Storage

Innovative AI-a-a-S Solution Uses Fast Micron NVMe Storage with Excelero NVMesh Software

By Jason Echols - 2019-07-25

InstaDeep™, a global innovator delivering artificial intelligence (AI) products and solutions for the enterprise, is an African AI startup success story. Its goal to bring AI/Machine Learning workloads within reach of more organizations, large and small, created one of the world’s first AI-as-a-Service offerings. The innovative AI-a-a-S solution helps organizations gain the benefits of AI without investing to run their own AI stacks.

InstaDeep wanted speed and low latency for its service platform. “Finding a storage infrastructure that would scale modularly and was highly efficient for AI and ML workflows is no small challenge,” explained Amine Kerkeni, Head of AI Product at InstaDeep. “Our clients simply will not achieve the performance they need if an infrastructure starves the GPUs with slow storage or wastes time copying data to and from systems.”

Excelero

The right technology to serve its customers, maximize utilization, and reduce overall costs was a highly efficient data center infrastructure built from a scalable pool of Micron NVMe™ high-performance solid-state drives (SSDs) linked with the NVMe-over-Fabrics™ (NVMe-oF) protocol. The other key ingredient: NVMesh® software by Excelero®, which offers software-defined block storage solutions for web-scale applications.

Using NVMesh for NVMe-oF, InstaDeep’s GPU-optimized servers could access remote Micron SSDs as if they were local flash – with full IOPs and bandwidth capabilities. Excelero’s NVMesh enables its customers to maximize GPU utilization with connectivity and low-latency (5µs), delivering the high IOPs of local NVMe in a distributed and linearly scalable architecture.

Excelero NVMesh and Micron NVMe SSDs let InstaDeep deliver huge bandwidth and huge IOPs in a very small and cost-effective storage form factor. Get speed and latency with a small footprint. “The only storage that is fast enough to keep up with these GPUs is local NVMe flash, due to the high competition for valuable PCIe connectivity amongst GPUs, networking and storage,” explained Yaniv Romem, Excelero’s CTO.

Read the full case study: Micron NVMe SSDs and Excelero NVMesh Shared Storage Speed AI Innovation for InstaDeep. Learn more at micron.com/AI about how Micron storage and memory can help you be successful in your next AI/ML project.

Stay up to date with Micron by following us on Twitter @MicronStorage and connecting with us on LinkedIn.

Jason Echols

Jason Echols

Jason Echols brings more than 15 years of experience in data center and cloud to Micron where he leads the Technical Marketing organization. Prior to Micron, Jason led product and channel marketing at a network and security startup.

+