Category: Tech Notes
-
Azure OpenAI Private Link and VNet Isolation (Azure Gen AI Series, Part 9)
Private networking for Azure OpenAI, done in the right order: private endpoints, the DNS zones that make them work, when a managed VNet earns its keep, and the On Your Data trap that closes every call.
-
Azure OpenAI Regions, Quotas, and Data Zones (Azure Gen AI Series, Part 8)
Azure OpenAI quota is a grid: per region, per subscription, per model, per deployment type. Here are the real 2026 numbers, the new quota tiers, how data zones change residency, and the two commands I run before promising any capacity.
-
Azure ND-Series GPUs and the Maia Accelerator (Azure Gen AI Series, Part 7)
Azure’s ND-series gives you NVIDIA H200, Blackwell GB200, and AMD MI300X GPUs by the node, while Maia is Microsoft’s own inference silicon you cannot rent. What each one is, how to size a GPU to your model, and the short list of times self-hosting beats the managed endpoint.
-
Azure OpenAI Deployment Types, Standard vs PTU vs Batch (Azure Gen AI Series, Part 6)
Standard, provisioned throughput units, and Batch are the three ways to bill an Azure OpenAI deployment. How to pick with a utilization break-even, size PTUs, and use spillover so you are not throttled or overpaying.
-
Azure OpenAI vs Foundry vs Azure Machine Learning, and Which One Your Project Belongs In (Azure Gen AI Series, Part 5)
Azure OpenAI, Microsoft Foundry, and Azure Machine Learning look interchangeable and are not. A practical guide to which resource your project belongs in, and how to move between them without a rebuild.
-
Azure Foundry Model Catalog and Models-as-a-Service (Azure Gen AI Series, Part 4)
The Foundry catalog splits into models sold by Azure and partner models, and every one deploys as serverless Models-as-a-Service or as managed compute. Here is how to pick, with the permission and rate-limit traps that bite first.
-
Azure OpenAI in Foundry Models, and How to Choose One (Azure Gen AI Series, Part 3)
Azure OpenAI in Foundry Models is the set of OpenAI models you deploy inside a Foundry resource. Here is which models you get in 2026, how to pick one, and how the call goes out with the v1 API.
-
Azure AI Foundry Platform and Foundry Projects Explained (Azure Gen AI Series, Part 2)
Microsoft renamed Azure AI Foundry and rebuilt it around one resource, projects, and a single SDK. Here is what the platform is, how projects work, and your first API call against it.
-
Azure Generative AI Stack, End to End (Azure Gen AI Series, Part 1)
The whole Azure generative AI stack in one map: Azure AI Foundry, the model catalog, the three deployment types that decide your bill, retrieval, agents, and the compute underneath. Where to start and what to skip.
-
AWS Generative AI vs the Field, the Verdict (AWS Gen AI Series, Part 30)
After thirty parts on the AWS generative AI stack, the verdict: where Bedrock, Amazon Nova, and AWS silicon earn the default, and the exact cases where Azure, Google Cloud, or IBM watsonx fit better.
-
Bedrock Reference Architectures for Chatbot, RAG, Agentic, and Batch (AWS Gen AI Series, Part 29)
Most AWS generative AI features are one of four shapes: chatbot, RAG, agentic, or batch. Here is how each maps to Amazon Bedrock services, what it costs, and which one to reach for first.
-
LLMOps and CI/CD for Amazon Bedrock and SageMaker (AWS Gen AI Series, Part 28)
LLMOps on AWS is two pipelines, not one. Version Bedrock prompts as code, gate every change on evaluation, register and deploy models through SageMaker, and keep a rollback you have actually tested.
Architect’s Toolkit
PJ’s Tools
VMware Cloud Foundation
- VCF Documentation
- VCF 9 Planning & Preparation Workbook
- VCF Bill of Materials (BoM)
- VMware Compatibility Guide
- VMware Interoperability Matrix
- VMware Configuration Maximums
- VMware Ports & Protocols
- VMware Hands-on Labs
- RVTools Download
Nutanix
AI & Cloud-Native Platform
- NVIDIA Build (Model Catalog)
- NVIDIA AI Enterprise Reference Architecture
- NVIDIA NIM Performance Benchmarking
- NVIDIA NGC Catalog
- NeMo Microservices Helm Chart
- Helm Charts Repository
- Hugging Face Models
Architecture & Design
About the Author
Dr Pranay Jha
Dr. Pranay Jha is a Cloud and AI Consultant with 18+ years of experience in hybrid cloud, virtualization, and enterprise infrastructure transformation. He specializes in VMware technologies, multi-cloud strategy, and Generative AI solutions. He holds a PhD in Computer Applications with research focused on Cloud and AI, has published multiple research papers, and has been a VMware vExpert since 2016 and a VMUG Community Leader.
You May Have Missed
DrJha