When a high-profile hardware startup posts a 147 percent revenue jump in six months, people pay attention. When that same company maps out an international expansion strategy through a Hong Kong secondary listing, the stakes get much higher.
Moore Threads just posted its first-half 2026 financial results, bringing in 1.736 billion yuan and proving that domestic GPU development has officially cleared the experimental hurdle. Founded by former Nvidia China executive Zhang Jianzhong, the Beijing-based chip designer isn't just surviving under heavy US export restrictions. They are scaling at a pace few industry watchers anticipated.
Moving Past the Startup Phase
Let's look at the actual numbers because hardware tech reporting loves hype over balance. In the first half of 2026, Moore Threads pulled in 1.736 billion yuan, marking a 147.42 percent increase compared to the same period a year prior. Gross profit crossed 989 million yuan, up over 103 percent.
More importantly, the company's net losses narrowed dramatically. For years, building high-end graphics processors required burning through massive amounts of cash with little near-term return. Moore Threads reported cumulative research and development spending approaching 5.9 billion yuan since 2022, with another 769 million yuan funneled into R&D during the first half of 2026 alone.
This isn't a company coasting on government subsidies. They are aggressively spending on engineering while commercial revenue finally catches up to the cost structure.
The Hardware Driver Behind the Surge
Why is revenue doubling over and over? The answer sits inside data centers rather than consumer desktop rigs.
While Moore Threads originally gained attention for graphics cards aimed at visual rendering and gaming, the real financial engine now is artificial intelligence infrastructure. Their flagship MTT S5000 AI computing cluster has achieved scaled commercial sales, landing deployments across major tech hubs like Beijing, Wuxi, and Hangzhou.
Training trillion-parameter large language models demands massive compute clusters. When international hardware options disappeared from the local market due to trade barriers, domestic labs needed an alternative that actually worked at scale. The MTT S5000 became that alternative, passing official evaluation checks and proving its worth in high-end training scenarios.
Hardware alone doesn't secure enterprise clients. Software lock-in is real, and companies hate migrating away from established developer ecosystems.
Moore Threads poured substantial resources into expanding its MUSA software stack ecosystem. Crossing a threshold of over 800,000 registered developers changes the equation. When an engineer can write code that runs natively on domestic accelerators without rewriting half their pipeline, adoption rates spike. That software stickiness explains why their Kuae AI computing clusters are moving so fast from warehouses into active deployment.
The Real Reason for a Hong Kong Listing
Market observers shouldn't view the planned Hong Kong listing as a simple cash grab. The company's shares have already surged more than 420 percent since their initial debut on the Shanghai STAR Market, where they raised 8 billion yuan in a heavily oversubscribed offering.
A Hong Kong dual listing serves a different master: global talent acquisition and internationalization.
As global supply chains fracture and tech pools isolate, cross-border visibility matters. Accessing international capital markets gives firms based in mainland China a structural advantage when hiring top-tier engineering talent from overseas hubs. It also signals to international enterprise clients that the infrastructure complies with broader financial transparency standards.
Chinese tech companies have flooded into Hong Kong as regulatory walls tighten elsewhere, with share sales hitting multi-year highs. Moore Threads wants a seat at that table to cement its status as a permanent fixture in the global semiconductor conversation.
Building silicon from scratch is brutal. Most hardware startups fail quietly before ever taping out a competitive chip. Moore Threads survived the initial valley of death by pivoting hard toward enterprise AI workloads at the exact moment the market needed domestic options most.
The upcoming Hong Kong entry will test whether public markets outside the mainland are ready to price domestic Chinese AI hardware on par with global giants. Keep an eye on their R&D burn rate versus deployment velocity over the next two quarters. If they keep narrowing losses while scaling the MUSA ecosystem, the narrative shifts from substitution to genuine competition.
