Futurenest, recognized for AI governance and infrastructure, joins the inaugural cohort of the “TAI1 AI Accelerator”, launched by StarFab in collaboration with key local and global industry leaders and leveraging resources from the NVIDIA Inception program for startups. Timed for GTC 2026
— Futurenest Joins the Inaugural Cohort of TAI1 AI Accelerator, Showcases Its Flagship Product Xparse at GTC 2026
Futurenest Selected for TAI1 Accelerator
Recognized for its innovative AI governance and execution infrastructure, Futurenest has been officially selected as a member of the inaugural cohort of the “TAI1 AI Accelerator,” a program jointly launched by StarFab and NVIDIA Inception. Through the program, the company will receive business mentorship and AI hardware–software enablement support, along with an early-stage investment of NT$3 million in SAFE (Simple Agreement for Future Equity) from ITIC (Industrial technology investment corporation), accelerating its technology development and market expansion.
While many enterprises remain at the proof-of-concept stage of AI adoption, Taiwan-based AI startup Futurenest is addressing a more critical challenge: how to integrate AI into daily enterprise operations. The company is committed to solving the structural barriers to scalable AI deployment, including cost, energy consumption, and compliance.
Its flagship product, Xparse, is a sovereign governance infrastructure for enterprise AI execution that enables rapid on-premises deployment of large-scale models. With built-in task-level governance, energy attribution, and full audit traceability. The test on the Compal SGX30-2 with a 10U NVIDIA HGX™ B300 server (8 x NVIDIA Blackwell Ultra GPUs) shows that it not only has lower power consumption but also achieved a benchmark breakthrough of 50 concurrent users and 25.5X throughput scaling. It is a practical architecture for running enterprise AI services more efficiently. In this test, we reduced large-model GPU compute energy by 41.1%, lowered total system power by 13.3%, while still maintaining interactive latency and strong concurrency performance.
In addition, GeniAltX, Futurenest’s enterprise AI agent powered by GraphRAG technology, delivers deep analytical capabilities that help organizations accumulate long-term intellectual capital. The company is accelerating its expansion into finance, manufacturing, consulting and accounting professional services, as well as government sectors, driving the scalable adoption of enterprise AI.
Futurenest Collaboration with Compal Electronics and ASRock Rack for NVIDIA GTC 2026
Futurenest, a selectee of the TAI1 Accelerator, is unveiling its latest breakthrough technologies at the NVIDIA GTC 2026 global conference in San Jose, USA. In a joint exhibition with Compal Electronics (TPE: 2324) and ASRock Rack (TPE: 7711), the collaboration integrates Compal’s advanced AI servers boasting massive high-performance computing power and ASRock Rack’s world-first waterless liquid cooling R&D technology with Futurenest’s Xparse AI infrastructure and Xerno enterprise AI knowledge platform. This showcase presents a comprehensive integrated solution spanning High-Performance Computing (HPC) platforms, on-premise deployment, and enterprise-grade AI task governance and audit mechanisms, creating a governable computing architecture designed for the AI generation.
Compal Electronics integrates the Xparse AI infrastructure into this exhibition, highlighting its spotlight product, the SGX30-2, a 10U high-performance rack-mount AI server built on the NVIDIA HGX™ Blackwell Ultra GPUs platform. This system provides ultra-high-density GPU inference and training capabilities, enabling enterprises to run large-scale models and high-performance workloads on-premise. Leveraging its epoch-making on-premise hybrid deployment architecture combined with NVIDIA’s underlying framework and Futurenest’s exclusive patented technology, the solution drastically reduces on-premise deployment time. It supports a privatized AI environment capable of Agent governance and auditing, featuring task-level energy and resource attribution. This ensures complete traceability for every AI task regarding decision-making, compute usage, and cost allocation.
ASRock Rack combines the Xerno enterprise AI knowledge platform with its latest generation NVIDIA HGX B300 Waterless Liquid Cooling accelerated computing server platform. This demonstration illustrates how enterprises can transform data dispersed across file systems, databases, and operational systems into knowledge assets capable of real-time AI search, reasoning, and generation atop high-performance computing infrastructure. Centered on its “Event-Driven” architecture, the solution integrates data search, comprehension, analysis, and generation into a continuous task flow. Paired with ASRock Rack’s globally unique ZutaCore waterless heat dissipation technology, it ensures extreme operational reliability and safety during high-load knowledge modeling on-premise. This creates a highly governable and scalable environment that perfectly supports diverse application scenarios, including document analysis, operational decision support, and professional knowledge querying.
Futurenest’s core patented technology: GeniAltX AI
Futurenest’s core patented technology, GeniAltX AI, enables enterprises to build a complete pipeline from on-premise data integration and knowledge generation to AI task execution. The platform integrates three core modules: Xerno for enterprise knowledge search and analysis, Xtan for rapid AI assistant deployment, and Xparse, the foundational AI infrastructure supporting sovereign on-premise deployment.
This architecture allows enterprises to convert fragmented data into AI-ready knowledge assets and embed AI directly into operational workflows, forming a closed loop from decision support to execution. Futurenest plans to accelerate global expansion and deepen integration within the NVIDIA Inception ecosystem. According to Co-founder and Chairman Bob Hsu, the platform has been adopted across finance, manufacturing, and government sectors. The latest release of Xparse introduces governance, energy optimization, and concurrency capabilities that make high-performance on-premise AI deployment a practical reality.
Contact Info:
Name: Rachel Ching | Marketing Lead
Email: Send Email
Organization: Futurenest
Website: https://futurenest.ai/
Video URL: https://www.youtube.com/watch?v=cUKkCPz4WfA
Release ID: 89186351
If you detect any issues, problems, or errors in this press release content, kindly contact [email protected] to notify us (it is important to note that this email is the authorized channel for such matters, sending multiple emails to multiple addresses does not necessarily help expedite your request). We will respond and rectify the situation in the next 8 hours.
