London Stock Exchange Group (LON: LSEG) has announced a collaboration with NVIDIA to let financial firms analyse large volumes of its Tick History data far faster. The markets infrastructure and financial data group says the approach uses NVIDIA’s cuDF software on NVIDIA GPUs, the graphics chips used for fast parallel computing.
The shares stood at 8,292p at 8.10am London time today, down 0.69% from yesterday’s close of 8,350p. The announcement came at 7am, before the market opened. The shares had already fallen 4.68% in yesterday’s session, before the news.
LSEG says its own benchmark, run with cuDF and the Polars data-processing library, cut processing time from hours to seconds. The company gave no details of how the test was run.
Tick History is LSEG’s product holding every recorded trade and quote. The company says the method needs no second full copy of the data and keeps the licences, permissions and controls that govern access.
LSEG’s reasoning is that equity trading is moving towards near 24-hour operation. That, it says, is pushing firms away from loading data at market close and running analytics overnight.
The dataset is large: 110 trillion rows across 100 million instruments, more than 30 years of history and more than 580 exchanges, trading venues and contributors. In the first half of 2026, traffic on LSEG’s real-time network rose 29% from a year earlier, reaching a new peak of 26 million messages a second in June, while customer API requests for Tick History data rose 48%.
Ron Lefferts, Co-Head of Data & Analytics at LSEG, framed the benefit as speed without added risk.
Customers want to work with detailed historical data at speed, without the burden and risk of creating another full copy.
Ron Lefferts, LSEG
Ioana Boier of NVIDIA said cuDF helps researchers analyse the data at scale without downsampling, which means cutting detail to shrink size.
LSEG says the work is part of its Everywhere AI strategy and is intended to support quantitative research, execution analysis and risk work, including market replay and AI-enabled research. The company gave no revenue figure, commercial terms or launch date.
