Low Latency Server

1. Low-Latency: The Key to Success

Recently, the upsurge in low-latency trading has allowed investors to achieve tremendous profitability by executing trades at blazing-fast speed. With securities transactions occurring in microseconds and approaching nanoseconds, securities firms whose trading machines are capable of rapidly capturing market data and creating orders will prosper, while slower competitors will fall behind. For this reason, extreme low-latency has become not only key to building trading infrastructures, but also critical to successful investmentbanking and hedge-fund operations.

2. Staying Competitive

To stay competitive in the low-latency trading world, securities firms must deploy the fastest and most powerful servers. These servers, collocated in the same premises as the servers in the securities exchanges, retrieve massive market information from these exchange servers, analyze it with trading strategies, create orders for execution, and send them back to the exchange servers

While achieving low latency is critical, servers must also deliver consistent trading performance. Jitter in data transmission can cause orders to lag or algorithms to corrupt, leading to undesirable results and significant losses at the end of trading. Thus, careful hardware optimization is necessary to design low-latency, jitter-free servers so our customer firms can gain competitive advantage.

3. Supermicro’s 5th generation Hyper-Speed & HyperTurbo Technologies for Extreme Low-Latency Trading

Sigma Primary Supermicro introduces the 5th generation Hyper-Speed and Hyper-Turbo technologies: proprietary server-board level optimizations for extreme low-latency trading. These technologies are made possible with the latest VRMs as well as optimized firmware to focus on flexible tuning.

With Hyper-Speed, the CPUs, memory, and PCI-E cards are pre-accelerated in lockstep mode for the most reliable performance. Coupled with Hyper-Speed, Hyper-Turbo allows server system CPUs to maintain maximum Turbo Mode frequency under intense workloads