Lenovo's Artificial Intelligence Workstations
By To meet this demand, Lenovo has created a new class of AI workstations developed precisely for these areas. Based on Nvidia Quadro RTXTM GPUs, Lenovo's P900, P500 and P300 series workstations can count on up to 96GB of GPU memory, perfect not only for managing, but also for accelerating AI, Machine & Deep Learning workflows such as data preparation, model training and visualization.
The new Lenovo AI workstations
With a configuration suitable for loads of artificial intelligence, and an Nvidia CUDA-X software stack GPU-accelerated AI, Lenovo's workstations stand as solutions that can guarantee the highest levels of performance, compatibility, and reliability. Available in single and dual GPU configurations to conform to a variety of data science workflows, dedicated series aim to maximize productivity, reduce time to obtain insights, and cut the cost of IT projects .A great example is the ThinkStation P920, with two Intel Xeon SP 8C + up to 4.4 GHz, two Nvidia Quadro RTX 8000, 384 GB DDR4 memory, 1 TB storage, Ubuntu 20.04 system and Nvidia AI software Data Science. But also the ThinkStation P520, with Intel Xeon W Class 6C +, one or two Nvidia Quadro RTX 8000, 64GB or 128FB DDR4 memory, 512GB or 1TB storage, Ubuntu 20.04 and Nvidia Data Science.
The "augmented" world of Ducati Motors
Among customers who already use Lenovo's workstation configurations for complex, AI-driven workloads, is Ducati Motors. As Konstantin Kostenarov, Ducati Chief Technology Officer explains: "Our HPC environment is the engine that drives the development and design of road and racing bikes"."We use advanced aerodynamic and fluid dynamic modeling tools to calculate how a particular design or feature of the bike will react under different riding conditions. We do this not only for the superbikes on the track, but also for our road models, so that all riders who choose Ducati can have an exceptional riding experience. "
Recently, Ducati has discovered that the its HPC infrastructure no longer offered the performance, reliability or flexibility its engineers needed to test new designs within tight deadlines. Additionally, the company's legacy cluster did not offer built-in data visualization capabilities, adding resistance to test workflows.
For this, Ducati worked closely with Lenovo to design and implement an HPC cluster new generation, based on Lenovo ThinkSystem SD530, SR630 and SR650 servers, equipped with latest generation Intel Xeon scalable processors and Nvidia Tesla M10 and V100 GPUs.
Stefano Rendina, IT Manager of Ducati Corse, explains that: “Not only were we impressed that Lenovo's solution allows us to distribute some of our workloads across a CPU, but also how inherently flexible it is. This is fundamental, because we plan to evolve our HPC capabilities with hybrid cloud computing, which will allow us to benefit from elastic and economical scalability in the years to come. "
At the same time as the HPC project, Ducati has two new data centers were founded: one at the headquarters in Borgo Panigale and a second disaster recovery (DR) site in Ponte San Pietro. Together, the data centers contain 20 ThinkSystem SR630 servers and four ThinkSystem SR550 servers, configured as active-active clusters for seamless failover in the event of a site outage.
Watch Free Webinar
Lenovo and Ducati will hold a webinar dedicated to how the recognized Italian brand exploits the potential of Lenovo's workstations to carry out innovative projects. Ducati CTO Kostantin Kostenarov will talk about how the racing department has managed to optimize mechanical, aerodynamic and fluid dynamics research and analysis with Lenovo, Intel and Microsoft solutions, obtaining a competitive advantage in testing and developing next-generation motorcycles.To register for free and have the opportunity to ask questions to the Ducati CTO, just register on this page.
REGISTER FOR FREE TO THE LENOVO WEBINAR TO FIND OUT HOW THE DUCATI CTO EXPLOITES IT ARTIFICIAL INTELLIGENCE TO IMPROVE THE PRODUCTION PROCESS OF THE MOTORCYCLES.
