Jyoti Ingle

I’m a Systems Performance Engineer with 5+ years of experience in distributed systems, high-performance computing (HPC), and low-latency networking. I specialize in MPI runtime internals and performance optimization for large-scale Linux systems.

At Trinetra Interconnect and C-DAC, I enabled MPI runtimes (MVAPICH2, MPICH via libfabric/OFI) on a custom 100 Gbps interconnect by resolving compatibility issues and validating bring-up. My work included diagnosing protocol and memory-handling defects, and delivering measurable performance improvements—66% latency reduction (6μs → 2μs) and 2.4× cluster efficiency improvement (35% → 85%).

I’m strongest when debugging complex multi-node failures end-to-end: correlating cross-node behavior with custom instrumentation and large-scale log analysis, then using tools like GDB, Valgrind, QCachegrind, and Intel VTune to drive root cause analysis. I also serve as an SME for MPI runtimes and benchmarking, mentoring engineers during HPC platform setup and guiding interconnect design improvements.