Tingyu Technology

What Tingyu Technology does

Tingyu Technology (庭宇科技) is a Beijing-headquartered, privately held edge cloud and AI compute infrastructure operator founded in 2019. The company designs, builds, and operates a three-tier national compute network spanning a national hub at the Wulanchabu (Ulanqab) green-energy base, regional data centers in the Beijing-Tianjin-Hebei area, and 1,500+ distributed edge nodes covering 2,000+ county-level cities, delivering sub-20ms intra-provincial latency for real-time interactive workloads. Core technology assets include a proprietary elastic fusion distributed edge computing platform with heterogeneous computing pooling (95%+ server compatibility across GPU SKUs), a self-developed compute-power coordination scheduling system, 400GbE RDMA networking, cold-plate liquid cooling (PUE < 1.15), and an OpenAI-compatible multi-model API routing engine called SilvaMux (千木).

The company monetizes through four primary revenue streams: (1) usage-based GPU cloud compute rental ranging from single GPU to 10,000+ GPU clusters across bare metal, containerized, and virtualized delivery modes; (2) AI FDE (Forward Deployed Engineering) professional services where on-site engineering teams deliver private AI deployments; (3) end-to-end Smart Compute Base solutions covering site selection, power, cooling, network, hardware, and software for greenfield intelligent computing centers; and (4) per-token SilvaMux inference API services. Customer segments span real-time entertainment and gaming (Tencent Games, miHoYo, NetEase Games, iQIYI, Tencent Video, BestTV), short video and internet platforms (ByteDance, Kuaishou, Douyin, Baidu), automotive (Li Auto, GAC XLab, Chery), AI/smart home (Xiaodu), and energy/government (State Grid). The go-to-market combines direct enterprise sales, channel integrations with major cloud platforms (Baidu Cloud, China Mobile Cloud, Volcano Engine, Tencent Cloud), self-service API provisioning, and embedded engineering services.

The business is structured around 110 employees, 80% of whom are in R&D, and reports three consecutive years of profitability alongside 100%+ YoY revenue growth. A Series B round of several hundred million RMB closed in June 2024 (co-led by Guohai Investment and Fosun RZ Capital) with extension rounds in 2025; total disclosed funding is in the hundreds of millions RMB range. Recent strategic emphasis has shifted from pure infrastructure IaaS toward higher-value AI applications, including the 2025 launch of the Lybic AI Agent platform and the SilvaMux token service, both of which seek to convert GPU capacity into recurring token-based consumption.

Name

Tingyu Technology

Legal name

北京庭宇科技有限公司 (Beijing Tingyu Technology Co., Ltd.)

Website

https://tingyutech.com

Company type

Private

Founded year

2019

Operating status

Operating

Short description

Tingyu Technology is a Beijing-based edge cloud and AI compute infrastructure operator (founded 2019) serving 55+ top-tier Chinese enterprises across gaming, short video, automotive, and government with GPU clusters, a national three-tier compute network, multi-model API aggregation (SilvaMux), and forward-deployed AI engineering services.

Ownership category

akta.pro rank

Product category

AI Cloud Infrastructure

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Computer Systems Design and Related Services (54151), Computer Facilities Management Services (541513)

SIC

Services-Computer Integrated Systems Design (7373), Services-Computer Processing & Data Preparation (7374), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)

akta.pro secondary industries

Edge Networking & SD-WAN Integration (SASE/SD-WAN Edge On-ramps) (HDABAGAJ), Cloud Data Platforms & Analytics Migration (BPAEACAI), Network, Infrastructure & Data Center Integration (BPAEAAAK)

Keywords

• GPU cloud computing
• Edge cloud services
• AI compute infrastructure
• Real-time interactive cloud
• Intelligent computing base

Headquarters

HQ city

Beijing

HQ country

China

HQ region

Asia

Offices4 records

Markets served

GTM type

B2B

Offering type

Services

Cost components

Infrastructure, Technology or R&D, Personnel, Operations, Marketing or Sales, Supply Chain

