connections at all levels, including with Zhang Wen. Through their interactions, Li Wei perceived Zhang Wen as "open-minded, possessing a strong sense of contractual spirit, and full of vitality and innovation."

The gears shifted to 2019. With deep, long-term involvement in emerging industries like integrated circuits and AI, and drawing on years of entrepreneurial and investment experience, Zhang Wen acutely recognized a historic inflection point arriving for China's semiconductor industry. He resolved to join the entrepreneurial wave focused on domestic GPUs, and thus Biren Technology was born.

During this period, Li Wei and former Songhe Capital partner Gu Wenting decided to visit a circle of entrepreneur friends before an attempt to climb Mount Everest. In discussions with Zhang Wen, his vision of creating a "China Core" resonated strongly with Songhe Capital's philosophy of investing to solve critical "chokepoint" challenges. From Songhe's perspective, the GPU is the core component of artificial intelligence, and at that time, 99.9% of GPUs in China relied on imports, representing a critical bottleneck for industrial development.

Upon learning that Biren Technology's new funding round was nearing closure, Li Wei immediately expressed his intent to invest. "The available investment quota was extremely tight at that time, with only a $10 million slot available," Li Wei recalled to investment community media. The internal investment decision at Songhe Capital was made swiftly. In June 2020, merely nine months after its founding, Biren Technology completed a Series A funding round of 1.1 billion RMB, setting a record for the largest Series A round in the domestic industry at that time, with Songhe Capital among the shareholders.

Not long after, during Biren Technology's Pre-B round, Songhe Capital made an additional investment of $10 million, bringing its total commitment to $20 million—undoubtedly a substantial bet.

There is a lesser-known episode here: after investing in Biren, other GPU startups emerged in the market. Songhe Capital researched two other leading GPU companies, but their response was essentially, "Since you've already invested in Biren, let's forget it." Zhang Wen's attitude towards this matter was completely opposite. "He consistently maintained an open mindset and would not interfere with our potential investments in other companies within the same field," Li Wei recalled.

The subsequent story is widely known: Biren Technology rose rapidly, not only securing multiple funding rounds but also achieving technological breakthroughs. It developed its first-generation GPGPU architecture and multiple chip products like the BR106, BR110, and BR166, launching a series of hardware products based on these and becoming a representative enterprise in the domestic GPU field.

The entrepreneurial journey is never smooth, and Biren Technology also faced its share of challenges. Throughout, Songhe Capital remained a steadfast supporter. For this Hong Kong IPO, Songhe Capital acted as a lead cornerstone investor. According to the prospectus, Songhe Capital held a 1.89% stake in Biren Technology pre-IPO, making it a significant shareholder. Calculated based on the valuation at Songhe's initial entry, this investment has yielded a return exceeding 60 times.

Three IPOs in 18 days reveal a hard tech investment map.

Looking back, Biren Technology serves as a validation for Songhe Capital—it confirms the path the firm has persistently followed over the long term.

Centered around digital technology with AI at its core, precision medicine guided by genetic science, and innovative materials encompassing import substitution, Songhe anchors itself at the critical "chokepoint" links within industrial chains. Starting from leading companies as focal points, it delves deeply upstream and downstream, aiming to identify high-potential enterprises that genuinely solve core pain points early on and accompany them long-term.

As one of the earliest domestic VCs to heavily invest in AI, Songhe initially focused on targets with the potential to become platform companies, such as its 2017 investment in SenseTime. It was through deeply engaging with SenseTime's upstream and downstream ecosystem that the Songhe team formed a clear judgment: the development of China's AI industry is highly dependent on future providers of computing power chips. These two fields are inextricably linked.

"Understanding the industry chain down to its very foundation." The investment in Biren Technology was a natural extension of this logic at the intersection of semiconductors and AI.

Yet, Songhe Capital's hard tech portfolio extends far beyond this. Over the years, the team has deeply penetrated the broader chip industry, making systematic investments across semiconductor sub-sectors like GPU, CPU, and DPU, as well as in sensor fields including lidar, millimeter-wave radar, and high-precision inertial navigation. A robust map centered on computing power and perception has emerged.

In computing power, besides Biren Technology, Songhe has also invested in Hongjun Micro, a domestic Arm CPU developer, and Nebula Matrix, a DPU chip designer. At the design tool level, it invested in Xinyahui, which recently completed its IPO辅导. Extending downstream to AI applications, it has backed a roster of star companies like 4Paradigm, Pony.ai, and SmartMore.

Just half a month ago, Songhe Capital harvested another chip IPO—OnMicro. Founded in 2012, this company focuses on the RF front-end chip sector, long dominated by international giants like Qualcomm. Songhe invested in OnMicro in 2020, accompanying it as it broke the international monopoly on 5G high-end RF modules, achieved mass production and adoption by leading clients, and continued to expand into new scenarios like satellite and automotive. In December 2025, OnMicro successfully listed on the STAR Market, giving Songhe another hard tech IPO valued in the tens of billions.

