Scaling GPU-based Analytics Using NVMe Drives - Seguridad Mania.com - España y América Latina

Data is doubling every two years, driven by human and machine-generated data, such as the Internet of Things (IoT). The analysis of patterns and structure in this data and its transformation into actionable insights is driven by Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) workloads. IDC predicts that in the next three years over half the businesses will use analytic workloads. This has given rise to massively parallel GPU-based servers like the Nvidia DGX-1, IBM Power AC922 and HPE Apollo 6500 delivering massive compute power to run these machine learning frameworks. In order to accelerate the AI/ML training and operational cycles, storage systems that support these GPU-based applications must deliver ultra-low latency, massive ingest bandwidth, and heavy mixed random and sequential read/write I/O operations. Application architectures that use direct attached storage (DAS) or all-flash arrays (AFAs) see limited performance and complex data provisioning, mobility, and growth. Vexata addresses these issues with its VX-100FS Scalable NVMe File Array. Vexata is teaming up with leading GPU Compute vendors to develop reference architectures for these new data-driven applications. Hear from an expert about these reference architectures and how the Vexata VX-100FS, with its transformative VX-OS, is purpose-built to overcome these machine learning challenges. This webinar is targeted to those that want to leverage GPU for data analytics. He will cover: - Dealing with large training and inferencing data-sets by leveraging NVMe media - Using the Vexata VX-100FS to eliminate movement between data pipeline stages - Dealing with data security, protection, and efficiency concerns Learn how you can leverage the VX-100FS to deliver transformative performance with breakthrough economics at scale for your GPU-based applications.