Job Description – Performance / Infrastructure Engineer (HFT Domain)
A leading diversified financial services group in India is seeking a highly skilled Performance Engineer to support and optimize its ultra-low-latency trading infrastructure. The role focuses on hardware tuning, Linux kernel customization, network stack optimization, and end-to-end latency reduction for high-frequency trading platforms.
Key Responsibilities
- Optimize server hardware (CPU, BIOS/UEFI, NUMA, NICs, SmartNIC/FPGA).
- Customize and tune Linux kernels for deterministic, low-latency performance.
- Enhance network performance across TCP/UDP, RDMA, DPDK, kernel bypass, multicast, and time-sync (PTP/NTP).
- Build microbenchmarks, measure tick-to-trade latency, and maintain latency dashboards.
- Collaborate with trading, development, and infra teams to diagnose bottlenecks.
- Work closely with hardware vendors for performance improvements.
Required Experience & Skills
- 6–10 years of experience in Linux performance engineering, kernel internals, and low-latency systems.
- Strong expertise across major Linux distributions and custom kernel builds.
- Hands-on knowledge of GRUB tuning, CPU pinning, hugepages, IRQ affinity, isolcpus, nohz_full, etc.
- Experience with low-latency networking: DPDK, Onload, VMA, RDMA, AF_XDP.
- Strong understanding of CPU microarchitecture, NUMA, cache hierarchy.
- Proficiency in C/C++, shell scripting, and a scripting language (Python/Perl).
- Exposure to HFT/market data feed handlers/tick-to-trade systems is preferred.
- RHCE certification preferred.
Ideal Candidate
- Someone deeply passionate about nanosecond-level optimization, system determinism, and high-performance computing.
- Highly analytical, detail-oriented, and comfortable in a fast-paced, mission-critical trading environment.