Revenue model

1. GPU Cloud Compute Rental: Usage-based billing for GPU compute resources from single GPU to 10,000+ GPU clusters. Customers pay based on monthly usage volume or per-license model. Supports bare metal exclusive, containerized sharing, and virtualized GPU partitioning.
2. Enterprise AI Solutions (AI FDE): Forward-deployed engineering services where team embeds at customer site to discover AI opportunities, design architecture, build systems, and deploy to production. Includes private deployment, model customization, and system integration.
3. Smart Compute Base Solutions: End-to-end intelligent computing center delivery and operation, including site selection, power systems, cooling architecture, network topology, hardware clusters, and software platform.
4. Token Services (SilvaMux Platform): Multi-model inference API aggregation platform charging per token consumed. Provides precise token accounting and billing per project and API key.

Pricing tiers

Go-to-market motion3 records

Distribution channels4 records

Marketing channels5 records

Core offering

Tingyu Technology operates as an independent third-party AI compute infrastructure operator providing GPU cloud computing rental (from single GPU to 10,000+ GPU clusters), a unified multi-model inference API aggregation platform (SilvaMux/千木), an OS Agent deployment platform (Lybic/灵臂), and end-to-end intelligent computing base construction services. The company runs a three-tier computing power network spanning national hubs, regional centers, and 1,500+ edge nodes across China, integrating green energy via a proprietary compute-power coordination scheduling system.

Product overview

Tingyu Technology is an AI computing infrastructure operator offering a comprehensive product and solution portfolio. The core offerings include GPU Cloud Computing (GPU 云算力) for elastic GPU rental from single GPU to 10,000+ clusters, SilvaMux (千木推理平台) as the unified Token service and multi-model inference API platform, and Lybic (灵臂) as the OS Agent intelligent platform for enterprise deployment. The solution suite encompasses Intelligent Computing Solutions (智算解决方案) for end-to-end AI computing base construction and AI FDE (Forward Deployed Engineering) for on-site AI implementation services. The company operates a three-tier computing power network infrastructure (算力网络) connecting national computing hubs, regional centers, and 1,500+ distributed edge nodes. The products work together to provide seamless AI computing infrastructure from API-level access to complete intelligent computing base delivery.

Differentiator

Problem solved

Functional benefit

Brands

• Silvamux (千木): Unified token service platform providing multi-model API aggregation for enterprises and developers, with intelligent routing for optimal cost-performance balance
• Lybic (OS Agent)

Products and services

• GPU Cloud Computing (GPU 云算力) Elastic GPU computing power rental service ranging from single GPU to 10,000+ GPU clusters. Supports bare metal exclusive, containerized shared, and virtualized partitioned deployment modes, covering large model training, online inference, and cloud rendering scenarios. Target customers are AI companies, large model developers, cloud rendering customers, and enterprises needing on-demand GPU resources.
• SilvaMux (千木推理平台) Token Service Platform Unified Token service and inference API aggregation platform providing access to multiple mainstream large language models through an OpenAI-compatible protocol with intelligent routing, precise Token metering, and enterprise security features. Serves AI developers, enterprises, and end users needing multi-model LLM access.
• Intelligent Computing Solutions (智算解决方案) End-to-end intelligent computing infrastructure solutions covering site selection planning, power systems, cooling architecture, network topology, hardware clusters, and software platforms for building AI computing bases from scratch. For enterprises and government customers needing full AI computing center delivery.
• AI FDE (Forward Deployed Engineering) On-site AI implementation service where engineering teams embed in customer scenarios to discover, design, build, and deploy AI private solutions end-to-end, from scene identification to production launch. Includes private deployment, model customization, and system integration. Target customers are large enterprises and government/industry leaders in finance, healthcare, and manufacturing.
• Lybic (灵臂) AI Agent Platform OS Agent intelligent platform enabling deployment of vertical domain AI Agents across multiple industries. Each Agent continuously consumes Tokens, forming high-density load matrices as unified commercialization output for the Agent matrix. For enterprise customers deploying AI agents in business processes.
• AIGC Workbench (AIGC工作台) AI content generation workstation providing multi-modal capabilities and pipeline-style workflows to assist users in production from text to images to video, significantly improving content creation efficiency. For content creators and enterprise marketing teams.

