Capital markets firms are processing larger datasets within tighter operating windows. As workloads expand, compute-heavy operations can consume more of the time available for analysis, end-of-day processing, and reruns.
This ebook explores how CPU and GPU resources can handle different stages of the same workflow. It explains where selective GPU acceleration may add throughput for repeated calculations across large datasets while established q workflows continue to run on CPU.
The guide covers transaction cost analysis, end-of-day processing, backtesting, risk analysis, large q operations, and simulation. It also sets out the questions teams should answer before testing KDB-X GPU Acceleration.
What You’ll Learn
- How CPU and GPU compute can support different stages of the same workflow
- Which capital markets workloads are stronger candidates for GPU acceleration
- How to determine whether compute is the bottleneck
- What to consider when testing performance on a representative workload
- How to connect faster processing with a measurable business outcome
Who Should Read This Ebook
This guide is for people responsible for capital markets technology, analytics, and compute infrastructure, including:
- Chief Technology Officers, Chief Information Officers, and Chief AI Officers
- Heads of platform, infrastructure, and electronic trading technology
- Quants, quant developers, and q and kdb+ developers
- Transaction cost analysis, execution analytics, and risk analytics teams
Download the Ebook
Learn how to identify the workloads where selective GPU acceleration could create more capacity within a fixed processing window.
Download your copy
"*" indicates required fields