5Y News | Congratulations to Moore Threads on its successful STAR Market IPO!

On December 5, 2025, Moore Threads Intelligent Technology (Beijing) Co., Ltd. (abbreviated as Moore Threads, stock code: 688795.SH) officially listed on the STAR Market of the Shanghai Stock Exchange, becoming China's first full-function GPU company to enter the capital markets.

After five years of entrepreneurship, this IPO marks a critical milestone in the company's development. It represents recognition of the company's long-standing commitment to independent innovation in core technologies and rapid product iteration, and signals that Moore Threads has initially established a systematic capability integrating technology, products, market reach, and capital in the strategic domain of GPU computing power. For the path toward domestically developed high-end GPUs, it also represents a symbolic new starting point.

Founded in 2020, Moore Threads focuses on full-function GPUs and is dedicated to providing accelerated computing infrastructure and one-stop solutions globally, delivering powerful AI computing support for the digital and intelligent transformation of industries across the board.

5Y Capital has supported Moore Threads across four consecutive rounds since 2020. 5Y Capital partner Fei Zhang commented: "Jianzhong Zhang is a truly distinctive founder. He possesses strong judgment on the rhythm of technology evolution and long-term industry trends, combined with powerful organizational charisma and remarkable composure in the face of difficulties and uncertainty. These qualities have been critical to the company's sustained progress on the high-barrier, long-cycle development path of full-function GPUs. Moore Threads' positioning within China's domestic computing infrastructure has become increasingly clear. The company has already achieved scaled delivery of 10,000-card clusters and supports mainstream large model training and inference. GPU is an ecosystem comprising hardware, underlying software, frameworks, and applications — it requires long-term accumulation, and Moore Threads has entered a critical stage of ecosystem building.

Hard tech is a marathon spanning more than a decade. GPU evolution still has 5–10 years of runway ahead, and the path toward more general-purpose intelligence may extend beyond two decades. In such a lengthy and uncertain cycle, what truly determines how far a company goes is not how early it started, but judgment, resilience, and sustained execution. In the vast domestic market, the importance of these long-term capabilities is especially pronounced. The IPO is a significant milestone. Congratulations to Moore Threads, and we look forward to the team maintaining its composure and advancing steadily on this ultra-long track."

5Y Capital partner Kai Liu commented: "Computing power is one of the most fundamental links in the AI industry chain. Over the past decade, NVIDIA has proven that in the large model era, compute chips determine the speed and ceiling of the entire ecosystem. Moore Threads is among the earliest domestic teams to achieve deep compatibility with NVIDIA's software ecosystem, demonstrating leading advantages in product performance, software compatibility, and R&D velocity. After 2024, as large model training scales expanded rapidly, industry concerns grew about whether domestic GPUs could build large-scale clusters. Moore Threads was the first to break through this bottleneck, developing and delivering 10,000-card GPU clusters and collaborating with multiple leading large model companies and cloud service providers on joint technical breakthroughs. From technology evolution speed to execution capability, Moore Threads is at the forefront of the domestic GPU camp, and we have full confidence in the team's technical accumulation and execution power.

The IPO is a critical node on the company's development path. We look forward to Moore Threads gaining greater market recognition after listing, attracting more outstanding talent and industry resources, and increasing its penetration and ecosystem influence in the construction of China's domestic computing infrastructure."

Building a Foundation of Independent Technology, Leading Full-Function GPU Innovation

Over the past two decades, GPUs have continuously driven and led the accelerated computing revolution, from graphics processing to artificial intelligence. Today, amid intensifying global technology competition, computing power equals national power, and advanced AI computing has risen to become a critical strategic resource for national development initiative. As the core compute engine of the AI revolution and digital economic development, achieving autonomous control over GPUs has become a strategic high ground that China must secure in global computing competition.

Since its founding in 2020, Moore Threads has been a rare domestic leader in full-function GPUs, consistently rooted in independent innovation and guided by forward-looking vision, continuously deepening its design and R&D of full-function GPU chips. The company's development path resonates with the national strategic direction of promoting high-level sci-tech self-reliance and building a secure, reliable computing infrastructure.

Based on its self-developed MUSA unified system architecture, Moore Threads was the first to achieve the technical breakthrough of supporting AI computing acceleration, graphics rendering, physics simulation and scientific computing, and ultra-high-definition video encoding and decoding on a single chip, setting a key technical benchmark for the autonomous development of domestic high-end GPU chips. Relying on the "full-function + MUSA" collaboratively driven technical system, Moore Threads has built solid differentiated advantages and full-stack technical capabilities. This not only establishes the company's unique position and competitiveness in the industry, but also equips it with the critical architectural capabilities needed for future computing scenarios including embodied intelligence, digital twins, physical AI, and AI for Science, thereby laying a sustainably evolvable technical foundation for participating in building and leading the next-generation intelligent computing ecosystem.

