La convergence des workloads d‘IA et de HPC
Découvrez comment la convergence des charges de travail IA et HPC exécutées sur une infrastructure HPC d'architecture Intel® peut accélérer l'innovation. Cela commence avec l'architecture de calcul intensif de référence commune d'Intel : l'association du matériel haute performance et de logiciels optimisés qui peuvent prendre en charge simultanément l'analyse, la simulation et la modélisation, ...l'apprentissage machine, le deep learning et d'autres charges de travail impliquant des calculs complexes et beaucoup de données. Une infrastructure de calcul intensif (HPC) équipée de processeurs Intel® Xeon® Scalable et d'une architecture Intel® Omni-Path (Intel® OPA) offre les performances de calcul évolutives dont ont besoin les applications d'analyse et d'intelligence artificielle.
Artificial intelligence, AI, has the potential of transforming our lives, from enabling breakthrough scientific discoveries, personalized medical diagnosis and treatments, more-accurate storm forecasting, advanced fraud detection, autonomous driving, and more.
A common misconception is that AI workloads must run on separate, highly specialized hardware, which means you have to purchase, configure, manage, and support a new, disparate infrastructure and software stack. But what if you could run your AI workloads without those inefficient, costly efforts? You can by running your workloads on your familiar Intel-based HPC infrastructure. And that same HPC infrastructure is ideal for your big-data analytics workloads as well.
Intel is working with leaders across those ecosystems to create a common HPC infrastructure foundation optimized to run all of those diverse HPC, big-data, and AI workloads. And that improves resource utilization and cost efficiency while making it much faster and easier to explore new AI opportunities.
It also enables data from all your workloads to be located in a single, always-available data pool. So moving data in and out of your infrastructure isn't required.
It starts with Intel's common HPC reference architecture, a combination of high-performance hardware with optimized software that can simultaneously support analytics, simulation and modeling, machine learning, deep learning, and other compute and data-intensive workloads. An HPC infrastructure featuring Intel Xeon scalable processors and an Intel Omni-Path architecture fabric delivers the highly-scalable compute performance AI and analytics applications need.
You'll also be ready to take advantage of Intel's ongoing open-source contributions, many tools to optimize software and investments in leadership in HPC, AI, and analytics innovation, including current and future innovations like Intel Optane SSDs, Intel FPGAs, Intel persistent memory, and more. The result is an infrastructure that is highly scalable today and ready for tomorrow.
Imagine your existing known HPC infrastructure running all your game-changing workloads, delivering unprecedented insight, discovery, and competitive advantage. Find out more about how a converged Intel-based HPC infrastructure can accelerate your innovation.