OpenAI appointed Uday Ruddaraju as Chief Technology Officer, Compute on 19 July 2026; he joined OpenAI in July 2025 as Head of Compute and Infrastructure.
Role: CTO, Compute
- Mandate: Set compute architecture, procurement strategy and operational reliability for model training and inference.
- Scope: Hardware selection, cluster orchestration, data‑centre design, networking and cost-performance optimisation.
Compute Infrastructure for Large Models
- Core components: Accelerators (GPUs/TPUs), CPU hosts, NVMe storage, high‑bandwidth interconnects (InfiniBand/RoCE), power and cooling.
- Performance metrics: FLOPs, GPU‑hours, petaflops/exaflops, training throughput and inference latency.
- Distributed methods: Data‑parallel, model‑parallel, tensor/pipeline parallelism for scaling transformer models.
Career and Industry Links
- Appointment details: Promoted to CTO, Compute after one year at OpenAI; prior roles include infrastructure leadership at xAI and engineering roles at X, Robinhood and eBay.
- Relevance: Experience spans hyperscale infrastructure, site reliability engineering and cloud‑adjacent deployments.
IASPOINT Booster Facts
- OpenAI: AI R&D firm operating proprietary clusters and partnering with cloud providers for training large language models.
- xAI & X: xAI founded 2023; X is the rebranded Twitter platform.
- Data centre: Facility housing servers, networking and power systems; critical for latency, sovereignty and energy footprint considerations.
- Regulation: Export controls (notably US) affect supply of advanced AI accelerators and enterprise compute procurement.