Between Overpriced and Inadequate,
There’s a Better Option.
Enterprise load balancers cost too much. Switch-based hashing doesn’t do enough. The Quantum Framework lets you build stateful load balancing that fits exactly what your application needs — at 40G to 400G on standard COTS hardware.
The Load Balancing Market Has Two Problems
The expensive problem: incumbent stateful solutions
The established load balancing vendors — the ones with the feature-complete platforms, the HA clustering, the application delivery controllers — build for the enterprise market. Their pricing reflects that. If you’re an OEM building a product, or an operator looking for a focused solution, you’re buying a platform built for global enterprises. You pay for SSL offload, application firewalling, global server load balancing, DNS services, and a management plane designed for large operations teams.
You need load balancing. You end up paying for a data center in a box.
The cheap problem: switch-based hashing
On the other end of the spectrum, modern high-speed switches offer ECMP hash-based load distribution at very low cost. It works until something fails. When a backend goes down, hash-based distribution reconverges by rehashing — and every connection that wasn’t mapped to the failed backend gets remapped too. In a high-availability architecture, that reconvergence event looks like a partial outage even when only one backend failed.
Hash-based distribution is not load balancing. It’s traffic spreading. The distinction matters the moment you need true HA.
The gap BlueDot fills
Stateful load balancing, purpose-built for your application’s exact requirements, at throughputs from 40G to 400G, on standard COTS hardware, without the platform pricing of incumbent vendors.
Purpose-Built for Builders and Operators
OEMs Building High-Performance ADC / Load Balancing Appliances
You’re building a load balancing product and you’re evaluating whether to license an incumbent platform or build your own. Licensing puts you in a cost position that makes your product uncompetitive. Building from scratch on raw VPP takes 18 months before your first feature ships. Quantum gives you a third option — a complete forwarding platform, a 90-day custom plugin, your product.
Operators Needing High-Throughput Stateful Distribution
You’re operating at 40G–400G and you need stateful session distribution with fast failure detection and reconvergence that doesn’t disrupt non-failed sessions. Hash-based ECMP isn’t good enough for your HA requirements. Enterprise ADCs are priced for someone else’s budget.
Carriers with Custom Distribution Requirements
Your load balancing requirement doesn’t fit a standard algorithm — subscriber-aware distribution, protocol-specific session pinning, custom health check logic, weighted distribution tied to real-time capacity signals. You need an algorithm that’s designed for your application, not adapted from a generic enterprise feature set.
Stateful. Custom Algorithm. Both Worlds.
The Quantum Framework’s plugin model is exactly suited to load balancing. We design a custom VPP plugin that implements your load balancing algorithm — whether that’s a standard approach like least-connections or weighted round-robin, or a completely custom distribution model built around your application’s specific requirements.
The Quantum Controller manages the backend pool at runtime — adding and removing backends, updating weights, responding to health check failures — without touching the forwarding plane. Sessions already established stay pinned to their backend. Only new sessions are redistributed. No reconvergence event. No collateral disruption.
Standard algorithms we can implement:
Custom capabilities available:
Failure Shouldn’t Affect More Than the Failed Backend
This is where hash-based approaches break down and where Quantum delivers.
When a backend fails in a Quantum load balancing deployment, the Quantum Controller detects the failure via health check and updates the backend pool in real time. Sessions established to healthy backends are unaffected — they stay pinned. Only sessions that were on the failed backend need to be redistributed.
The result is failure containment — a backend failure affects only the sessions on that backend, not the entire session table. For high-availability architectures where reconvergence events are operationally significant, this is the difference between a contained event and a visible service degradation.
40G to 400G. Your Hardware.
The Quantum Framework scales load balancing throughput by adding CPU cores — not by buying the next vendor appliance tier. A dual-socket COTS server gives you a different throughput profile than a single-socket edge platform. The plugin is the same. The algorithm is the same. The cost model scales linearly with hardware, not exponentially with vendor licensing tiers.
Where Quantum Sits in the Market
Tell Us What You’re Distributing
Every load balancing engagement starts with your application — the protocols, the session model, the failure requirements, the throughput target. From there we design the plugin and scope the engagement.