Suanova Technology Limited ("Suanova") is a wholly-owned subsidiary of Yeebo (International Holdings) Limited (0259.HK), a Hong Kong listed company. As a pioneer of domestic cloud computing and computing power, Suanova is dedicated to providing enterprises with professional, autonomous, and innovative one-stop computing, cloud, and application services. Suanova has established branches in Hong Kong and Shanghai to deliver better localized services to clients.
About UsSuanova Services
Powered by advanced domestic chips, we build the "Harvest Series" domestic computing power rental service, offering enterprise clients flexible and diverse AI computing rental solutions.
Powered by the advanced cloud computing technology and rich cloud service experience of "Suanova Cloud", we provide enterprises with secure, stable, and elastic cloud solutions.
Backed by localized service capabilities, deep technical expertise, and strong computing power supply chain integration, we provide clients with comprehensive, professional one-stop cloud, computing, and application integration services.
Leveraging advanced domestic computing technology and high-speed interconnects, we deliver highly integrated computing facilities, providing enterprise clients with domestic high-density computing solutions.
As an investor in and Hong Kong regional distributor of advanced domestic chips, we provide enterprise clients with high-performance domestic chips.
Suanova Services
The "Harvest Series" domestic AI computing rental service offers enterprise clients flexible and diverse computing rental options, combined with pre-configured models and industry-specific solutions, enabling one-click access for enterprise users.
Smart Management Platform:
Provides computing resources with intelligent management and operations, greatly improving resource utilization.
Robust Infrastructure:
Provides the essential infrastructure for AI development, including elastic bare-metal clusters and high-performance storage.
One-Stop AI Development:
Simplifies the entire AI development workflow with built-in mainstream frameworks, letting developers focus on innovation.
Large Model Development:
Pre-configured with a rich set of large models, supporting multiple fine-tuning methods and one-click deployment.
Industry-Specific Large Models:
Provides customized industry large models, continuously optimized based on industry knowledge and data.
Compatibility & Adaptation:
Fully adapted to domestic AI chips, enabling seamless migration of open-source models.
Suanova Services
"Suanova Cloud" provides enterprise clients with a neutral and secure cloud computing platform, offering comprehensive digital transformation services. With 31 availability zones across 25 regions worldwide — covering China, Southeast Asia, Europe, North America, South America, and Africa — it delivers a full range of cloud solutions including public cloud, private cloud, hybrid cloud, and multi-cloud management to support enterprise innovation.
30+ availability zones worldwide, delivering consistent high-quality cloud computing services.
Outstanding performance with Intel Cascadelake CPUs, self-developed network enhancement 2.0, disk Binlog technology, and full-NVMe RSSD cloud disks — delivering up to 1.2 million IOPS and 10 million PPS network performance.
Reliable service commitment with 99.95% availability, kernel hot-patching support, and online upgrades for cloud hosts without rebooting.
Quick creation within minutes — simplified host provisioning steps bring cloud hosts from creation to accessibility in minutes.
Suanova Services
We provide enterprise clients with one-stop, high-performance cloud, computing, and application integration solutions. By integrating diverse hardware resources and optimizing software configurations, we ensure clients can efficiently deploy and manage complex computing tasks. Our services cover the full spectrum of needs — from infrastructure setup and AI model development to big data processing — supporting enterprises in achieving both innovation and efficiency gains during digital transformation.
Localized Service:
Local-language service teams provide support with seamless communication.
Customized Service:
Tailored service support that adapts to client needs.
Comprehensive Integration:
Integration spans chips, software platforms, and model applications, providing one-stop cloud, computing, and application support.
Suanova Services
Shanghai Cube is a full-stack, hardware-software integrated AI data-center-grade product. It addresses the core strategic computing power needs of China's AI sector and, aligned with the latest trends in computing, brings together multiple core technology companies in Shanghai to jointly develop this benchmark computing product.
Full-Stack Optimization:
Shanghai Cube uses a full-stack optimization approach at the system level to increase computing density, meeting the enormous computing demands of current AI development.
Technical Collaboration:
The project brings together the combined strength of core chip technology companies, server hardware technology companies, platform software technology companies, and computing operators, backed by deep technical expertise.
Product Form:
Highly integrated computing equipment and highly integrated computing data centers.
Product Vision:
Shanghai Cube aims to create a reference design architecture for AI users, delivering leading AI data center products that can be rapidly deployed and delivered, with complete end-to-end service and support.
Suanova Services
As an investor in and Hong Kong regional distributor of high-performance general-purpose domestic GPU chips, we provide enterprise clients with chips for AI inference, AI training, and graphics rendering — meeting the demand for "high energy efficiency" and "high versatility" computing power.
Autonomous IP:
Fully self-developed GPU IP and instruction set, plus a fully proprietary architecture, providing greater flexibility for chip customization and optimization.
High-Performance Computing:
Delivers high-performance computing with mixed-precision support, meeting the high computing demands of AI training and inference.
High Energy Efficiency:
An innovative reconfigurable GPU architecture optimizes the core computing energy-efficiency ratio, effectively breaking through the efficiency bottleneck of traditional GPUs.
Mainstream Ecosystem Compatibility:
The MXMACA® software stack is compatible with mainstream GPU ecosystems, letting developers seamlessly migrate existing applications and lowering development and migration costs.
Multi-Card Interconnect:
Supports multi-card interconnect technology, enhancing full-interconnect capability for multi-GPU single-node systems and providing an excellent solution for building high-density computing.
Broad Application Scenarios:
Suitable for a wide range of scenarios, including AI large model training and inference, content generation, recommendation systems, speech recognition, speech synthesis, and image segmentation and detection.
Powerful Parallel Computing:
Powerful parallel computing makes it a core computing chip for fields such as scientific computing and AI, providing strong computing support across new infrastructure industries.
Security & Reliability:
Designed with security and reliability in mind, providing users with a more secure and dependable computing solution.
