Juniper Networks has unveiled a new solution for neocloud, traditional SPs and other AI cloud providers deploying and managing GPUaaS (GPU-as-a-Service) and AIaaS (AI-as-a-Service) offerings.
"Managed AI services, such as GPUaaS and AIaaS, have been growing rapidly. To capture the demand, neocloud, traditional SPs and other AI cloud providers need to move fast and deliver exceptional value. Juniper is excited to deliver a purpose-built solution for these providers, expediting their journey towards offering unique and compelling services with the performance, security, automation and open flexibility their customers demand," said Praveen Jain, senior vice president & general manager of data Centre & AI, Juniper Networks.
Key benefits
Juniper's latest Networks for AI solution claims to accelerate deployment and simplify Day 0/1/2+ operations by leveraging a combination of intent-based networking and Mist AI. Integrated with Red Hat OpenShift, it achieves even more automation benefits in Kubernetes environments.
It also deploys Juniper's Zero Trust DC Security portfolio, EVPN VXLAN capabilities within Juniper switches, and Juniper's SRX 4700 next-generation firewall to protect the AI infrastructure, models and confidential data from threats.
Further, Juniper's Ops4AI Lab allows managed providers to validate their AI models before deployment across various networking, compute and storage platforms and configurations. Juniper Validated Designs (JVDs) also allow multi-vendor AI blueprints, Weka and VAST storage to expedite GPUaaS and AIaaS deployment times.
Juniper's solution claims to stand out with its unique multi-vendor approach to DC fabric management and automation. It also offers distinctive capabilities to optimise AI workload performance over Ethernet, a unique approach allowing providers to use open and proven technologies and products, avoiding vendor lock-in.
