Lenovo Launches AI Servers with AMD MI300X GPUs

Print Friendly, PDF & Email

Lenovo today announced AI-centric infrastructure systems in collaboration with AMD, including the ThinkSystem SR685a V3 8GPU server (pictured here), designed for demanding AI workloads, including genAI and large language models. The server, powered by 4th Generation AMD EPYC Genoa CPUs and eight new AMD Instinct MI300X GPUs, are designed for on-prem and public AI cloud service providers.

The company also announced the Lenovo ThinkAgile MX455 V3 Edge Premier Solution with AMD EPYC 8004 processors. Designed for turnkey integration with on-prem and Azure cloud, Lenovo described it as an AI-optimized platform “with the best power efficiency of any Azure Stack HCI Solution.”

Lenovo said it complements these solutions with Lenovo TruScale professional services, designed to provide flexibility, scale and support to simplify the on-ramping of AI workloads as-a-service.

“Lenovo is working to accelerate insights from data by delivering new AI solutions for use across industries delivering a significant positive impact on the everyday operations of our customers,” said Kirk Skaugen, President of Lenovo ISG. “From advancing financial service capabilities, upgrading the retail experience, or improving the efficiency of our cities, our hybrid approach enables businesses with AI-ready and AI-optimized infrastructure, taking AI from concept to reality and empowering businesses to efficiently deploy powerful, scalable and right-sized AI solutions that drive innovation, digitalization and productivity.”

In financial services, the ThinkSystem SR685a V3 is designed to help with fraud detection and prevention and KYC (know your customer) initiatives. The system is also built for applications supporting algorithmic trading strategies, risk management (real-time transaction monitoring and value-at-risk), credit issuance (loan analysis and approval recommendations), wealth management and advisory services, streamlining and augmenting regulatory compliance, and forecasting.  Features include:

  • Interconnected with AMD Infinity Fabric, providing 1.5TB of high-bandwidth (HBM3) memory capacity and up to 1TB/s peak aggregate theoretical GPU I/O bandwidth performance for training and deploying AI models.
  • With thermal headroom via air-cooling, the Lenovo SR685a V3 can sustain maximum performance of high-power CPUs and GPUs.
  • Support for NVIDIA’s latest HGX (H100/H200/B100) GPUs and drop-in support for the next generation of AMD CPU.


Lenovo’s ThinkAgile MX455 V3 Edge Premier Solution, suited for retail, manufacturing and healthcare, includes:

  • Lenovo Open Cloud Automation (LOC-A) designed for “near zero-touch” provisioning to deploy faster and manage from in the cloud using Azure Arc enabled architecture.
  • Single point of contact support and centralized, cloud-based fleet management.
  • Continuous testing and automated software updates, validated by Microsoft and Lenovo, for security, reliability and reduced downtime.

“We hear from our customers every day that they need the power of Microsoft Azure to extend into their diverse, real-world environments – from manufacturing floors and retail spaces to fast-food outlets and other distributed locations. Our collaboration with Lenovo allows us to deliver just enough Azure to these critical edge sites while simplifying deployment and life cycle management,” said Douglas Phillips, Corporate Vice President, Azure Edge + Platform. “The ThinkAgile MX455 V3 Edge Premier Solution gives our shared customers a seamless experience from the cloud to the edge, wherever their data lives. This new solution and the power of our adaptive cloud approach enables businesses to unify teams, streamline operations, and harness the power of cloud-native and AI technologies so they can focus on transforming their business, rather than managing infrastructure.”

Lenovo and AMD also unveiled 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 Gen AMD EPYC processor and is engineered to maximize processing power and thermal efficiency for workloads including cloud computing and virtualization at scale, big data analytic, high-performance computing and real-time e-commerce transactions for businesses of all sizes.  Features include:

  • Support of CPU technology venders in a 2U 4 node offering – designed AMD or Intel architecture in the same 2u chassis.
  • Scale agility allowing configuration of 2, 3 or 4 nodes installed within one chassis.
  • Simplify large scale automation, orchestration, and deployment with Lenovo XClarity system management software.
  • Unified power and cooling implementation delivering up to 30 percent reduction in power consumption vs. standard 1U rack server, according to Lenovo.

“Together with Lenovo, we are accelerating AI transformation for the enterprise in multiple manners – helping guide customers through this monumental shift in computing, while providing robust end-to-end solutions at scale,” said Forrest Norrod, executive vice president and general manager, Data Center Solutions Group, AMD. “We have a proven track record of bringing amazing solutions, jointly, to the market and are excited to add AMD Instinct MI300X to Lenovo’s portfolio, helping enterprises to rapidly adopt AI.”

The company also introduced Lenovo AI Advisory and Professional Services designed to help businesses navigate the AI landscape and find optimized AI solutions. This includes:

  1. AI Discover: Lenovo said it tailors AI Discover engagements to help customers visualize and maptheir strategy and resources for AI adoption. Lenovo assesses the organization’s AI readiness across security, people, technology, and process – using proven methodology – with AI recommendations.
  2. AI Advisory: Lenovo’s service professional recommend security postures, change management plans, technology architectures and adoption plan for AI based on customer strategy and objectives from pocket to cloud, and co-develops the roadmap.
  3. AI Fast Start: Lenovo builds AI elements to show how customer outcomes can be achieved, including proof of concept or MVP (minimum viable product) in either Lenovo or customer labs.
  4. AI Deploy and Scale: Lenovo deploys the tools and frameworks to implement a secure and scalable AI system or fleet. Customers may choose to deploy via Lenovo TruScale as-a-service, which can reduce upfront costs to help deploy genAI and scale.
  5. AI Managed: Lenovo offers guidance to manage and optimize the AI system and cultivate an ecosystem of AI Innovators so that AI use cases can mature.

Speak Your Mind