skip to Main Content

100G DIY load balancer solution
benefits OEM vendor

Case Study


A network OEM vendor needed a flexible data delivery solution to distribute the traffic load on their DPI appliance and across numerous external monitoring appliances.


In accommodating this need, Napatech expanded the key accelerator feature set with a number of new advanced in-line features. This enabled efficient traffic load distribution at CPU level within the DPI appliance and across external appliances.


The solution helped the OEM vendor to use their valuable compute resources on revenue-generating SW-processing instead of packet-forwarding – while providing them freedom to customize and scale as desired.

Client challenge
As a network OEM vendor was developing a DPI security monitoring solution for a global client, they needed to optimize resource usage. To enable the heavy traffic analysis, they had to distribute the traffic load from 100G links between numerous appliances.

The OEM vendor wanted a fully scalable solution that would give them the needed agility to tune their system to changing industry needs. They were not looking to invest in a costly full-fledged platform with an abundance of unnecessary capabilities – but rather to integrate dedicated load distribution capabilities to supplement their own solution without adding undesired proprietary limitations. Napatech was brought on to engineer the needed technology.

The solution was based on two 100G Napatech FPGA SmartNICs, guaranteeing zero packet loss and nanosecond time precision. The solution both needed to address the load distribution at CPU level to enhance the performance of the DPI load distribution appliance, but also needed to facilitate efficient distribution across external monitoring appliances. To accommodate these needs, a number of advanced in-line features were added to the Napatech FPGA SmartNICs.

In-line support: Full parallel processing – no bottlenecks
To enable intelligent traffic load distribution within the DPI appliance, the following features were developed:

  • 128 Rx/Tx streams in hardware, enabling parallel processing with up to 128 CPU cores
  • Full sustained throughput 100G Rx and 100G Tx for any packet size
  • Customizable hash based load distribution

Advanced protocol API
To facilitate the traffic redirection and load distribution to the external monitoring appliances, an advanced API was engineered for prepending each individual packet with meta-data. This made it possible to apply VLAN IDs to retransmitted packets and redirect them via VLAN switching to the appliances dedicated to processing the traffic further.

The solution helped the OEM vendor to use their valuable compute resources on revenue-generating SW-processing instead of packet-forwarding – while providing them freedom to customize and scale as desired.

Napatech FPGA SmartNICs

Napatech FPGA SmartNICs capture data from networks at high speed and high volume using patented packet capture technology, enabling real-time insight into network traf c.

With a portfolio that scales from 1G to 100G, we provide more ef cient data delivery through such features as zero packet loss and nanosecond time precision.

Back To Top