Hi,
I’m interested in your project to optimize your application running on the EPYC 9375F server.
Approach:
1. Analysis:
* Profile the application to identify CPU, memory, and I/O bottlenecks.
* Analyze thread scheduling and NUMA memory allocation.
* Review how interrupts and context switches affect latency.
2. Optimization Plan:
* Pin critical threads to dedicated cores (CPU isolation) to avoid interference.
* Tune kernel parameters (e.g., CPU governor, interrupt affinity).
* Optimize memory usage to leverage DDR5 bandwidth.
* Adjust network stack parameters for 10Gb links (e.g., buffer sizes, offloading).
* If necessary, refactor code sections causing delays (locks, waits, inefficient loops).
3 Deliverables:
* Detailed report of analysis and identified bottlenecks.
* Step-by-step description of optimizations applied.
* Validation tests showing latency improvements.
Why Work With Me:
* Strong experience in performance tuning for high-throughput, low-latency systems.
* Proficient in Linux performance tools (perf, numactl, htop, tcpdump).
* Clear communication and documented process.
Timeline:
I can start immediately and deliver the initial analysis in 3–5 days, followed by implementation and validation.
If you have any specific performance metrics you’d like to target, please let me know.
Thank you for considering my proposal—I look forward to helping you achieve maximum performance.
Best regards,
Youness