TokenPulse, a next-generation AI inference infrastructure company, has inaugurated its Asia-Pacific Operating Hub in Singapore. The launch event, attended by industry leaders and ecosystem partners, highlighted the growing demand for AI compute and global infrastructure. The company aims to connect North American AI technology with Southeast Asia’s expanding infrastructure, enhancing its capacity across North America, Asia-Pacific, and Europe.
The global AI infrastructure market is expected to double from US$98.2b in 2025 to US$198t by 2027, driven by increasing inference demand. TokenPulse offers high-performance, cost-efficient AI compute by integrating data centre resources, GPU infrastructure, and an optimised inference engine. This positions the company to support foundation model companies, AI platforms, and AI-native applications.
Singapore serves as a strategic hub for TokenPulse’s operations, enabling partnerships across Malaysia, Indonesia, Thailand, and beyond. The hub will facilitate regional operations and infrastructure partnerships, bringing compute capacity closer to customers and AI workloads. CEO Steven Zhou emphasised the shift in AI compute demand from training to inference, noting that “inference is becoming a core competitive factor.”
TokenPulse provides flexible consumption models, including Serverless, On-Demand, and Reserved options, allowing customers to scale their compute capacity without replatforming. The company also offers deployment options such as TokenPulse Cloud, Private VPC, and BYOC (Bring Your Own Cloud).
With a leadership team experienced in AI, machine learning, and global business operations, TokenPulse is poised to optimise inference for performance and cost. Chairman Thomas Pang highlighted Singapore’s mature data centre ecosystem and strong network connectivity as key factors for choosing the location. The company plans to continue investing in compute capacity and ecosystem support to deliver optimal performance at the lowest cost per token.
