GTC 2024: Lenovo and NVIDIA launch new hybrid AI solution

Today, the global AI event GTC (GPU Technology Conference) 2024 officially opened. Lenovo Group and NVIDIA announced their cooperation to launch a new hybrid artificial intelligence solution to help enterprises and cloud providers obtain artificial intelligence. The key accelerated computing power required for success in the intelligent era turns artificial intelligence from concept to reality.

These hybrid solutions are the culmination of a joint engineering collaboration to efficiently bring artificial intelligence to customer data where and when they need it most, whether in their pocket or in the cloud. This move will advance Lenovo Group’s vision of AI for All and provide support for the next generation of innovative architectures for large-scale generative artificial intelligence to quickly enter the market.

Lenovo hybrid solutions have been optimized to run NVIDIA AI Enterprise software for safe, supported and stable production-grade AI, and will now also provide developers with just-released NVIDIA microservices, including NVIDIA NIM and NeMo Retriever.

Today, industries around the world are looking to use artificial intelligence to analyze huge data sets, with power efficiency becoming a key factor. Enterprises are under tremendous pressure to undertake the energy consumption of these computationally intensive tasks.

Lenovo Group is a leader in enabling high-performance and high-power computing, launching innovative Lenovo Neptune liquid cooling technology many years ago. Not only does this technology earn Lenovo the No. 1 spot on the Green500 list, but its cutting-edge design is powered by NVIDIA GPUs, enabling faster computing and ensuring the device remains cool even in high temperatures and multi-GPU environments.

Lenovo ThinkSystem AI Server: Together with NVIDIA, rapidly improve generative AI inference and efficiency

In terms of efficiently enhancing artificial intelligence workloads at scale, Lenovo released an expansion of the ThinkSystem artificial intelligence product portfolio , an industry-changing milestone that includes two new powerful 8-way NVIDIA GPU systems designed to deliver massive computing power and unparalleled energy efficiency to accelerate the implementation of artificial intelligence. Designed specifically for generative artificial intelligence, natural language processing (NLP) and large language model (LLM) development, providing support for the NVIDIA HGX artificial intelligence computing platform to enter the market.

The new Lenovo ThinkSystem SR780a V3 is a 5U system using Lenovo Neptune liquid cooling technology with a power usage efficiency (PUE) of up to 1.1. By using direct water cooling of the CPU and GPU and NVIDIA NVSwitch technology, the system can maintain maximum performance without hitting any thermal limits.

For more than ten years, Lenovo’s industry-leading Neptune direct water cooling solution has recycled warm water to cool data center systems, enabling customers to reduce power consumption by up to 40% and improve heat dissipation performance by 3.5 times. More efficient than traditional air cooling systems.

As industry metrics used to determine data center energy efficiency, PUE and power consumption are among the most talked about sustainability methods, according to the Uptime Institute's 2021 Annual Global Data Center Survey. Because liquid cooling offers a more energy-efficient alternative than air, systems can achieve higher sustained performance while consuming less energy. It also allows the ThinkSystem SR780a to be installed in a dense 5U package, helping to save valuable data center space.

In addition, the new Lenovo ThinkSystem SR680a V3 is an air-cooled dual-socket system that provides maximum acceleration for complex artificial intelligence with Intel processors and NVIDIA GPUs. The highly accelerated system delivers massive computing power and uses industry-standard 19-inch server racks, allowing for dense hardware configurations that maximize efficiency while avoiding excessive space usage.

Lenovo also showed off the new PG8A0N, the ultimate 1U node designed specifically for AI.

Customers can run NVIDIA AI Enterprise, an end-to-end cloud-native software platform for developing and deploying production-grade AI applications, on Lenovo’s portfolio of NVIDIA certified systems. Customers can run the NVIDIA NIM inference microservice included in NVIDIA AI Enterprise on Lenovo’s enterprise-grade infrastructure to perform high-performance inference on popular AI models from NVIDIA and its ecosystem.

The product portfolio includes Lenovo's XClarity system management software, which improves the efficiency of infrastructure deployment for enterprises, and Lenovo's Intelligent Computing Platform (LiCO), which simplifies the use of cluster computing resources for artificial intelligence models Development and training and HPC workloads. The portfolio also supports 4th and 5th generation Intel Xeon Scalable processors and provides thermal headroom for future higher-power GPUs.

