GigaSMART Appliance: Architected to Keep Pace With Hybrid AI Scale and Complexity

As enterprise AI evolves toward Hybrid AI, organizations will increasingly bring together frontier and open-weight models, plus specialized and private models. They will match models and services to the requirements of specific AI workloads and workflows. As those models interact with more applications and agents across distributed hybrid cloud infrastructure, they create more data in motion and more network traffic for security and IT teams to understand.

The new GigaSMART^® appliance, GigaVUE-GSA110, extends the Gigamon^® Deep Observability Pipeline with the processing capacity and density to keep pace. It scales GigaSMART processing as network traffic grows. It also helps organizations avoid a proportional increase in physical infrastructure and downstream telemetry volume.

Doing that requires more than processing more packets. It requires an architecture that can add processing density where needed, extract valuable application context from high-volume traffic, and deliver the right data to tools and platforms that can use it.

Adding Processing Capacity Where It Is Needed

The GigaVUE-GSA110 is designed to add GigaSMART processing capacity without requiring organizations to redesign the visibility fabric around it. The appliance includes two independent GigaSMART engines. Each provides dedicated processing resources for one or more GigaSMART applications.

Its disaggregated architecture connects to existing GigaVUE^® TA Series and GigaVUE HC Series deployments through two 400Gbps bidirectional ports. This provides up to 800Gbps of input capacity. The appliance can then apply GigaSMART functions such as De-duplication and Application Metadata Intelligence (AMI). It can return optimized packet traffic or export metadata and flow records to downstream tools.

That separation matters. As network traffic volumes grow, organizations need the ability to add processing where demand is increasing without forcing every function onto the same appliance or requiring infrastructure to grow at the same rate as traffic. The GigaVUE-GSA110 delivers additional processing density in a compact 1RU form factor. It helps the Gigamon Deep Observability Pipeline evolve as infrastructure requirements change.

Different Data, Different Destinations

Scaling processing capacity is only part of the challenge. Not every downstream tool needs the same data or consumes it in the same way.

Packet-based security and performance tools may require optimized packet traffic. Security and observability platforms may instead benefit from application metadata. This metadata provides context about what is communicating, how applications are behaving, and where dependencies exist.

The GigaVUE-GSA110 supports both paths at once.

Network traffic can enter through GigaVUE TA or HC nodes and be forwarded to the GigaSMART appliance for processing. Metadata can then egress directly to the tools that require it. Optimized packet traffic can return through TA or HC nodes for delivery to packet-based tools. The visibility fabric is centrally managed through GigaVUE-FM. GSA processing is automatically recognized and integrated, which removes the need to manually shift traffic between the two devices.

This architecture provides flexibility in how network data is processed and delivered. The Gigamon Deep Observability Pipeline can deliver optimized packet traffic to packet-based tools. It can also deliver application-aware metadata to downstream platforms that require it.

Turning High-Volume Network Traffic Into Application Intelligence

Gigamon Application Metadata Intelligence (AMI) is central to making that model more efficient.

AMI derives application and protocol context directly from network traffic and delivers it as trusted, high-fidelity network-derived telemetry. That context can include application identity, round-trip time (RTT), and certificate information. It can also show relationships and dependencies. This supports cybersecurity, performance assurance, and compliance use cases. It can also support certificate visibility, shadow AI visibility, and application dependency mapping.

The GigaVUE-GSA110 supports AMI across up to 400Gbps of network traffic. But throughput alone is not the full story.

As traffic volumes grow, organizations need richer context without requiring downstream telemetry volumes to grow at the same rate. Approximately 100Gbps of packet traffic can yield 1Gbps of application metadata. This makes valuable network-derived context available to security and observability tools without requiring equivalent packet volumes downstream.

Gigamon is also improving metadata efficiency. New HTTPS metadata optimization consolidates two metadata records into a single unified record for HTTPS traffic. It reduces metadata records for those flows by approximately 50 percent while preserving valuable application context.

The objective is to extract the context that matters from growing volumes of data in motion and efficiently deliver it to downstream tools.

Delivering Network-Derived Telemetry to Security, Observability and AI Tools

Generating application metadata only creates value if organizations can put it to work across the tools and platforms they already rely on.

The GigaVUE-GSA110 supports metadata delivery to security information and event management (SIEM) systems, observability platforms, and data lakes. It also supports data engines and data buses, plus telemetry pipelines. New software capabilities support JavaScript Object Notation (JSON). They also support Common Event Format (CEF) and IP Flow Information Export (IPFIX). This provides flexibility in how organizations integrate network-derived telemetry into existing workflows.

AMI can export close to 6,000 metadata elements across applications. Organizations can select application and protocol attributes based on their specific use cases. This can reduce tool overload while enriching SIEM and observability workflows with trusted network-derived telemetry.

That flexibility becomes increasingly important as consumers of network-derived telemetry expand beyond traditional security and network tools. Data platforms and AI workflows need context too. Making that context available in flexible formats helps organizations derive more value from the network data already moving across their infrastructure.

Extending the Pipeline for What Comes Next

Hybrid AI is increasing the scale and complexity of the infrastructure organizations need to secure and manage. As applications and workloads become more interconnected with models, agents, and services, the visibility infrastructure beneath them must evolve as well.

The GigaVUE-GSA110 extends the Gigamon Deep Observability Pipeline with the processing capacity and density to support that growth. It also helps organizations manage the volume of telemetry delivered downstream. Combined with AMI enhancements and flexible metadata delivery, high-scale packet optimization transforms growing data in motion into trusted application context.

Keeping pace requires more than processing more traffic. It requires increased processing capacity, the right context, and efficient delivery to security and observability tools. It also supports data and AI tools that can put the information to work.

See How the GigaSMART Appliance Scales Application Metadata Intelligence for High-Speed Environments

CONTINUE THE DISCUSSION

People are talking about this in the Gigamon Community’s AI Exchange group.

Share your thoughts today