With today’s technological advancements, for a trading product to be successful it often has to utilize everything that hi-tech has to offer. As FPGAs become widely adopted and low-latency wireless links between exchanges gain popularity, the competition for certain trades moves into nanosecond space. For other types of trading products it is very important that hardware and software aspects are carefully designed as a whole: a system that productively uses both the speed of current processors and FPGA’s predictability and ability of parallelism.
Our team has been working in trading industry since 2005 participating in all aspects of trading: from software feed handlers and order gateways to back-office tools and risk checks; from implementing advanced analysis tools and high-frequency trading strategies in software all the way to implementing market feed handlers and complete trading strategies on modern FPGAs. We are the Subject Matter Experts when it comes to products for trading industry: we know how the industry works, we know all major exchange protocols, all the implementation drawbacks and how to overcome them. We did it all, we did it more than once and we did it with the lowest latency possible at the time.
Exchanges (market data and order routing):
- S. Stocks: NASDAQ, NYSE ARCA, BATS/Direct Edge
- KRX (multiple broker gateways)
- Software and hardware-based risk check implementations on a client side and on a network device (bump-on-the-wire) for centralized risk evaluation.
- 15c3-compliant risk engines in software and on FPGA
Strategies and Trading Engines
- Complete trading strategies on the FPGA for sub-microsecond tick-to-trade latency
- Specialized FPGA-based packet processing engines for early detection of certain trade triggers
- Designing ultra-low latency software trading systems:
- Consulting on hardware (servers, NIC cards, etc.) and system optimization
- Efficient low-latency software development with optional tasks handled by FPGAs.
- Large computational tasks optimizations using OpenCL on FPGAs and GPUs