After five years of innovative cultivation, Moore Threads has successfully mass-produced five chips, iterating through four generations of GPU architecture and intelligent SoC products, achieving diverse layout from chips to compute cards to intelligent computing clusters, forming a complete computing acceleration product matrix covering artificial intelligence, scientific computing, and graphics rendering, with full support for cloud-edge-device scenarios. At the same time, the company has continuously strengthened its independent intellectual property system. As of June 2025, the company had obtained 514 authorized patents, ranking among the top domestic GPU companies. The deep integration of these solid patent achievements with full-stack product capabilities jointly constructs a technical moat centered on sustained independent innovation.

"AI + Graphics" Dual-Wheel Drive, Achieving Sustainable Commercial Growth

On the path toward AGI, computing power is not only the engine but also the foundation shaping the future of intelligence. Moore Threads' KUAE intelligent computing cluster supports 10,000-card scale expansion capability, aiming to build a domestic general-purpose accelerated computing platform with petaflops-level floating-point computing power. Based on the MUSA unified architecture, this platform maintains compatibility with mainstream international GPU ecosystems while achieving technical autonomy, capable of providing autonomous, reliable, and efficient computing infrastructure for training trillion-parameter large models. With KUAE at its core, Moore Threads is pursuing system-level engineering innovation to build an "AI super factory" for the AGI era that can continuously produce advanced models.

In the consumer and professional graphics markets, Moore Threads launched the MTT S80, China's first domestic gaming graphics card. Through continuous system optimization and driver iteration, it can now smoothly run top-tier AAA games including Black Myth: Wukong. Based on its autonomous MUSA architecture, the company was the first to release a graphics acceleration engine supporting DirectX 12, with full compatibility with mainstream graphics and computing ecosystems including OpenGL and Vulkan, and seamless adaptation to multiple operating systems and CPU platforms — a breakthrough that elevates domestic graphics card compatibility and experience to new heights. Similarly in the professional graphics domain, its products have been deeply applied in key industries including industrial design and geographic information, passing multiple domestic certification processes, continuously driving the maturation and improvement of the domestic high-performance graphics ecosystem.

Solid technical accumulation and commercialization pathways, driven by both policy guidance and market demand, have translated into robust performance and growth resilience. Between 2022 and 2024, the company's operating revenue achieved a compound annual growth rate of 208.44%, realizing leapfrog development. In the first half of 2025, revenue reached 702 million yuan, surpassing the total of the previous three years, with AI intelligent computing business contributing 94.73% of revenue, fully demonstrating the company's focus and leading advantage in its core track. During the same period, gross margin significantly improved to 69.14%, confirming its products' excellent competitiveness and market recognition.

Behind this series of achievements lies the company's long-term commitment and high-intensity investment in R&D. From 2022 to June 2025, cumulative R&D investment exceeded 4.3 billion yuan, with R&D personnel accounting for over 77% of total staff. Sustained investment has yielded abundant innovative results: on key performance indicators, the company's products have approached or reached international advanced levels. In particular, the KUAE 10,000-card intelligent computing cluster built on its latest AI flagship intelligent computing product MTT S5000 can efficiently support trillion-parameter large model training, with computing efficiency matching that of foreign same-generation GPU clusters at equivalent scale.

Grounded in National Strategy, Forging a Foundation for High-End Domestic Computing Power

Grounded in the strategic vision of national sci-tech self-reliance, the 14th Five-Year Plan explicitly designates high-end chips as a key breakthrough area, focusing on solidifying the computing foundation for AI development. This top-level design not only establishes the core position of GPU-centered compute chips in driving industrial upgrading and shaping future competitiveness, but also opens an important strategic window for building an autonomous and controllable domestic computing system and achieving scaled substitution.

Seizing this historical opportunity, Moore Threads' STAR Market listing comes at precisely the right moment. The proceeds from this offering will be directed primarily toward R&D of next-generation autonomous and controllable AI training-inference integrated chips, graphics chips, and AI SoC chips. The advancement of these core projects aims to further consolidate and expand the company's technical leading advantages in AI computing power and graphics computing, building a more complete and internationally competitive autonomous product system, and providing solid, reliable, and autonomous computing support for the intelligent transformation of thousands of industries.

A successful IPO is a milestone of the journey past, and even more so a new starting point facing the future. Standing at this new beginning, with unchanged original aspiration and mission on its shoulders, Moore Threads will continue to uphold its belief in independent innovation, continuously deepen its work at the frontier of high-performance GPUs and intelligent computing clusters, drive the computing industry toward greater generality and intelligence, dedicate itself to solidifying the autonomous computing foundation for the AI era, and contribute solid and far-reaching foundational strength to the flourishing development of the national digital economy and the era's mission of sci-tech self-reliance.

5Y Capital seeks out, supports, and inspires founders who dare to be different, providing support from spirit to all operational matters. We believe that if the "crazy" you in others' eyes begins to be believed in, the world will become a different place.

BEIJING · SHANGHAI · SHENZHEN · HONG KONG

WWW.5YCAP.COM