Avnish Chhabra

About Avnish Chhabra

Technology executive with 25+ years of experience building and operating large-scale cloud, compute, networking, and distributed systems platforms.Currently Vice President of GPU Compute at Oracle Cloud Infrastructure, leading AI infrastructure platform delivery for large-scale training and inference workloads in partnership with NVIDIA and AMD.My experience spans the core infrastructure layers behind modern cloud and AI platforms: networking and communications systems in Windows, early inference and decisioning systems in Bing Ads, distributed storage and compute platforms in Bing, hyperscale cloud deployment and compute platforms in Azure, infrastructure economics and capacity efficiency at Meta, and GPU / AI infrastructure leadership at Oracle.At Oracle, I lead GPU Compute engineering for OCI, with responsibility for the GPU server and AI infrastructure software stack spanning provisioning, validation, diagnostics, automated repair, customer APIs, maintenance handling, and fleet health monitoring. During my tenure, OCI GPU platform capacity grew from roughly 126K to 436K GPUs, with the same platform foundation enabling an additional 150K GPUs in Oracle’s raw-metal cloud.I have led platform delivery and scale-out for strategic AI customers including OpenAI, Meta, and ByteDance, enabling new customer activation, major capacity expansion, and adoption across GPU generations from H100 through GB300.Previously at Meta, I led capacity efficiency, infrastructure economics, and quota management systems, helping deliver more than $1 billion in capacity savings. At Microsoft, I led Azure Pilotfish (Autopilot) Cloud, company-wide deployment platforms, and Azure air-gapped cloud SRE, supporting major Microsoft businesses and Azure IaaS deployments at hyperscale.

Experience

1. Vice President of Software Engineering for Gpu Compute

   Oracle

   Apr 2024 — Present

   Lead Oracle GPU Compute engineering, driving AI infrastructure platform delivery across the GPU server and software stack with a global engineering and SRE organization.Own the GPU server and AI infrastructure software stack, spanning provisioning, validation, diagnostics, automated repair, customer APIs, maintenance handling, and fleet health monitoring.Expanded OCI commercial GPU platform capacity by roughly 3.5x during my tenure, with the same platform foundation leveraged to enable substantial additional GPU capacity in Oracle’s raw-metal cloud.Led the GPU Compute side of platform delivery for strategic AI customers including OpenAI, Meta, and ByteDance, directly enabling new customer activation, major capacity expansion, and adoption of successive GPU generations from H100 through GB300.Delivered GB200 commercial GA in 2 months from QS rack availability, enabled customer API milestones more than 10 weeks ahead of GA, and drove major improvements in ingestion, diagnostics, repair automation, and operational quality across the GPU fleet.

Education

• SVKM's Narsee Monjee Institute of Management Studies (NMIMS)

  MBA (incomplete), Finance, General

  1998 — 1999

• University of Mumbai

  Bachelor of Engineering (B.E.), Computer Engineering

  1994 — 1998

• Virginia Tech

  Master of Science (M.S.), Computer Science

  1999 — 2001

Skills

• Software Development
• Scrum
• Xp
• Windows Server
• Servers
• Operating Systems
• Firewalls
• Cloud Computing
• Databases
• Security
• Software Design
• Windows
• Distributed Caching
• Ipsec
• Low Latency
• System Architecture
• Vpn
• Agile Methodologies
• Big Data
• Cloud Storage
• Windows Azure
• Testing
• Scalability
• Software Engineering
• Networking
• High Availability
• Integration
• C
• Project Management
• Web Services
• Storage
• Tcp/Ip
• Algorithms
• .net Framework
• C#
• .net
• Distributed Systems