Inspur is driving high performance and sustainability with vigorous collaborations with Intel® technologies to deliver solutions for new generations of advanced applications.
Inspur Information, a leading IT infrastructure solutions provider, announced support for the Intel Artic Sound-M GPU (ATS-M) in its servers, in addition to showcasing a broad array of Intel-powered technology solutions that are empowering breakthroughs and advances in innovation from AI to edge computing at Intel Innovation 2022. The event marks the gathering of software and hardware developers across the tech industry to learn about Intel’s latest technologies and tools to create open ecosystems, collaborate on innovative new computing solutions, participate in hands-on labs and tutorials, and network with the community.
Inspur Information Announces Integration with Intel’s Arctic Sound-M GPU for Advanced Performance in Two GPU Configurations
Designed for virtual desktop infrastructure (VDI), video streaming and transcoding, and AI inferencing, ATS-M comes in two configurations: a full-height 150W option for maximum peak performance and a low-profile, dual-GPU 75W option for high-density use cases. At Intel Innovation, the Inspur NF5280M6 2U server supporting 4x ATS-M 150W GPUs and NF5468M6 4U server supporting up to 16x ATS-M 75W GPUs were both showcased. Leveraging ATS-M, Inspur servers can deliver leading transcoding and compute performance at up to 500 TOPS (trillion operations per second) to support visual inference, media analytics, and real-time AAA gaming experiences.
Latest Data Center Solutions Powered by Intel Acceleration Technologies
In partnership with Habana, Inspur is developing AI servers based on open computing standards that can be integrated with Gaudi2® deep learning training processors. The Gaudi2 mezzanine card is utilized in Inspur’s next-gen OAM platform to accelerate AI development on open architecture to support a variety of AI workloads.
Also featured at Intel Innovation is Inspur’s F26A Intel® Agilex™ I-Series CXL-based reconfigurable FPGA. CXL is a cache and memory coherent processor interconnect for high-speed, low-latency, and efficient performance between the CPU and FPGA, enabling better use of the accelerator’s onboard memory. The accelerator and CPU can share memory data, address in unity and work cooperatively. Compared with traditional PCIe DMA, the average data access latency between the CPU and accelerator is reduced by 80%, and the memory capacity can be expanded by 2 times.
Inspur Solutions and Innovations Empowering Breakthroughs
The innovations from Inspur Information continue with major advances in liquid cooling, edge servers, etc.
Inspur Information recently announced plans to make cold-plate liquid cooling available for its full product portfolio, and demonstrated at Intel Innovation its end-to-end data center liquid cooling capabilities. A cooling demo of the NF5260FM6 shows how this general-purpose server is optimized for hybrid air-liquid cooling: a liquid-cooling mode for high-power consumption modules, such as processors, and an air-cooling mode for low-power consumption modules, such as drives and I/O cards.
The latest in Inspur’s edge infrastructure portfolio is the EIS800 Edge Microserver. Its modular design and environmental adaptability provide three different deployment modes —indoor, outdoor, and micro-center— allowing for rapid deployment across various edge scenarios. It delivers a range of computing power from 21 to 130 TOPS to support near-end AI inference applications.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!
