Conference Presentation
Keynote by Alex Saroyan, Co-Founder & CEO of Netris | RAISE Summit 2026
- Netris plans to deploy approximately two to three new AI clusters monthly while aiming to manage a cumulative total of one million GPUs.
- The sector is projected to expand from hundreds to thousands of clusters as manual network engineering and in-house automation via Ansible or Python become unviable due to increasing complexity.
- Market transition is expected toward commercial software solutions offering hard multi-tenancy enforced at the hardware layer to secure resource sharing and separate maintenance from tenant data access.
- Rapid evolution in reference architectures, specifically GB300 and Vera Rubin, will drive the shift from SDN to Network Automation Abstraction and Multitenancy (NAM) as the new generation of networking.
- An abstraction layer is anticipated to standardize cloud networking requests across diverse technologies including GB200, InfiniBand, and Ethernet, insulating users from underlying hardware differences.
- Netris leverages a "learning reservoir" within its product algorithm to adapt to nuances in new architectures, while ecosystem partners including NVIDIA, VAST, Red Hat, and SpectroCloud are expected to continue integrating their solutions.
- Neocloud operators will require network reconfiguration capabilities to slice resources between different customers and tenants as they scale their services.