Cloud Software Development Engineer
CurrentDemonstrate new Intel technology and accelerators for integration into Microsoft Data products (SQL Server and Synapse) and investigate performance opportunities and bottlenecks for data benchmarks such as TPC-E/HammerDB/TPC-DS on SQL Server and Synapse Spark. Responsibilities:Design experiments and summarize findings to present to internal stakeholders and Microsoft.Use profiling tools like EMON, perfmon, vtune, xperf to characterize data workloads encompassing application, OS, storage and network stacks. Develop automation tools and scripts to make experiments repeatable, reliable and transferable.Responsible for setup, debug and maintenance of multiple server hardware systems in lab.Key Highlights: 1. Drove adoption of Intel QAT accelerator for database backup/restore in SQL Server. Identified bugs and opportunities for performance optimization in QAT driver and SQL Server resulting in 2x+ speedup on bare metal and virtualized scenarios.2. Reduced storage cost and complexity overhead by 2x by architecting an NVME SSD based storage solution for TPC-E workload to replace SATA SSD solution at 0 performance impact3. Demonstrated optimization opportunities in SQL Server for OLTP (TPCE) database workload. 4. Enabled and optimized Hybrid Buffer Pool feature into MS SQL Server using Persistent Memory leading to 20% improvements in throughput for OLTP (TPCE) database workload.