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

Hardware Acceleration – A Roundup and Update with Nomura, Burstream and Fiberblaze

Subscribe to our newsletter

What’s been happening in the world of hardware acceleration of late? Despite a couple of killjoys on LinkedIn who reckon it’s still unproven technology, the answer is – quite a lot.

It was good to get an update from the coalface a couple of weeks ago at our London Low-Latency Summit. There, I chatted to Nomura’s Ben Stephens, who sells HFT and electronic trading services for the investment bank.

Nomura’s NXT platform – it stands for Nomura Extreme Technology – incorporates Field Programmable Gate Array (FPGA) technology for market data handling and pre-trade risk checks.

Stephens reckons that using FPGAs – and avoiding the I/O data bottlenecks between network and main CPU – saves NXT between 10 to 50 microseconds in overall latency. It also allows for very deterministic latency – i.e. minimal jitter – which is “really important” for customers running arbitrage trading.

Recent news includes a new managed service from Burstream, which began offering Nasdaq market data using FPGA-based data feed appliances from NovaSparks. Latencies of less than a microsecond – even at peak periods – are being observed.

Then the name Fiberblaze cropped up. A Danish company that builds FPGA-based network cards. Ullink has started to use them in its pre-trade risk offering, presumably offloading checks on to the card.

Also on our radar: Celoxica, with an offering for HFT Futures, and Redline Trading Solutions, which is making Goldman Sachs’ Sigma X dark pool zip along. Unlike many others in the hardware acceleration space, Redline uses Cell processors instead of FPGAs.

Subscribe to our newsletter

Related content

WEBINAR

Recorded Webinar: The Role of Data Fabric and Data Mesh in Modern Trading Infrastructures

The demands on trading infrastructure are intensifying. Increasing data volumes, the necessity for real-time processing, and stringent regulatory requirements are exposing the limitations of legacy data architectures. In response, firms are re-evaluating their data strategies to improve agility, scalability, and governance. Two architectural models central to this conversation are Data Fabric and Data Mesh. This...

BLOG

Bloomberg Enhances RMS Enterprise to Unlock Proprietary Models and Strengthen Research Oversight

Bloomberg has announced significant enhancements to its enterprise-level Research Management Solution (RMS Enterprise), introducing two new capabilities: Custom Fundamentals and Digest Alerts. The updates are designed to address long-standing data interoperability challenges within investment firms, allowing research teams to better integrate proprietary financial models into their workflows and strengthen oversight across their organisations. For many...

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

Entity Data Management Handbook – Fourth Edition

Welcome to the fourth edition of A-Team Group’s Entity Data Management Handbook sponsored by entity data specialist Bureau van Dijk, a Moody’s Analytics company. As entity data takes a central role in business strategies dedicated to making the customer experience markedly better, this handbook delves into the detail of everything you need to do to...