Leading portfolio chosen to deliver high performance networks for AI clusters that are easy to deploy, operate and scale
Juniper Networks, a leader in secure, AI-Native Networking, today announced that it has been selected by AI pioneer ionstream.ai to provide data center switching and automated operations for its AI managed services to enterprises. Following an extensive competitive evaluation process, ionstream.ai selected the Juniper AI Data Center solution to deliver high throughput, low latency connectivity for their GPU as a Service and AI workloads over Ethernet with simplified fabric management and troubleshooting to minimize the ongoing costs of managing an AI service.
Also Read: Secure with Simplicity: Why IT Teams Need Better Backup Processes
For organizations facing the rapid pace of AI transformation, ionstream.ai offers innovative business models including AI infrastructure as a service (IaaS) to deliver smooth cloud experiences to enterprises trying to navigate the complexities of AI. For those enterprises interested in building and deploying their own on-premises AI data centers, ionstream.ai also offers deployment and system integration services for these scenarios.
The Juniper AI data center solution is foundational to all ionstream.ai managed AI service offerings. As a key element of Juniper’s AI-Native Networking Platform, the AI data center solution is built on a spine-leaf architecture using AI-optimized 800G Juniper Networks QFX5240 switches. Managed through Juniper Apstra, the solution streamlines operations with key Ops4AI capabilities, such as intent-based networking and multivendor switch management, to maximize network utilization. It enhances GPU performance with RDMA over Converged Ethernet version 2 (ROCEv2) and uses Ethernet VPN (EVPN) to secure environments for multi-tenant environments, making it a reliable and efficient choice for AI-driven networks.
With Juniper’s full Networking for AI solution, customers and partners typically lower AI training Job Completion Times (JCTs), reduce latency during inferencing and increase GPU utilization while decreasing deployment times by up to 85 percent and reducing operations costs by up to 90 percent in some instances.
Supporting Quotes:
“In the rapidly evolving field of AI and ML, selecting the right GPU infrastructure and deciding whether to build, buy or rent it can significantly impact a company’s success. Juniper provides us with easy-to-deploy, scalable and AI-optimized networking so that we can deliver accessible, first-class GPU as a Service and AI transformation for our customers, however they choose to consume it. After careful evaluation of networking options, including Infiniband, we found that Ethernet’s superior availability of resources, performance capabilities, and scalability made it the clear choice for our infrastructure. This aligns with the industry trend we’re seeing among hyperscalers, who are increasingly adopting Ethernet solutions for their proven reliability and broader technical support ecosystem.”- Jeff Hinkle, CEO, ionstream.ai
Also Read: TrueData Introduces Low-Latency Identity API That Fits Into Any Data Workflow
“ionstream.ai understands that organizations require high-performance, economical AI infrastructure, but that they also need flexible consumption models. This is the ideal time for an enterprise to use high-performance data center solutions to execute new AI projects and reap the benefits of this technology without the need for in-house expertise or resources. Having been the first vendor to ship 800G switches, Juniper is excited to announce that we have already shipped over one hundred thousand 800GbE ports globally. This underlines how well our three key tenets for the solution, automation, performance and open flexibility, is resonating in the industry.”- Praveen Jain, Senior Vice President & General Manager, Data Center & AI, Juniper Networks.
[To share your insights with us as part of editorial or sponsored content, please write to psen@itechseries.com]
