AI Infrastructure Faces New Challenges: Models Grow, Heterogeneous Computing and Agent Complexity Reshape the Field

As AI model architectures become increasingly complex and diverse, AI infrastructure must adapt to heterogeneous computing, multi-model pipelines, and rising agent-driven workloads. Experts from Huawei, Inferact vLLM, and 4Paradigm discuss key challenges: managing diverse AI accelerators (NVIDIA, Ascend, Cambricon, etc.), optimizing multi-modal inference, and how AI may eventually automate much of infrastructure management, shifting engineers from implementation to planning and SLA definition.

Huawei 4Paradigm (Fourth Paradigm) DeepSeek Alibaba/Qwen Cosmos (NVIDIA) Google/DeepMind Meta Anthropic

Mineng Technology Secures Tens of Millions in Funding for Self-Developed SNN Brain-Like Chip

Chinese startup Mineng Technology, specializing in physiological brain-like chips, has completed a tens-of-millions RMB equity financing round led by Blue Bay Capital and Wuxi Venture Capital. The funds will be used for mass production of medical-grade standard modules, full-chain closed-loop physiological regulation system development, and ecosystem expansion. The company has developed an event-driven spiking neural network (SNN) brain-like core that mimics the asynchronous processing of the human brain, aiming to solve power consumption and latency issues in wearable medical devices.