New Stock News | T-Head, under Alibaba (09988), Said to Plan Hong Kong IPO Next Year, Seeking to Raise Up to HK$23.4 Billion

According to reports, T-Head, the semiconductor subsidiary under Alibaba (09988), plans to list in Hong Kong in 2027, seeking to raise at least US$2 billion to US$3 billion, equivalent to approximately HK$15.6 billion to HK$23.4 billion. Information shows that T-Head Semiconductor was established in 2018 and is the wholly-owned semiconductor business entity under Alibaba Group, with end-to-end capabilities spanning chip architecture design, R&D, and product implementation. Currently, the company's product portfolio covers CPU, GPU, NPU, networking, storage, and other fields, with a focus on building a complete chip product system around data centers and AI computing infrastructure. In the AI computing field, T-Head has continued to accelerate product iteration in recent years. In September 2026, Alibaba unveiled its new-generation AI chip Zhenwu V900 at the Apsara Conference. According to Alibaba's disclosure, the chip's performance is approximately 3 times higher than the previous generation Zhenwu M890, features 216GB of memory and 1,200GB/s inter-chip bandwidth, supports FP8 and FP4 precision computing, and can connect up to 500,000 chips to form an ultra-large-scale cluster for frontier large-model training and inference. Alibaba expects Zhenwu V900 to achieve mass production and commercial launch in the first quarter of 2027. Notably, T-Head's commercialization progress is no longer limited to Alibaba's internal ecosystem. Alibaba previously disclosed that, as of early August 2026, Zhenwu-series chips had served more than 650 customers, covering industries such as automotive, finance, energy, and manufacturing, and super-node instances based on the M890 had also been launched on Alibaba Cloud for large-scale sales. This means T-Head is gradually expanding from an internal computing power supplier for the group to the external market. From an industry perspective, T-Head's pursuit of an independent listing comes as global investment in AI computing infrastructure continues to heat up. With the expansion of large-model parameter scales and the accelerated implementation of applications such as AI Agents and embodied intelligence, market demand for high-performance GPUs, servers, network interconnection, and storage chips continues to grow. At the same time, US export restrictions on advanced chips to China have further pushed Chinese tech companies to accelerate the buildout of domestic computing power supply chains, providing new market space for domestic AI chipmakers.