Hi, I'm Sai Saketh Cherukuri, an AI Performance Engineer at Red Hat. I studied Mathematics with a Computer Science specialization at Sri Sathya Sai Institute of Higher Learning. I build and study machine learning systems with a focus on inference performance, GPU optimization, and efficient model serving.

My work centers on understanding how workloads behave on real hardware. I use MLPerf inference benchmarking, CUDA, vLLM, PyTorch, Nsight Systems, and Nsight Compute to investigate throughput, latency, KV cache behavior, batching, scheduling, memory utilization, and kernel efficiency.

I approach optimization as a measurement driven engineering discipline. Recent projects include an MLPerf study of Llama 3.1 8B, an H200 quantization benchmark suite, and a periocular age and gender recognition system.