Co-engineered NVIDIA MGX Solutions: Bringing custom AI, NVIDIA Omniverse and high-performance computing to CSPs and enterprises

From workstation to cloud support, Lenovo develops for NVIDIA OVX and NVIDIA Omniverse Leading provider of platform design, engineering and support. To help enterprises quickly build and deploy a wide range of purpose-built AI, HPC and Omniverse applications, Lenovo has partnered with NVIDIA to build accelerated models faster using the NVIDIA MGX modular reference design.

Leveraging these designs, cloud service providers can cost-effectively deliver accelerated compute for AI and Omniverse workloads at scale, receiving custom models faster. These systems will empower scientists and researchers to solve some of the world's most challenging problems by accelerating complex AI and HPC applications that process terabytes of data.

Lenovo’s new products designed using NVIDIA MGX modular reference servers include:

The new Lenovo HG630N - MGX 1U—an open-standard server that uses Lenovo Neptune direct liquid cooling, supporting the highest performance GPUs while reducing power consumption.

The new Lenovo HG650N - MGX 2U - a highly modular, GPU-optimized system that is air-cooled and supports industry-standard racks.

New Lenovo HG660X V3 - MGX 4U - This system supports up to eight 600W NVIDIA GPUs in an air-cooled environment, making it ideal for NVIDIA Omniverse and AI workloads. Lenovo is the design partner of NVIDIA MGX 4U.

New Lenovo HR650N - MGX 2U - A high-performance Arm CPU server with multi-core and the flexibility of storage and front-end IO.

Lenovo ThinkStation with NVIDIA RTX: Generative AI at your fingertips

Lenovo enhances its data science workstations by providing up to 4 NVIDIA RTX 6000 Ada Generation GPUs for large-scale AI Train, fine-tune, infer, and accelerate graphics-intensive workloads. This increases AI developer productivity through automated workflows:

New Lenovo workstations with NVIDIA AI Workbench can be used to develop and deploy powerful AI solutions for inference, large-scale simulation, and demanding industries and scientific workflow. NVIDIA AI Workbench makes generative AI and machine learning development easy for all developers.

Featuring dual CPUs and up to 4 NVIDIA RTX professional graphics cards, the Lenovo ThinkStation PX is the most powerful workstation on the market today. Lenovo’s portfolio of AI-ready ThinkStation desktop and ThinkPad mobile workstations offers the broadest and most powerful workstations for AI to handle the most intensive real-time AI solution development.

Lenovo Professional Services: Quickly start the AI ​​journey and accelerate industry progress

Lenovo real-time solutions powered by NVIDIA are helping customers accelerate the realization of smarter solutions through artificial intelligence and Lenovo infrastructure. future and unlock new potential in various industries. In the retail industry, these solutions optimize foot traffic distribution and manage inventory more effectively in real time by analyzing customer movement and behavior patterns. In manufacturing, Lenovo and NVIDIA's joint solution enhances safety between workers and machines by improving assembly line inspection processes. In terms of urban planning, Lenovo's integrated edge AI solutions are helping planners use data to optimize the use efficiency of space, infrastructure and assets, effectively manage traffic flow and reduce energy consumption.

Lenovo’s newly announced Center of Excellence (COE) for AI services brings together business consultants, data scientists and AI-optimized infrastructure-as-a-service to maximize outcomes and satisfy customers on the road to AI. needs during the trip.

Based on Lenovo’s proven AI readiness experience in security, people, processes and technology, the AI ​​COE provides expertise and capabilities that enable enterprises to quickly implement powerful, responsible and sustainable artificial intelligence. Lenovo’s new artificial intelligence professional services include:

• New AI discovery tools – helping customers explore the “infinite possibilities” of artificial intelligence. By holding interactive workshops and assessments, looking at the entire ecosystem, and formulating an AI strategy, Lenovo has built a blueprint for AI success.

• New quick-start generative AI service with NVIDIA - Leverage powerful data insights and achieve competitive advantage with generative AI. Lenovo provides full-stack solutions to support the entire product lifecycle, as well as services to implement, adopt and scale generative AI solutions.

• New TruScale GenAI as a Service – Delivers AI capabilities through an as-a-service model to increase flexibility, scalability and predictability. With Lenovo TruScale, infrastructure can always configure, deploy and manage AI innovations just right.

• Enhanced AI Professional Services – Help clients accelerate their AI transformation by providing business consultants, data scientists and AI-optimized infrastructure-as-a-service to ensure seamless utilization of sustainable AI.

The above is the detailed content of GTC 2024: Lenovo and NVIDIA launch new hybrid AI solution. For more information, please follow other related articles on the PHP Chinese website!