About a-team Marketing Services
The knowledge platform for the financial technology industry
The knowledge platform for the financial technology industry

A-Team Insight Blogs

Exegy Unveils Nexus Market Data Platform to Tackle Latency, Scale and Cost Trade-Offs

Subscribe to our newsletter

Exegy has launched Nexus, a new market data platform designed to streamline infrastructure for high-volume, low-latency trading environments. The FPGA-based system replaces traditional, server-heavy feed handlers with a combination of managed appliances and network interface cards that deliver microsecond-level performance and reduce datacentre footprint by over 40%.

“Across the buy side and sell side, market data has traditionally been processed using two common approaches: ticker plants that centralise processing and distribute data across servers, and embedded software libraries that run alongside applications on the same server,” observes Laurent de Barry, VP, Ultra-Low Latency Trading Products at Exegy, in conversation with TradingTech Insight. “Both methods come with significant drawbacks, i.e. higher latency, limited scalability, or high CPU consumption.  Neither is well suited to today’s environment of record-breaking data volumes and constrained data centre capacity.”

De Barry explains how Nexus takes a different approach. “What’s unique about Nexus is that it rethinks the entire architecture. We’ve separated the workload: feed processing is handled centrally by FPGA-enabled appliances that publish raw market data without merging feeds. On the client side, a Nexus SmartNIC, also FPGA-enabled, subscribes to those feeds and performs aggregation, conflation, and shaping based on the application’s needs. This creates a full end-to-end FPGA and Layer 1 data path, enabling extremely low latency with minimal jitter. We’ve also implemented a new PCIe shared memory interface that allows the FPGA to push data directly into application memory structures, without polling. This means the application server needs only one CPU core for market data tasks. The rest is available for trading logic.”

Exegy reports that Nexus has delivered notable cost savings in early deployments. A U.S. equities broker saw server count fall from 32 to 22, saving $1.4 million annually per site. Further reductions were achieved when paired with Nexus SmartNICs, dropping the server count to 17 and delivering close to $2 million in yearly savings.

Initial use cases have focused on equities and futures trading, but the company plans to expand support to other asset classes and features, including advanced order handling for agency brokers. Nexus integrates with Exegy’s execution solutions – SREX and nxAccess – providing a single platform for market data normalisation, distribution, and tick-to-trade execution.

“With Nexus, we’re not just accelerating market data; we’re aligning it with what applications actually need to do, especially in latency-sensitive environments,” says de Barry. “For example, a pricing model may only need top-of-book data, while a Smart Order Router (SOR) requires full depth. The application controls how the data is consumed, and the SmartNIC ensures it’s delivered efficiently, eliminating bottlenecks. And because Nexus integrates with Exegy’s broader FPGA tick-to-trade stack, it supports everything from normalised market data delivery to ultra-low-latency execution across multiple venues. For both sell-side and buy-side firms, it offers a scalable way to reduce infrastructure complexity and deploy high-performance trading systems more efficiently.”

Beyond performance improvements, Nexus is intended to simplify the deployment of large-scale FPGA environments by delivering raw and normalised market data directly to FPGAs. This supports ultra-fast order routing to exchanges and remote locations via wireless networks.

“What makes Nexus possible is the combination of FPGA technology from Exegy and Enyx, and the software feed handler expertise from Vela and Exegy,” note de Barry. “The ability to support both hardware and software feed handlers is essential, because not every market justifies FPGA-level latency. We can start clients with a software-based solution, and if a venue becomes latency-sensitive for their strategy, they can seamlessly upgrade to an FPGA-based handler without changing their API or control interface. This dual approach also improves system resilience. For example, running an FPGA-based primary with a software-based secondary significantly reduces the likelihood of shared faults, increasing reliability by design. Finally, we bring nearly two decades of experience operating production systems. That informs how we deliver Nexus as a managed service, covering everything from remote monitoring and version upgrades to supporting integration. The feed handler appliance is fully managed by Exegy, while the SmartNIC runs on the client’s server. Although we typically don’t have direct access to those client environments, we’ve focused on providing the right tools to ensure the SmartNIC doesn’t feel like a black box. Instead, it becomes a transparent, controllable part of their trading infrastructure.”

Nexus is now available for order, with production delivery expected in Q4 2025. Early adopters have access to a pre-production integration environment. The platform is offered as a managed service, including deployment, monitoring and ongoing support.

Subscribe to our newsletter

Related content

WEBINAR

Recorded Webinar: Unlocking value: Harnessing modern data platforms for data integration, advanced investment analytics, visualisation and reporting

Modern data platforms are bringing efficiencies, scalability and powerful new capabilities to institutions and their data pipelines. They are enabling the use of new automation and analytical technologies that are also helping firms to derive more value from their data and reduce costs. Use cases of specific importance to the finance sector, such as data...

BLOG

Trading Technologies Taps SIGMA AI to Build Dedicated AI Hub and Names Andy Simpson as Head of AI

Trading Technologies (TT) has deepened its partnership with fintech firm SIGMA AI through a strategic minority investment, tasking the company with building a proprietary AI and innovation hub to embed artificial intelligence across TT’s global trading platform. The move, which expands on an existing collaboration established in 2024, aims to accelerate TT’s AI adoption for...

EVENT

AI in Capital Markets Summit London

Now in its 2nd year, the AI in Capital Markets Summit returns with a focus on the practicalities of onboarding AI enterprise wide for business value creation. Whilst AI offers huge potential to revolutionise capital markets operations many are struggling to move beyond pilot phase to generate substantial value from AI.

GUIDE

A-Team Group’s Valuations Vendor Directory 2009

An indispensable guide to valuations professionals seeking providers of services in the asset valuations market. A-Team Group’s latest release in its series of directories – available for FREE download – focuses on vendors of valuations data, models and analytics. But this is not just another list of firms with their telephone numbers – you can get that...