Shanghai-based semiconductor company Iluvatar CoreX has secured approval to list on the Hong Kong Stock Exchange, according to exchange filings published on December 19. The company confirmed the development when contacted by KrASIA, describing the approval as a key step toward bringing a domestic general-purpose GPU maker into China’s public capital markets.

For decades, modern computing has relied on a division of labor between central processing units (CPUs) and graphics processing units (GPUs). CPUs serve as the system’s control center, optimized for decision-making, branching logic, memory management, and task coordination. GPUs, by contrast, function as high-throughput engines, designed to apply the same mathematical operations across large volumes of data in parallel, a capability that historically made them well suited to computer graphics.

The rise of general-purpose GPUs, also commonly known as GPGPUs, followed breakthroughs in artificial intelligence. Developers found that many computational problems could be reformulated as large tensor or matrix operations, enabling CPUs to offload structurally simple but computationally intensive workloads to GPUs. This shift significantly improved both performance and energy efficiency.

Within China, Iluvatar is one of several domestic companies focused on GPGPU development, positioning itself as part of the country’s broader AI infrastructure buildout. According to the company, it is among a small group that has not only developed GPGPU technology but also achieved mass production and commercial deployment, translating technical progress into revenue.

Iluvatar has also said it was the first domestic company to develop GPGPU chips and the first to mass produce and deploy seven-nanometer GPGPU products in real-world applications.

GPGPUs are widely regarded as one of the more technically demanding segments of the semiconductor sector, given their architectural complexity and deep integration with software systems. Iluvatar attributes its progress to a proprietary innovation strategy pursued since its founding. According to the company, it has built an internal IP system spanning computing units, cache architectures, data transmission, compilers, performance-optimized libraries, and system interconnection protocols. This approach, Iluvatar said, allows it to maintain control over core technologies while integrating hardware and software into deployable systems.

The company’s GPGPU development milestones can be traced back to 2021, when it powered on its first TG Gen 1, or TG 100, chip. Iluvatar characterized the event as a breakthrough for domestic GPGPU development. The same year, it formally launched the TG series and entered mass production and commercialization the same year. Iluvatar said the TG 100 achieved performance parity with mainstream international products. It later introduced the TG Gen 2 and TG Gen 3 to further consolidate its position in the market.

In 2022, Iluvatar introduced the ZK Gen 1, or ZK 100, chip, targeting AI inference workloads. The company said the chip was the first GPGPU designed in China specifically for inference use cases. ZK 100 and a related variant, ZK Gen 1X, are currently in mass production, with sales underway, according to Iluvatar.

Across its TG and ZK product lines, the company aims to address computing needs ranging from AI model training to cloud-based and edge-side inference. Iluvatar has emphasized its use of seven-nanometer manufacturing processes for GPGPU production, which it said was a first for China, citing Frost & Sullivan.

Iluvatar has also highlighted the scalability of its commercial model. Company-provided data shows that GPGPU shipments reached approximately 7,800 units in 2022, 12,700 units in 2023, and 16,800 units in 2024. In the first half of 2025, shipments totaled about 15,700 units, representing a 227% increase from roughly 4,800 units during the same period a year earlier.

As of June 30, Iluvatar reportedly delivered more than 52,000 GPGPUs to over 290 customers across multiple industries. Based on shipment volumes recorded in 2024, the company said it ranked third among domestic GPGPU suppliers in China.

Beyond chip sales, Iluvatar has focused on offering integrated computing solutions, aiming to embed its products deeply within customer systems used for model training, AI inference, and high-performance computing. The company said this strategy has helped increase customer retention and lifetime value.

Looking ahead, Iluvatar expects sustained growth driven by rising demand for AI-related computing and China’s push for technological self-sufficiency. Frost & Sullivan estimates that China’s GPGPU market shipped around 1.6 million units in 2024. Shipments are projected to grow at a compound annual rate of 33% from 2025–2029, indicating continued strong expansion.

As its Hong Kong listing process advances, Iluvatar said it plans to draw on capital market resources to strengthen its technology, product development, and market reach, with the goal of securing the leading position in China’s GPGPU market.

This article was published in partnership with Iluvatar CoreX.