Chinese chipmakers and model labs are scaling up domestic AI hardware and open-weight models, from PyTorch Foundation seats and IPOs to a proposed BRICS open-source AI zone.
The headline signals a shift: Chinese GPU vendors are no longer aiming only to match Nvidia’s H100 in single benchmarks. They are building scalable platforms where interconnects, software, and cluster-level throughput matter. For on-premise deploymen...
Xi Jinping's push for an open-source AI zone among BRICS nations shifts the focus from proprietary APIs to shared, auditable code and local infrastructure. It reinforces the case for self-hosted deployments, data sovereignty and hardware independence...
Moonshot AI reports more than US$1 billion in annual recurring revenue as of August and targets US$2 billion by end-2026. The gains come from Kimi subscriptions and services built on the open-weight K3 model. The licence imposes commercial agreements...
Teams running H100 clusters with policies that exclude Chinese models start at a disadvantage: in the open 120B+ vision segment, the gap with GLM, Qwen, and DeepSeek is measurable. Among the Western candidates cited, Inkling Small and Command A+ cove...
Enflame's listing closes the IPO cycle of China's four main GPU producers. The financial signal is clear: Beijing treats AI chips as a strategic priority. But capital alone cannot close the gap in software ecosystem, VRAM, bandwidth, and compatibilit...
Enflame's September 11 listing completes the IPO cycle of China's four major domestic GPU makers. The event goes beyond finance: it signals a structural bet on an alternative AI hardware supply chain, with implications for on-premise deployments, dat...
Chinese chipmaker Loongson is focusing on CPUs and GPUs to meet inference demand, marking a shift in domestic silicon strategy. The news offers no financial details but confirms that inference workloads are becoming the driver for hardware roadmaps a...
PyTorch Conference China 2026 in Shanghai revealed an open source ecosystem increasingly multi-chip and multi-cloud. Alibaba Cloud, Ant Group, and Cambricon joined the PyTorch Foundation; the DeepSeek-R1 case on GB300 showed 2.77x throughput and 60 p...
Cambricon has gained a seat on the PyTorch Foundation board, aiming to strengthen the software ecosystem for Chinese AI chips. The move highlights how competition is shifting from raw hardware to framework integration. For self-hosted and on-premise ...
Cambricon has taken a seat on the PyTorch Foundation board. Beyond representation, it signals that Chinese AI chip makers are focusing on software ecosystem control. With advanced GPU export restrictions, compatibility with dominant frameworks become...
Biren, a Chinese AI GPU developer, is extending R&D into optical interconnects and advanced cooling. The move signals that the next bottleneck for on-premise LLM workloads is not just silicon, but chip-to-chip data movement and heat dissipation. It s...
Qwen has published an open-weight model derived from a 4B checkpoint and fine-tuned for driving. The full BF16 checkpoint is 9B. It signals Chinese labs pushing toward autonomous driving with open models, with implications for local inference, data s...