Quantifiable outcome

• 70%+ customer repurchase rate
• +7 more outcomes

Named customers18 records

Segments4 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

AI capability10 records

Feature6 records

Partnerships

Three partnerships are on record, tiered core and minor.

• BestTV (百视通)coreStrategic or Co-development Partner · 18 November 2025Tingyu provides edge compute infrastructure for BestTV's cloud gaming platform. Jointly launched 'BestTV Cloud Gaming' product commercially deployed on Oriental Cable, Changhong TV and other major platforms, streaming 3A titles like Black Myth: Wukong to home TV screens. Tingyu CEO Zhu Bin attended BestTV's 20th anniversary ceremony and co-unveiled the cloud gaming strategic partnership
• Microsoft (微软)minorStrategic or Co-development Partner · 18 November 2025Participated as game vendor representative in BestTV cloud gaming strategic partnership unveiling alongside Tingyu and BestTV
• Tencent Cloud (腾讯云)coreStrategic or Co-development Partner · 16 September 2025Strategic cooperation announced at 2025 Tencent Global Digital Ecosystem Conference. Tencent Cloud will provide high-performance, low-latency cloud services leveraging its global infrastructure to help Tingyu improve user experience for massive users. Both parties will collaborate on edge cloud solutions and jointly promote edge cloud innovation, leveraging Tencent's cloud computing and AI LLM capabilities with Tingyu's edge cloud domain expertise

Scale indicators10 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• CoreWeave: US-based specialized GPU cloud provider purpose-built for AI training and inference at massive scale. Closest direct peer to Tingyu on the specialized-GPU-cloud-for-AI business model, though CoreWeave is more focused on hyperscale training workloads while Tingyu emphasizes edge and real-time interactive.
• Lambda Labs: US-based GPU cloud and deep learning workstation company offering on-demand NVIDIA GPU compute for AI training and inference. Comparable to Tingyu as a specialist GPU cloud provider with both self-service and enterprise offerings.
• Together AI: AI-focused cloud platform providing GPU compute for training, fine-tuning, and inference of large models. Direct overlap with Tingyu's SilvaMux inference API aggregation and GPU cloud compute offerings, with similar emphasis on serving AI-native customers.
• RunPod: GPU cloud platform offering on-demand and serverless GPU compute for AI developers, with both pod-based and serverless deployment. Comparable self-service GPU cloud model to Tingyu's product-led growth motion.

Emerging players

• Bitdeer Technologies: Cryptocurrency mining-turned-AI-cloud provider with significant data center footprint and GPU capacity, particularly in Asia and North America. Comparable in transitioning from crypto/blockchain infrastructure to AI compute, though more scale and geographic diversification than Tingyu.

Broad incumbents

• Volcano Engine (火山引擎): ByteDance's enterprise cloud platform offering GPU compute, AI inference, and MaaS services. Competes with Tingyu in the same GPU cloud market while also being a distribution channel partner, creating both partnership and competitive overlap.
• Alibaba Cloud (阿里云): China's largest cloud provider offering GPU compute instances, PAI for AI/ML workloads, and extensive edge node coverage. Represents the dominant incumbent that Tingyu must compete against or partner with across all segments.
• Tencent Cloud (腾讯云): Major Chinese hyperscale cloud with GPU compute, edge computing, and gaming-optimized services. Strategic partner of Tingyu (Sept 2025) while also operating directly in the same GPU cloud and edge computing markets.
• Huawei Cloud (华为云): Major Chinese cloud provider with strong AI compute stack (Ascend GPUs, ModelArts) and significant edge node footprint. Competes in the same enterprise AI compute and edge cloud markets as Tingyu, with much greater capital scale and domestic chip supply.
• Baidu Cloud (百度智能云): Chinese cloud provider with strong AI capabilities (ERNIE models, AI compute stack) and one of Tingyu's key distribution partners. Competes in the same enterprise AI compute and inference market while also channeling demand to Tingyu.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

Compliance3 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles3 records

Funding overview

Funding rounds6 records

Investors12 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Tingyu Technology