INFRA
INFRA
INFRA
Lenovo Group Ltd. today introduced a collection of data center systems optimized to run artificial intelligence workloads.
The product additions are headlined by a compute appliance called the ThinkSystem SR680a V4. According to Lenovo, it combines nearly two dozen processors in an air-cooled chassis. Those chips enable the appliance to run inference workloads up to 11 times faster than the company’s previous-generation hardware.
Most of the SR680a V4’s AI processing capacity is provided by eight Blackwell B200 graphics cards from Nvidia Corp. Introduced last May, the B200 features 208 billion transistors made using a four-nanometer node. According to Nvidia, the chip can run trillion-parameter language models using 25% less power than previous-generation silicon.
The SR680a V4’s 8 graphics cards are supported by six of Intel Corp.’s Xeon 6 central processing units. Introduced in February, the Xeon 6 CPU series includes chips with up to 288 cores. It also ships with optimizations designed to speed up AI models. One of those optimizations is a feature called PCT that can boost some of a Xeon 6 chip’s cores when they run demanding workloads.
The SR680a V4’s chip roster is rounded out by eight Nvidia SuperNICs and the chipmaker’s BlueField-3 DPU. The processors offload some tasks from a server’s main processors to leave more computing capacity for AI workloads. The BlueField-3 offloads tasks related to cybersecurity, storage hardware management and networking, while the SuperNIC focuses on only the latter use case.
Lenovo debuted the system today alongside two other new hardware offerings. Both are based on machines from its existing ThinkSystem SR675 server lineup.
The first system combines SR675 V3 servers with Cisco Systems Inc.’s Nexus switch series. It can be equipped with up to eight of Nvidia’s RTX PRO 6000 Blackwell Server Edition graphics cards. Those chips are based on the same architecture as the B200, but aren’t focused solely on AI workloads. They include circuits called Tensor Ray Tracing Cores and Streaming Multiprocessors that are optimized for graphics rendering workloads.
Lenovo’s third new AI system is likewise based on its SR675 server series. It’s optimized to run IBM Corp.’s watsonx software suite, which helps companies build AI models and manage the data that they process.
The systems are rolling alongside four new Hybrid AI Advantage offerings. Those are product bundles that combine Lenovo hardware with third-party AI software. One of the new offerings uses computer vision models to help manufacturers detect product defects, while another is geared towards hospitality companies. The two other product bundles are optimized for workplace safety and retail use cases.
Lenovo is rolling out the products alongside a new suite of professional services. AI Adoption and Change Management Services, as the offering is called, promises to ease tasks such as training employees in using AI tools. Lenovo can also help companies ensure that their workers use those AI tools responsibly.
Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.
Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.