GTC China - NVIDIA today announced that China's leading original equipment manufacturers (OEMs) -- including Huawei, Inspur and Lenovo -- are using the NVIDIA® HGX reference architecture to offer Volta architecture-based accelerated systems for hyperscale data centers.
Through the NVIDIA HGX Partner Program, NVIDIA is providing each OEM with early access to the NVIDIA HGX reference architecture for data centers, NVIDIA GPU computing technologies, and design guidelines. HGX is the same data center design used in Microsoft's Project Olympus initiative, Facebook's Big Basin systems and NVIDIA DGX-1™ AI supercomputers.
Using HGX as a starter "recipe," OEM and original design manufacturer (ODM) partners can work with NVIDIA to more quickly design and bring to market a wide range of qualified GPU-accelerated AI systems for hyperscale data centers to meet the industry's growing demand for AI cloud computing.
New HGX server designs coming to market -- with eight NVIDIA Tesla® V100 GPU accelerators in a hybrid cube mesh with NVIDIA NVLink™ interconnect technology -- include Lenovo HG690X and HG695X, Inspur 2U-8GPUAGX-2 and Huawei G-series heterogenous servers.
With GPUs based on the NVIDIA Volta architecture offering three times the performance of its predecessor, manufacturers can meet market demand with new products based on the latest NVIDIA technology.
"As companies increasingly harness the capabilities of artificial intelligence, demand continues to grow for accelerated computing in the data center," said Ian Buck, general manager of Accelerated Computing at NVIDIA. "Our new Volta-based HGX design sets the standard for lightning-fast, energy-efficient data centers that can support the most demanding AI training and inference requirements."
Highly configurable to meet workload needs, HGX can easily combine GPUs and CPUs in a number of ways for high performance computing, deep learning training and inferencing.
The standard HGX design architecture includes eight NVIDIA Tesla GPU accelerators in the SXM2 form factor and connected in a cube mesh using NVIDIA NVLink high-speed interconnects and optimized PCIe topologies. With a modular design, HGX enclosures are suited for deployment in existing data center racks across the globe, using hyperscale CPU nodes as needed.
Industry Support for Volta HGX Platform
"NVIDIA is the world's leading artificial intelligence and accelerated computing technology provider. Inspur has been focusing on the research and development of deep learning and AI computing systems for years; in May this year, Inspur and NVIDIA jointly released an HGX-based system equipped with the latest NVIDIA Tesla V100 GPU and NVLink high-speed interconnect technology, which is designed to provide maximum throughput with higher power consumption efficiency for high performance computing in scientific research and engineering, taking AI computing to the next level."
-- Leijun Hu, vice president of Inspur Group
"NVIDIA HGX architecture provides exceptional capabilities and energy efficiency. Combined with Huawei's strength in computing and connectivity, we can provide an outstanding computing solution with AI capabilities to growing numbers of enterprises."
-- Wu Zhan, vice president of IT Server Product Line at Huawei
"New AI workloads are creating strong demand for high-performance and flexible data center architectures. NVIDIA's modular HGX design offers best-in-class performance for these new AI workloads. Volta HGX designs will help us meet our customers' needs while reducing their energy requirements."
-- Paul Ju, vice president and general manager of Global Hyperscale Segment at Lenovo
NVIDIA also announced today that Alibaba, Baidu and Tencent are incorporating new Volta architecture-based NVIDIA Tesla V100 GPU accelerators into their data centers and cloud-service infrastructures.