Is a Nanoconda's Managed Environment Right for You? A Look at Our Approach

Many customers ask us whether our managed environment or standalone software is the right fit for their trading desk.

The traditional Standaone Software model is straightforward: buy the software, procure a server, arrange colocation and connectivity, and operate the infrastructure yourself.

Nanoconda Managed is a new way to achieve the same result, at a fraction of the cost and operational complexity. Your trading application still runs on dedicated, isolated CPU cores on the same physical machine as the execution stack, with the same low-latency APIs and no additional network hop between your strategy and the gateway.

There is no server to purchase, colocation to arrange, or infrastructure stack to build and maintain. The environment is already deployed and managed, while you retain the low-latency architecture and access expected from your own HFT stack.

 Cost, Time, and Complexity

Building a CME-certified DMA stack in-house typically takes 6 - 12 months, before accounting for the engineering headcount to maintain it. The shared environment reduces all three:

• Cost. "Colocated Essentials" starts at $2,100/month, and Colocated Professional at $3,200/month - flat rate, no order routing fees, no volume tiers. A full year on the top managed tier costs a fraction of a single year of in-house build cost.
• Onboarding time. Go live in days, not months. There's no hardware to provision, no certification to pursue, no network path to stand up - the infrastructure is already built and already certified.
• Complexity. One vendor, one subscription, one support line. Software, colocation, exchange certification, and 24/6 operations come as a single package instead of several vendor relationships to manage.

None of that comes at the cost of latency or performance. Every customer gets a dedicated and isolated environment, still reads market data from the same low-latency shared-memory path, and the order-to-wire numbers don't change based on how you deploy.

Everything, in One Place

Every Nanoconda tier runs the same certified software stack: iLink 3.0 execution gateway, MDP 3.0 market data feed handler, in-memory pre-trade risk, and the full trading GUI. The build is identical regardless of how you deploy it - only the infrastructure around it changes.

1. Colocated Essentials. One dedicated CPU core at CME Aurora colocation, GUI access, and fully managed operation. The shortest path to running a strategy live, with access to lower-cost CME market data licensing options.
2. Colocated Professional. Three dedicated CPU cores at CME Aurora colocation with full SSH access. Run your own trading applications alongside the Nanoconda stack with direct, low-latency access to execution and market data.

Your Data Stays Yours

Privacy is one of the most common questions about a shared environment. Each customer runs in an isolated environment - other customers cannot access your files, processes, memory, or applications.

Nanoconda is contractually prohibited from accessing your files. We access your environment only when necessary to resolve infrastructure or connectivity issues.

Your source code never needs to leave your hands. Build your C++ application locally and deploy only the compiled binary or a plugin.

Dedicated Performance

Every environment runs on fully dedicated, isolated CPU cores: bare metal, no kernel interrupts, no host processes, and no CPU contention.

Market data is distributed through a shared-memory ring buffer. Every customer reads the same feed directly from memory, so adding customers does not add another processing or network hop to your market data path. It is the same patent-pending distribution layer used across our deployments.

Order entry uses a dedicated virtual interface with TCPDirect over a 10 Gb connection to CME.

Trade With Less Capital

A $25,000 account and colocated HFT market making rarely go together. With Nanoconda, they can.

In the managed environment, clients cannot modify or bypass pre-trade risk controls, while the FCM retains independent control of risk limits and the kill switch. This structure can allow DMA accounts starting around $25,000-$50,000 instead of the $500,000 commonly required.

Clients are already market making at these account sizes using low-latency CME MSGW sessions.

Standalone Software: The Familiar Way

Owning the infrastructure is the what most trading firms are familiar with. Nanoconda's DMA Software Suite provides the same certified stack - iLink 3, MDP 3, pre-trade risk, GUI, simulator, and FCM integration - installed on your own hardware.

Standalone makes sense when:

• You already operate your own trading infrastructure. Nanoconda can run directly within your existing colocated environment and integrate with your internal systems.
• You want your own hardware and more compute. Procure the server configuration you need and dedicate as much CPU, memory, and hardware capacity as you want to your trading applications.
• Your compliance policy prohibits cloud or externally managed infrastructure. Your trading systems and data remain entirely on infrastructure you own and control.
• You use custom hardware or networking. FPGAs, specialized NICs, or custom network topologies may require full control of the underlying server and network.

With standalone software, you manage the hardware, colocation, networking, and ongoing maintenance. The Nanoconda software and execution path remain the same.

Managed Environment vs. Standalone Software

How Managed Latency Stays Low

Nanoconda's patent-pending architecture keeps your application, market data, and dedicated execution gateway on the same physical machine.

Each customer runs on dedicated CPU cores isolated from the kernel, host processes, and other customers. Each customer also has a dedicated execution gateway on isolated cores and dedicated TCPDirect virtual interfaces on the Solarflare NICs.

Market data is published directly into shared memory, with no per-customer queue, network hop, or additional processing layer. Customer applications consume it directly from memory.

Adding more customers does not add any processing or queuing latency to your path.

This architecture keeps the latency benefits of a dedicated HFT stack while sharing the infrastructure costs that do not affect the latency path.

Pick Your Tier

Colocated Essentials — 1 dedicated core with plugin deployment and GUI management through the Algo Control API. Same low-latency execution path, with lower infrastructure and market data costs.

Colocated Professional — 3 dedicated cores with full SSH access. Use the environment like your own colocated machine: deploy your applications, install additional software, and manage your trading processes directly.

Colocated Elite — 7 dedicated cores with full SSH access. The same unrestricted environment as Professional, with additional dedicated compute for more strategies and heavier workloads.

The Takeaway

Standalone software remains the established path, and it's still the right one for firms with existing colocation, custom hardware, or compliance requirements that call for full isolation - the DMA Software Suite runs the identical certified stack on infrastructure you control.

The shared environment is a different way to get the same result: dedicated cores, flat-latency shared memory, protected data, and a fraction of the usual capital requirement, without owning any of the infrastructure behind it.