// Compute & HPC Engineer

Gaurav Yadav

Building high-performance compute infrastructure, GPU-accelerated systems, and enterprise virtualization at scale

Oracle Cloud Infrastructure Private Cloud Appliance Edge Cloud

8+
Years Experience
67%
Boot Latency Reduced
50%
Deploy Time Saved

01. About Me

Senior Software Developer with 8+ years of expertise in cloud compute infrastructure, server virtualization, and high-performance computing workloads. Currently driving compute ecosystem innovation at Oracle Cloud Infrastructure's Private Cloud Appliance team, specializing in GPU-accelerated computing, hypervisor performance optimization, and edge cloud solutions. Deep hands-on experience with KVM/QEMU, libvirt, SR-IOV, GPU passthrough, and bare-metal provisioning — enabling enterprise HPC and AI/ML workloads at scale. Proven track record in reducing deployment latency, achieving FedRAMP compliance, and building converged infrastructure for mission-critical compute environments.
Technical Arsenal
Compute & HPC
KVM/QEMU libvirt GPU Passthrough vGPU SR-IOV PCI Device Mgmt Bare-Metal NUMA
Cloud & Infrastructure
OCI / PCA Edge Cloud AWS Kubernetes Docker/Podman Terraform
Programming
Python 3 Bash/Shell Rust JavaScript SQL
Performance
Boot Optimization Live Migration Autoscaling Load Balancing Workflow Orchestration
Security & Compliance
FedRAMP Non-root Hypervisor Certificate Mgmt ACLs sudoers Policy
Tools & Architecture
Git RabbitMQ Microservices IaC CI/CD MySQL/PostgreSQL

02. Experience

Software Developer 3 — Compute & HPC
Oracle Cloud Infrastructure
June 2022 — Present
Bengaluru, India
  • GPU & HPC Enablement: Developed GPU passthrough, vGPU, and SR-IOV capabilities for PCA/Edge Cloud, enabling on-premises AI/ML and HPC workloads. Built GPU capacity metric service and configured virtnodedevd for PCI device management.
  • Hypervisor Modernization: Migrated VM Agent from monolithic to modular libvirt daemons (virtqemud, virtproxyd, virtnodedevd, virtnetworkd), eliminating single points of failure. Led Python 2→3 migration and OL7→OL8 upgrade, reducing deployment time by 50%.
  • Performance Optimization: Reduced instance boot latency by 67% by cutting network calls from 3 to 1 and eliminating persistence DB dependency at instance-metadata service.
  • Live Migration & HA: Resolved critical cross-version live migration issues between OL7/OL8, ensuring zero-downtime operations for production HPC workloads.
  • Converged Edge Infrastructure: Enabled single-device PCA/Edge Cloud deployments for disconnected field operations with optimized shape selection for compute and GPU resources.
  • FedRAMP Compliance: Migrated hypervisor to non-root operations using sudoers policy, ACLs, and comprehensive certificate management across multi-node deployments.
  • Network & Load Balancing: Configured virtnetworkd for virtual networks. Developed NLB support and instance-pool configuration with attach/detach functionality.
  • Compute Services: Maintained autoscaling, ovm-agent, and instance configuration management. Resolved split-brain issues and implemented workflow model POC for instance launch reliability.
Product Developer
Exceleron
Dec 2019 — June 2022
Bengaluru, India
  • Designed AWS serverless solutions using Lambda, DynamoDB, API Gateway, S3, CloudWatch Events, and Secrets Manager.
  • Integrated third-party APIs (YouTrack, Google Drive, Sendbird) and built Alexa Skills.
  • Utilized Terraform and AWS CDK for Infrastructure as Code deployment and environment management.
Software Engineer
Oust Labs India Pvt Ltd (Acquired by Betterplace)
May 2017 — Dec 2019
Bengaluru, India
  • Developed scheduling microservice and event-driven Lambda functions for content delivery and data lifecycle management on AWS.
  • Managed AWS infrastructure (S3, CloudFront, Elastic Beanstalk, Route53, RDS) and led migration to Mumbai region.
Bachelor of Technology (B.Tech)
Dr. A.P.J. Abdul Kalam Technical University
2008 — 2012
Ghaziabad, India

Electronics and Telecommunications Engineering

03. Key Projects

04. Contact

+91 9310150558
Bengaluru, India