Gigamon adds GigaVUE-GSA110 appliance for hybrid AI visibility

Gigamon has launched the GigaVUE-GSA110, a 1RU network appliance that extends its Deep Observability Pipeline with dedicated GigaSMART processing capacity. The Santa Clara-based vendor says the hardware supports Application Metadata Intelligence (AMI) processing across up to 400 Gbps of network traffic, with optimised packet delivery up to 800 Gbps, and is available now alongside enhancements to AMI software.

The appliance is designed to slot alongside existing GigaVUE TA Series and HC Series deployments in a disaggregated architecture, allowing organisations to add processing capacity at specific points in distributed environments rather than replacing existing infrastructure. Native JSON export and support for CEF and IPFIX formats means telemetry can flow into security information and event management (SIEM) platforms, data lakes, Kafka pipelines, and AI-enabled observability tools.

The product case

Gigamon frames the launch primarily around the shift to what it terms Hybrid AI: enterprise architectures that blend frontier, open-weight, and private models across public cloud, private cloud, and on-premises infrastructure. The company argues that as AI agents and services multiply across this infrastructure, security and network operations teams need a consistent, network-derived view of what is communicating with what, including unsanctioned or "shadow AI" applications that employees may be using outside approved procurement channels.

AMI helps identify and classify generative AI and large language model application traffic, mapping relationships, protocol behaviour, and certificate activity. A further software enhancement consolidates two HTTPS metadata records into a single unified record, which Gigamon says reduces metadata volume for those flows by approximately 50 per cent while preserving application context.

Gareth Maclachlan, chief operating officer at Gigamon, said: "As network and security operations become more AI-enabled, they need consistent, trusted network context to analyse activity and make informed decisions."

Market context

Network observability and security telemetry is a competitive segment with a growing roster of vendors, including Cisco (via its Thousand Eyes and AppDynamics portfolios), Broadcom's former Symantec enterprise division, ExtraHop, and a number of well-funded startups targeting cloud-native environments. Gigamon differentiates on packet-level visibility rather than log or agent-only approaches, a positioning that appeals to large enterprises and government agencies with complex, hybrid infrastructure. The company cites a customer base of over 4,000 organisations, including 83 of the Fortune 100.

The broader market tailwind is real. A Gartner report cited in the release projects that by 2028, more than half of enterprises will use public cloud selectively for experimental workloads while anchoring sovereignty-sensitive AI workloads on private or localised infrastructure. That bifurcation creates precisely the kind of heterogeneous, multi-perimeter environment where network-level visibility becomes harder to maintain with agent-based or cloud-native tools alone.

Regulatory read-across

Shadow AI governance is increasingly in scope for compliance teams operating under frameworks such as the EU AI Act, which requires organisations to maintain records of AI systems in use and assess risks for those classified as high-risk. The NIS2 Directive similarly demands that operators of essential services demonstrate network monitoring and incident detection capabilities. Tools that can identify unsanctioned AI application traffic in real time sit at the intersection of these obligations, giving Gigamon a credible regulatory angle to work with enterprise buyers in regulated sectors including financial services, healthcare IT and critical national infrastructure. Whether the GigaVUE-GSA110's throughput credentials translate into named customer wins will be the cleaner test of commercial traction.