Juniper Networks has introduced the industry’s first AI-Native Networking Platform, meticulously crafted to harness the power of artificial intelligence (AI) for ensuring optimal end-to-end operator and end-user experiences. This innovative platform, backed by seven years of insights and data science development, guarantees reliable, measurable, and secure connections for every device, user, application, and asset.
This significant announcement comes weeks after Hewlett Packard Enterprise (HPE) announced the acquisition of Juniper for a substantial $14 billion.
Juniper’s AI-Native Networking Platform unifies all campus, branch, and data center networking solutions through a shared AI engine and Marvis Virtual Network Assistant (VNA). This integration enables end-to-end AI for IT Operations (AIOps), facilitating deep insights, automated troubleshooting, and seamless networking assurance. The platform empowers IT teams to shift their focus from maintaining basic network connectivity to delivering exceptional and secure end-to-end experiences for various stakeholders, including students, staff, patients, guests, customers, and employees.
AI for Networking: Advancements with Marvis VNA
Juniper’s new AI-Native Networking Platform introduces two enhancements to Marvis, the industry’s only AI-Native Virtual Network Assistant driven by Mist AI. The new capabilities empower Juniper customers and partners with more automation and insight:
Marvis Minis: This is the only AI-Native Networking Digital Experience Twin, utilizing Mist AI to simulate user connections proactively. It instantly validates network configurations, identifies problems without user presence, and employs unsupervised machine learning to learn network configurations. Marvis Minis integrates directly with the network, providing continuous feedback to the Mist AI engine for optimal AIOps responses.
Marvis VNA for the Data Center: This first and only AI-Native VNA for the data center offers unparalleled insight throughout the entire data center lifecycle across any vendor’s hardware. The Marvis conversational interface allows IT teams to pose direct queries and gain seamless insight into the data center’s documentation and knowledge base using GenAI.
Networking for AI: High Performance Networks for AI Training
Juniper is enhancing its AI Data Center solution, making it a fast and flexible choice for deploying efficient AI training and inference clusters with simplicity for limited IT resources. The solution includes a spine-leaf data center architecture supported by QFX switches and PTX routers managed by Juniper Apstra. This multivendor solution streamlines data center fabric management, automation, and assurance. With intent-based operations and the Marvis VNA for data center, Juniper simplifies AI Data Center networking complexities, offering customers increased productivity with fewer IT resources and flexibility while avoiding vendor lock-in.
Expanding on the strengths of Juniper’s AI Data Center architecture, the company is unveiling a series of new products and capabilities to enhance speed, efficiency, and scalability. These include:
- Juniper Apstra’s expanded capabilities for faster and more efficient processing of AI/ML traffic over Ethernet.
- New Express 5 silicon-based PTX routers and line cards promising industry-leading performance and energy-efficient sustainability.
- A new QFX switch with 2X the capacity of the previous generation, using the most advanced Broadcom Tomahawk 5 silicon for 800GE.
The new PTX and QFX platforms support high 800GE port density and essential AI infrastructure protocols, ensuring power-efficient and scalable AI Data Center networking.