On December 30th, 51WORLD, specializing in digital twin technology, officially listed on the Hong Kong Stock Exchange, creating the "first stock of Physical AI" with a market cap of approximately HK$17 billion. Thus, within less than 20 days, Songhe Capital clinched three IPOs, welcoming a true harvest season.

A broader review of its portfolio—from companies like Dynanonic, Ruifeng New Materials, BGI, MGI, to UFactory, Orbbec, CloudMinds, UBTECH, DJI Innovations—reveals that Songhe's hard tech landscape has taken shape precisely through this philosophy of "daring to invest early and committing to long-term companionship."

Three decades of market fluctuations have taught the firm to maintain a sense of reverence.

Within the venture capital landscape, Songhe Capital possesses a distinct character.

This is closely tied to the career trajectory of its leader, Li Wei. In 1981, Li Wei was admitted to the Chemistry Department of Peking University. His first job after graduation was at Peking University's Science and Technology Development Department, working on the commercialization of scientific research成果, making him one of China's earliest practitioners in "technology industrialization."

In 1991, Li Wei moved south to Shenzhen, even transferring his Beijing household registration to the non-agricultural urban collective of the then Bao'an County. In Shenzhen, he joined Bao'an Group, handling securities affairs. During this time, he participated in several pioneering innovations in China's capital markets, including designing and issuing China's first convertible bond, its first medium-to-long-term warrant, and taking part in one of the earliest secondary market acquisitions.

At the end of 1996, Li Wei co-founded Songhe Capital with Luo Fei, the name inspired by the ideals of "the pine, upright and unyielding; the grain, thriving and benefiting the world." From that point on, hard technology became the consistent, central theme of Songhe's investments—following the three directions of digital tech, precision medicine, and innovative materials, continuously deepening its layout, and gradually becoming a representative of "Southern-style venture capital."

Thirty years on, Songhe Capital manages assets exceeding RMB 24 billion, has cumulatively invested in 500 companies—over 70% of which were early-stage or small enterprises—and has seen exits through IPO or M&A for more than 80 portfolio companies. A significant portion of these are long-term projects that Songhe accompanied starting from the angel round.

These three decades also mirror the venture capital industry's evolution in China, from non-existence to摸索 to maturity. Throughout, investment trends have rotated, and market noise has been constant, but Songhe's focus on hard tech has remained unchanged, allowing it to fully witness and deeply participate in the rise of China's technology sector.

Behind this steadfastness lies its consistently upheld investment philosophy: maintain discipline, delve deeply only within areas of profound understanding; avoid excessively chasing trends, and refrain from轻易 stepping beyond established boundaries.

Not long ago, an unprecedentedly large, state-level venture capital guidance fund focused on hard tech was officially launched. Among the first 49 sub-funds selected, Songhe Capital was included. This speaks volumes.

Reflecting on the recently concluded year of 2025, Li Wei provided a concise summary: the firm raised over RMB 3.2 billion throughout the year, including a RMB 1.5 billion specialized cell and gene therapy fund within Shenzhen's "20+8" industrial fund cluster. On the investment side, adhering to its "invest early, invest small" principle, it deployed approximately RMB 1.2 billion across nearly 40 projects, many of which were nascent teams just emerging from laboratories. Regarding exits, a total of 10 portfolio companies achieved IPOs during the year, with another batch having received approval for Hong Kong IPOs and awaiting listing.

"Overall, 2025 was a year of recovery for the industry," Li Wei shared, describing his personal sense of the market.

Towards the end of the conversation, Li Wei discussed "capital for good." He believes that while capital seeks returns, it must also consider doing good, considering how to integrate with and develop alongside society.

First, maximize the value of the portfolio companies. Entrepreneurs are the artists; we, as appreciators, provide the ink, brushes, paper, and inkstone during their creative process—offering help without causing disruption. Second, we must be responsible to our investors. They trust us and provide capital for us to manage. As fiduciaries, we must strictly uphold our responsibility to them. Third, the young teams are all chasing their dreams. We must treat them well, create platforms for them, provide opportunities, and help them achieve their ideals.

As Li Wei stated, "Hard tech investment is 'non-consensus investment'; it requires growing together with entrepreneurs, using time and patience to wait for technology to bear fruit."

"Venture capital is about seeing a glimmer of light, entering through a narrow gate, and walking a perilous path." Li Wei has emphasized multiple times that steady progress on a challenging path requires maintaining reverence. Entering 2026, he believes the scenery in venture capital is worth anticipating—more positive developments are unfolding.