Enterprises shift AI workloads to private cloud infrastructure

Key Takeaways

AI Tipping Point: Private Cloud Dominates AI Workloads

Broadcom's Private Cloud Outlook 2026 report reveals a significant shift in AI workloads, with 56% of enterprises running or planning to run production AI inferencing on private cloud, while public cloud use drops to 41%. The report highlights the growing importance of private cloud infrastructure in supporting AI workloads, driven by the need for greater control, security, and flexibility. Key findings include a 15 percentage point drop in public cloud's share of production AI workloads in a single year, and 97% of IT leaders believing some of their public cloud spend is wasted. The report also notes that geopolitics is increasingly affecting IT strategy and operations, with data sovereignty and residency requirements overtaking jurisdiction-specific compliance as the leading geopolitical factor shaping infrastructure decisions.

Broadcom's Private Cloud Outlook 2026

Broadcom's Private Cloud Outlook 2026 report reveals a significant shift in the production inference landscape, with a decisive move towards private cloud infrastructure. The report highlights the growing importance of private cloud infrastructure in supporting AI workloads, with 75% of organizations planning to deploy AI and machine learning workloads in private cloud environments by 2026. This trend is driven by the need for greater control, security, and flexibility in managing AI workloads, as well as the desire to reduce costs and improve efficiency.

Codex Powers Data Science with Interactive Reports

OpenAI's Codex, an advanced AI model, can act as a data analytics agent, querying data sources like Databricks and generating interactive reports to diagnose business issues. The demonstration highlights how Codex can transform raw data into easily digestible and manipulable reports, saving valuable time for data scientists and analysts. Codex can create sophisticated visualizations and derive actionable insights with far less manual effort.

Linux Developers Use AI to Keep Vintage AMD GPUs Alive

Linux developers are using AI-powered coding to maintain Linux kernel drivers, specifically the R600 Gallium3D driver, which supports AMD/ATI HD 2000 through HD 6000 series graphics cards. The use of AI, such as GitHub Copilot, helps to clean up shader compiler code and automate driver maintenance, making it easier to keep older drivers alive and compatible with modern systems.

Senator Josh Hawley Warns About AI Dangers

Missouri Senator Josh Hawley is calling for regulations on the AI industry, warning that left unchecked, AI will prioritize profit over human values. He emphasizes the need for regulations to ensure AI aids workers without displacing them. Hawley's stance on AI has put him at the forefront of a debate among Republicans, with some colleagues promoting a hands-off approach to AI.

New AI Course Offers Hands-on Training

Berkeley Law has introduced a new course, 'AI and the Practice of Law,' which provides students with over 100 hours of hands-on training in using AI tools. The course focuses on enhancing legal skills without lowering the threshold for human judgment and input. Taught by Wayne Stacy, the course uses Anthropic's Claude platform to teach students how to effectively use AI in legal practice.

LogRocket MCP Connects AI Agents to User Experience

LogRocket has launched the LogRocket MCP, a tool that connects AI agents directly to user experience data, allowing them to diagnose issues and surface usability and technical problems. The tool integrates with various AI platforms, including Claude, ChatGPT, and Slack, and provides a complete picture of the user experience.

Blue Books Won't Save Our Children from AI

The author discusses the limitations of traditional methods, such as blue books, in preventing AI cheating. They argue that AI will make it easier for students to produce answers, making it more important for schools to teach them how to ask better questions. The author calls for a human-centric national regulatory framework for AI to protect children and prevent AI from serving corporate interests.

AI in Healthcare: The Governance Gap

The article highlights the growing use of AI in healthcare and the need for effective governance. AI is being used to reduce administrative burden, organize complex information, and help clinicians and patients move through the system more efficiently. However, the lack of clear governance and oversight can create new operational, legal, and patient-safety risks.

Raimondo: US Must Prepare for AI Transition

US Secretary of Commerce Gina Raimondo emphasizes the urgent need for preparation in the AI transition, advocating for workforce support and collaboration between government and industry. She stresses that while AI is creating new job opportunities, it also presents significant challenges that require proactive strategies from both the government and the private sector.

Model Quantization with NVIDIA TensorRT

NVIDIA's TensorRT can turn FP8 checkpoints into high-performance inference engines. Model quantization reduces model size and on-disk footprint, and when combined with TensorRT, provides faster inference, higher throughput, and more efficient GPU utilization.

Sources

• Broadcom's Private Cloud Outlook 2026 Reveals an AI Tipping Point as Production Inference Shifts Decisively to Private Cloud
• Broadcom's Private Cloud Outlook 2026 Reveals an AI Tipping Point as Production Inference Shifts Decisively to Private Cloud
• Codex Powers Data Science with Interactive Reports
• Linux developers are using AI vibe coding to keep vintage AMD GPUs alive — R600 driver cleaned up with GitHub Copilot gives HD 2000 to HD 6000 series a new lease of life
• Josh Hawley is warning about AI dangers. But regulations languish in Missouri
• New AI and the Practice of Law course offers students more than 100 hours of hands-on training
• Remove the blindfold from your AI agents: The LogRocket MCP shows agents how customers are actually experiencing your product
• Blue books won’t save our children from AI
• AI in Health Care: The Governance Gap
• Raimondo: US Must Prepare for AI Transition
• Model Quantization: Turn FP8 Checkpoints into High-Performance Inference Engines with NVIDIA TensorRT