- The Cloud Cover
- Posts
- The AI Arms Race Takes a Surprising Turn
The AI Arms Race Takes a Surprising Turn
This Week in Cloud — June 12, 2025
Welcome back to The Cloud Cover, your guide to the world of cloud. This week, we saw surprising new partnerships, proving that the demand for compute power now outweighs traditional rivalries. It's all part of the great cloud build-out, as providers spend billions on new data centers and race to embed smarter AI into every layer of their stack. From Oracle's "insatiable" demand to OpenAI's new partner, here's what you need to know.
⚡ In a Stunner, OpenAI Taps Google Cloud for AI Compute
The biggest story of the week is a testament to the raw, gravitational force of AI compute demand. In a move that turned heads across the industry, OpenAI, Microsoft's marquee AI partner, has reportedly signed a deal to tap into Google Cloud's infrastructure to train its models. While OpenAI remains deeply partnered with Azure, this development underscores a crucial reality: the thirst for specialized, high-performance AI capacity is so immense that it transcends traditional rivalries.
This "coopetition" signals two key things. First, it’s a massive validation for Google Cloud’s AI infrastructure, particularly its highly-touted Tensor Processing Units (TPUs), proving they are a compelling option for the world's leading AI labs. Second, it suggests that AI leaders are actively pursuing multi-cloud strategies to de-risk their supply chains, gain access to the best-fit hardware for specific jobs, and leverage price competition. For architects and IT leaders, this is a powerful reminder that the cloud landscape is becoming more fluid and complex, driven by the unique and voracious appetite of AI workloads.
🔍 The Rundown
Smarter AI for Developers: Amazon Q, the AI developer assistant, gained new "agentic" capabilities. It can now autonomously understand complex requests, select the right tools from over 200 AWS services, and execute multi-step plans, showing its work along the way.
GPU Price Cuts: Making AI training and inference more affordable, AWS slashed prices by up to 45% for several of its NVIDIA GPU-accelerated EC2 instances (P4d, P4de, and P5).
Specialized and Global Expansion: AWS announced plans for a second U.S. "Secret Region" to handle classified government data, reinforcing its public sector focus. It also officially opened its new commercial cloud region in Taipei, Taiwan.
Deepened Databricks Partnership: Microsoft and Databricks announced a multi-year extension of their strategic partnership. The collaboration will focus on tighter native integrations between Azure Databricks, Azure AI, and the Power Platform.
New Storage-Optimized VMs Available: Azure announced the general availability of its new Lsv4-series VMs. These storage-optimized instances are designed for workloads needing massive local disk throughput like big data and high-performance databases, offering up to 23TB of local NVMe SSD and 6.6 million IOPS.
New Databricks Tier for Regulated Industries: GCP and Databricks launched a new "Enterprise tier" of the Databricks Data Intelligence Platform. The tier adds security and compliance features like private connectivity and customer-managed encryption keys, aimed at customers in highly regulated sectors.
"Insatiable" AI Demand Fuels Growth: OCI was the big financial winner this week, reporting a staggering 52% year-over-year increase in IaaS revenue. Executives credited the surge to "insatiable" demand for cloud capacity to train AI models.
Expanded NVIDIA Partnership: Oracle is bringing NVIDIA's latest tech to its customers, announcing support for the powerful GB200 Blackwell GPU superchips and making the NVIDIA AI Enterprise software suite available directly from the OCI console.
📈 Trending Now: The Great Cloud Build-Out
If one story this week captured the sheer scale of the AI gold rush, it was Oracle's earnings report. Executives didn't just report strong growth; they described customer demand for AI capacity as "insatiable" and announced plans to nearly double capital expenditures to over $25 billion next year just to keep up. This isn't an isolated event; it's a snapshot of a massive, industry-wide phenomenon: the great AI infrastructure build-out.
Across the board, providers are in a frantic race to build data centers and secure the highly specialized chips needed to power the next wave of AI. AWS reaffirmed plans to invest at least $30 billion in AI infrastructure in the U.S. alone. Microsoft and Google are spending at a similar clip. This isn't just about adding more servers; it's a strategic land grab for the foundational resource of the 21st-century economy. For cloud customers, this means that while compute is becoming more powerful, capacity in the most sought-after regions and for the most advanced chips could become a key point of negotiation and a critical factor in strategic planning for the foreseeable future.
📅 Event Radar
20
Free registration
9-10
Registration open now!
16
Session catalog now available
💼 Job Spotlight
 | Solutions Architect, Migrations at Snowflake $128,000-$178,500 | Remote US Drive enterprise-scale data warehouse migrations to Snowflake, leading end-to-end transformation from legacy systems to cutting-edge cloud data platforms. |
 | Senior Solutions Architect at Apollo GraphQL $158,000-$186,000 | Remote US Help the world’s top engineering teams scale GraphQL with federated architectures as a trusted technical advisor at the forefront of API innovation. |
👋 Until Next Week
This week hammered home that AI isn't just a service offering; it's the central organizing principle of the cloud industry right now. It's driving multi-billion dollar investment decisions, forcing unconventional partnerships, and creating new winners. The main thing to watch next is how this incredible demand continues to reshape alliances and infrastructure strategy.
Do you enjoy these emails? Your friends and colleagues might, too! Help us grow the cloud community by sharing the newsletter with others.