10 GPU capacity planning experts
GPU capacity planning shows up in very different seats across this group: AI GPU demand forecasting at Microsoft, ML GPU planning at Google, AI compute capacity delivery at AWS, HPC and accelerator capacity planning at AMD, and compute capacity planning for inference at OpenAI. Talk to them when you are sizing next year's GPU needs, choosing between owned, reserved and cloud capacity, or checking whether a supply plan will land on time.
Professionals to explore
Isabelle Munn
LinkedInExperience: DGX Cloud Technical Program Management Leader · NVIDIA
Isabelle Munn is DGX Cloud Technical Program Management Leader at NVIDIA since 2025 and was Sr. Director Cloud Infrastructure Engineering at Microsoft (2023-2025). Munn's LinkedIn profile names DGX Cloud, GPU capacity, data center delivery and hyperscale operations, so questions on the cloud side of an owned-versus-cloud GPU split fit here.
Is this you? Claim your profile and get paid for calls.
Jennifer Hibbetts
LinkedInExperience: Former Principal Program Manager · Microsoft
Now Senior Manager, Customer Program Management Datacenter Engineering and Operations at NVIDIA (since 2026), Jennifer Hibbetts was Principal Program Manager at Microsoft from 2019 to 2026. Jennifer has led cross-functional teams to deliver Nvidia and AMD GPU capacity solutions and Azure AI infrastructure programs, aligned with finance and global data center buildout, a fit for turning a capacity plan into delivered hardware.
Is this you? Claim your profile and get paid for calls.
Jimmy P.
LinkedInExperience: Technical Program Manager - AI Compute Capacity Delivery · Amazon Web Services (AWS)
For questions on getting GPU servers live in data centers, Jimmy P. is Technical Program Manager - AI Compute Capacity Delivery at Amazon Web Services (AWS) since 2025. The role covers owning end-to-end AI/ML compute capacity delivery across a global data center fleet, including GPU/server capacity planning and hyperscale capacity expansions. Before that, Jimmy was Supply Chain Manager - Demand Planning & Analytics at Amazon (2024-2025).
Is this you? Claim your profile and get paid for calls.
Kedar Upasani
LinkedInExperience: Sr Manager Team Lead - Capacity Planning & S&OP · AMD
At AMD, Kedar Upasani is Sr Manager Team Lead - Capacity Planning & S&OP (since 2025), heading High Performance Computing (HPC) and Accelerators (AI/ML GPU) capacity planning and managing long term capacity, capex and S&OP. Sr Technical Infrastructure Program Manager (Sr TIPM) - GenAI ML EC2 at Amazon Web Services (AWS) (2023-2025) adds the cloud provider side of capex timing questions.
Is this you? Claim your profile and get paid for calls.
Manjunath Shivanna
LinkedInExperience: HPC Architect, AI Infrastructure · NVIDIA
Manjunath Shivanna has been HPC Architect, AI Infrastructure at NVIDIA since 2021 and Director, IT Infrastructure Operations at NVIDIA since 2017, after Sr Manager, Datacenter Operations (2012-2017). Capacity planning, GPU scheduler maintenance, high performance storage and InfiniBand are among the AI infrastructure team's responsibilities, so cluster scheduling and sizing questions fit here.
Is this you? Claim your profile and get paid for calls.
Michelle K. Powers
LinkedInExperience: Program Manager - Machine Learning Supply Chain & Operations Planning · Google
Program Manager - Machine Learning Supply Chain & Operations Planning at Google since 2024, Michelle K. Powers describes a focus on AI infrastructure capacity strategy, leading GPU demand and supply planning. Earlier roles include Supply Chain Planning Manager - Wireline Network Infrastructure at AT&T (2022-2024) and Senior Consultant - Supply Chain & Operations at EY (2009-2011), a fit for balancing GPU supply against demand.
Is this you? Claim your profile and get paid for calls.
Nathan Conway
LinkedInExperience: Manager, ML GPU Planning · Google
Nathan Conway's Google titles are Program Manager II, Machine Learning Demand Planning (2022-2024), Program Manager III, Machine Learning Demand Planning (2024-2026), Senior Program Manager, ML GPU Planning Lead (2026) and Manager, ML GPU Planning since June 2026. That pairing of ML demand planning and GPU planning titles fits questions on how machine learning demand forecasts feed a GPU plan.
Is this you? Claim your profile and get paid for calls.
Sandy Mohanakumar
LinkedInExperience: Compute Capacity Planning · OpenAI
Compute Capacity Planning at OpenAI since 2025, Sandy Mohanakumar describes leading capacity strategy for production inference, translating product and model demand into infrastructure requirements and allocation decisions, and shaping GPU fleet strategy. Earlier Google roles were Staff Program Manager AI ML Infrastructure Strategy & Capacity Planning (2024-2025) and Program Manager III AI ML Infrastructure Capacity Planning (2022-2024).
Is this you? Claim your profile and get paid for calls.
Sibel Aktas Egilmez, PMP
LinkedInExperience: Senior Demand Forecasting Manager - AI GPU Planning · Microsoft
Sibel Aktas Egilmez, PMP, holds Senior Demand Forecasting Manager - AI GPU Planning at Microsoft (since 2026), alongside Senior Technical Program Manager, Cloud Supply Chain(CSCP) at Microsoft (2024-2026). Senior Sales & Operations Planning (S&OP) Manager at Philips (2019-2021) and Senior Demand Planner - Global Supply Chain at Bio-Rad Laboratories (2016-2019) also appear, which suits questions on building the GPU demand forecast itself.
Is this you? Claim your profile and get paid for calls.
Vivek Chander Selva Kumar
LinkedInExperience: Former Demand & Supply Planning Manager, AI & Compute · Meta
Demand & Supply Planning Manager, AI & Compute at Meta (2022-2026) and Program Manager, Demand & Supply planning at Google since 2026 are the two most recent roles for Vivek Chander Selva Kumar. At Meta the work was commodity planning focused on GPU, NIC, power and liquid cooling, covering strategic sourcing and production planning, a fit for questions on the parts and power that gate a GPU build.
Is this you? Claim your profile and get paid for calls.
Choose the right perspective
Forecasting questions belong with the demand planners: Sibel Aktas Egilmez, Nathan Conway, Michelle K. Powers and Vivek Chander Selva Kumar. Jimmy P., Jennifer Hibbetts and Isabelle Munn hold delivery and program titles, so pick them for getting capacity installed, while Sandy Mohanakumar and Kedar Upasani are the closer match for allocation, capex and long-term capacity choices.
Questions to take into the conversation
- 01How many months ahead of a model launch does a GPU capacity forecast need to be locked, and what changes after that point?
- 02When GPU supply is short, how is the available capacity split between training, inference and internal teams?
- 03What share of GPU capacity would you keep owned versus reserved or on-demand cloud, and what would change that split?
About this directory
This is a professional research starting point based on business profile data retrieved on . Titles and companies reflect that source snapshot; roles marked former had ended by then. Check the linked profiles for current details. Inclusion does not imply Instant Expert membership or availability.