High-Performance Computing Research Engineer | Distributed Systems | Parallel Computing
I'm a Computer Science graduate (MS, 4.0 GPA) from NC State University, specializing in High-Performance Computing research and development. My expertise spans cutting-edge parallel computing, distributed systems optimization, and scalable infrastructure design. I've contributed to premier research institutions including Lawrence Livermore National Laboratory, where I develop next-generation HPC solutions that push the boundaries of computational performance.
My work bridges the gap between theoretical computer science and real-world impact, focusing on RDMA-based acceleration, GPU optimization, and cloud-native HPC architectures that solve critical scientific computing challenges.
- Asynchronous IO-Offloading: Revolutionizing HPC data movement with RDMA-powered pipelines and DPU acceleration
- Power-Aware Computing: Dynamic resource allocation frameworks for energy-efficient supercomputing
- Parallel Algorithms: GPU-accelerated scientific computing and distributed graph processing
- Systems Optimization: Low-level performance tuning for next-generation HPC infrastructure
Breaking the IO bottleneck in exascale computing
- Engineered RDMA-based data streaming pipelines achieving 100+ Gbps throughput (near hardware limits)
- Developed novel DPU offloading techniques reducing CPU overhead by 90% in checkpoint operations
- Containerized high-speed dataflow for LLNL's cutting-edge Rabbit storage infrastructure
- Tech Stack: C/C++, RDMA, Libfabric, UCX, DOCA SDK, Nvidia DPUs
Intelligent power allocation for overprovisioned supercomputers
- Designed hierarchical power provisioning framework with real-time CPU/GPU power reallocation
- Achieved 25% improvement in cluster-wide resource utilization through intelligent scheduling
- Integrated seamlessly with SLURM and production HPC environments
- Tech Stack: C/C++, MPI, Power APIs, HPC Schedulers
Compile-time insight into runtime performance characteristics
- Built custom LLVM compiler passes for automated performance bottleneck detection
- Implemented sophisticated control-flow analysis to identify performance-critical code paths
- Enables proactive optimization in scientific computing applications
- Tech Stack: LLVM, C++, Compiler Design, Static Analysis
State-of-the-art parallel graph algorithms
- Implemented cutting-edge MST algorithm from SC'23 conference using advanced CUDA techniques
- Achieved 200-250% performance improvement through thread and warp-level optimizations
- Advanced synchronization reduction and memory access pattern optimization
- Tech Stack: CUDA, GPU Computing, Parallel Algorithms
High-performance mathematical visualization platform
- Developed multi-threaded fractal generator with GPU acceleration using Vulkan
- Real-time parameter tuning and deep zoom capabilities for research applications
- Combines mathematical computing with advanced graphics programming
- Tech Stack: Vulkan, C++, Multi-threading, Mathematical Computing
Cloud-native computer vision for IoT security
- Architected event-driven security system using serverless computing principles
- Real-time intrusion detection with distributed processing and alerting
- Tech Stack: AWS (Lambda, SQS, SNS, DynamoDB), Computer Vision, IoT
Production-scale social networking with 5K+ daily users
- Full-stack development with automated CI/CD deployment pipelines
- Scalable cloud architecture handling real-time user interactions
- Tech Stack: MERN, AWS (CloudFormation, CodePipeline, EKS), Kubernetes
- MPI/OpenMP | CUDA Programming | Vulkan Graphics
- RDMA (Libfabric, UCX, Libverbs) | DPU Programming (DOCA SDK)
- GPU Computing (cuBLAS, OpenBLAS) | Distributed Systems
- AWS/GCP | Docker/Kubernetes | Terraform/CloudFormation
- CI/CD (GitHub Actions) | Infrastructure as Code
- Performance Analysis | System Optimization | Parallel Algorithms
- Computer Vision | Machine Learning | Scientific Computing
Jun 2025 - Present
- Developing breakthrough IO acceleration technologies for exascale computing
- Research focus: RDMA-based checkpoint acceleration and DPU optimization
Aug 2024 - May 2025
- HPC systems research under Dr. Frank Mueller
- Power-aware computing and scalable checkpoint systems
May 2024 - Aug 2024
- Big data pipeline development with Apache Spark and Databricks
- ML model development and deployment automation
Jan 2023 - Jul 2023
- Cloud operations automation reducing incident response time by 25%
- Enterprise-scale monitoring and alerting systems
Oct 2021 - Jan 2022
- Real-time LiDAR streaming for autonomous systems (published in IEEE Sensors Journal)
- 5G and wireless networking optimization
6+ peer-reviewed publications in top-tier venues:
- IEEE Sensors Journal (Impact Factor: 4.325)
- IEEE Xplore Conference Proceedings
- International Journals in AI/ML and Computer Science
Research Areas: Network optimization, deep learning for medical applications, real-time data streaming, nature-inspired algorithms
- Academic Excellence: 4.0/4.0 GPA (MS), 8.73/10.0 GPA (BTech)
- AWS Certified Developer - Associate
- 2nd Place: Internet-of-Things Hackathon 2021
- 6 Publications: In prestigious international journals and conferences
- Founder: GitHub Organization for Advanced Academic Center, mentoring 50+ students
- Technical Mentor: Led web development teams and conducted Git/software engineering workshops
- Training Lead: Designed curriculum and assessment programs for emerging developers
-
"Asynchronous IO-Offloading for Scalable Checkpoint-Restart" (Expected 2025)
- Collaboration: LLNL & NC State University
- Focus: Next-generation HPC data movement
-
"OverPower: Power-Aware Provisioning Framework for HPC" (Expected 2025)
- NC State University research
- Focus: Energy-efficient supercomputing
I'm passionate about pushing the boundaries of computational performance and solving complex distributed systems challenges. Whether you're working on:
- Infrastructure Engineering at scale
- Research & Development in HPC
- Performance Optimization challenges
- Distributed Systems architecture
I'd love to connect and explore how we can advance the field together.
Professional Network: LinkedIn
Building the future of high-performance computing, one algorithm at a time.