Lenovo ships all-AMD AI systems

Lenovo Systems announced AI-centric systems using an all-AMD processor design, along with infrastructure for on-prem and Azure cloud.

At the top of the line is the ThinkSystem SR685a V3 8GPU server, designed for demanding AI workloads like large language models and generative AI. The server comes with 4th generation AMD EPYC “Genoa” CPUs and eight AMD Instinct MI300X GPUs, putting it slightly ahead of the hardware configuration in Frontier, the all-AMD supercomputer that topped the most recent Top 500 supercomputer list.

The company also announced the Lenovo ThinkAgile MX455 V3 Edge Premier Solution with AMD EPYC 8004 processors. The 8004 series, codename “Siena,” uses smaller, less power-hungry cores than standard EPYC and is meant mostly for edge deployments. But in this case, the ThinkAgile system is designed for turnkey integration with on-prem Azure Stack HCI and Azure cloud.

The systems are specifically tuned for the financial industry. For example, the ThinkSystem SR685a V3 is designed to help financial services with fraud detection and prevention and KYC (know your customer) initiatives. The system is also built for applications supporting algorithmic trading strategies, risk management (such as real-time transaction monitoring), credit issuance, wealth management and advisory services, streamlining and augmenting regulatory compliance, and forecasting, according to Lenovo.

The ThinkAgile MX455 V3 Edge Premier Solution is designed for remote use cases like retail, manufacturing, and healthcare. The Lenovo Open Cloud Automation (LOC-A) is designed for “near zero-touch” provisioning for fast deployment, centralized support, cloud-based fleet management and continuous testing and automated software updates, validated by Microsoft and Lenovo.

Lenovo and AMD also introduced a multi-node server designed for intensive transaction processing. The Lenovo ThinkSystem SD535 V3 is a 1S/1U half-width server node powered by a single 4th generation AMD EPYC processor and is engineered to maximize processing power for intense workloads. It starts with one node that can be expanded to up to four nodes within one chassis. Through custom cooling and power management, AMD claims a reduction in power consumption of up to 30% compared to standard 1U rack servers.

Lenovo also introduced Lenovo AI advisory and professional services designed to help businesses optimize their AI solutions. This includes AI Discover, which helps companies visualize and map their strategy and resources for AI adoption; AI Advisory professional service providing consultation and strategy planning; AI Fast Start to quickly build AI elements for customers; AI Deploy and Scale for deployment of tools and frameworks in its AI system; and AI Managed, which offers guidance for managing and optimizing the AI system.

As with other products, these solutions are available through Lenovo TruScale professional services, a consumption-based lease service.