AI
articles
GPT-5 Lands With Market-Leading Performance, a Nod to Openness
OpenAI’s GPT-5 outperforms rivals in coding, context retention, and accuracy—setting a new bar for enterprise AI while signaling a subtle shift toward openness.
Mid-Year Report: AI’s Biggest 2025 Surprises Revealed
Market share shakeups, pricing shocks, and a tectonic shift in the open internet: Intel’s Lynn Comp unpacks developments in AI trends in 2025 that no one could have predicted.
MLPerf Storage v2.0 Eclipses Records in AI Benchmarking
Global surge in submissions reveals the pivotal role of storage in scaling AI training, with new checkpoint tests tackling failure resilience in massive accelerator clusters.
MLCommons Raises the Bar for AI PCs with MLPerf Client v1.0
MLCommons launches industry-standard benchmarks for LLM performance on PCs, cutting through marketing hype and giving developers and enterprises the transparent metrics they need.
AI’s New Canvas: The Visual Creators in the Zoo
From Midjourney to Firefly, Part 2 of our ‘AI Zoo’ series breaks down how today’s top image models work—and how TechArena uses them to create powerful, responsible visuals.
The Lamb Lays Down with the Lion to Avoid Being Eaten by the Wolf
As Chinese EV giants like BYD rise, German automakers are forging an unlikely alliance, but history shows such partnerships often crumble within months.
MLPerf Training v5.1: GenAI Drives, Scale Ramps, Field Widens
Two new genAI tests (Llama 3.1 8B, Flux.1) align with production stacks as multi-node results climb. NVIDIA posts many fastest times; University of Florida, Wiwynn, and Datacrunch expand the ecosystem.
Wanjiku Kamau on Her Book: Out of the Loop, Into the Algorithm
Allyson Klein talks with author and Google/Intel alum Wanjiku Kamau on moving past AI skepticism, learning fast, and using new tools with intention—so readers start where they are and explore AI with hope.
Cooling the Future: Where Does Cold Plate End and Immersion Begin?
AI racks are blowing past air’s limits. Here’s a frank framework for when cold plate still wins, when it fails, and how to plan the pivot to immersion—without stranding today’s investments.
KubeCon Atlanta: CNCF Sets AI Standard for Kubernetes
On Day 1 of KubeCon + CloudNativeCon Atlanta, CNCF unveiled Kubernetes AI Conformance to make workloads portable—arriving as inference surges to ~1.33 quadrillion tokens/month across Google’s systems.
How Financial Services Can Responsibly Scale AI
FinTech expert Anusha Nerella shares insights on staying ahead of fraud, navigating regulation, and building collaborative teams to scale responsible AI across the financial services sector.
Under the Hood: The Alphabet Soup of Car Connectivity
Modern software-defined cars blend multiple links—CAN/LIN, MIPI, SerDes, and Ethernet/TSN—to shrink wiring and cost, manage EMI, and deliver reliable, deterministic timing from sensors to actuators.
.svg)
Podcasts
How Hedgehog Brings Hyperscaler Networking to Enterprise AI
Hedgehog CEO Marc Austin joins Data Insights to break down open-source, automated networking for AI clusters—cutting cost, avoiding lock-in, and keeping GPUs fed from training to inference.
Nebius at SC25: Building the Neocloud for Enterprise AI
From SC25 in St. Louis, Nebius shares how its neocloud, Token Factory PaaS, and supercomputer-class infrastructure are reshaping AI workloads, enterprise adoption, and efficiency at hyperscale.
Inside Runpod’s GPU Cloud for Long-Running AI Agents
Runpod head of engineering Brennen Smith joins a Data Insights episode to unpack GPU-dense clouds, hidden storage bottlenecks, and a “universal orchestrator” for long-running AI agents at scale.
Walmart and Cornelis Expose the Costs of Network Inefficiency
Billions of customer interactions during peak seasons expose critical network bottlenecks, which is why critical infrastructure decisions must happen before you write a single line of code.
Giga Computing on Rack-Scale AI: From Training to Inference
Recorded at #OCPSummit25, Allyson Klein and Jeniece Wnorowski sit down with Giga Computing’s Chen Lee to unpack GIGAPOD and GPM, DLC/immersion cooling, regional assembly, and the pivot to inference.
Data Insights at OCP: DataraAI on Edge Robotics
Durgesh Srivastava unpacks a data-loop approach that powers reliable edge inference, captures anomalies, and encodes technician know-how so robots weld, inspect, and recover like seasoned operators.
How Hedgehog Brings Hyperscaler Networking to Enterprise AI
Hedgehog CEO Marc Austin joins Data Insights to break down open-source, automated networking for AI clusters—cutting cost, avoiding lock-in, and keeping GPUs fed from training to inference.
Nebius at SC25: Building the Neocloud for Enterprise AI
From SC25 in St. Louis, Nebius shares how its neocloud, Token Factory PaaS, and supercomputer-class infrastructure are reshaping AI workloads, enterprise adoption, and efficiency at hyperscale.
Inside Runpod’s GPU Cloud for Long-Running AI Agents
Runpod head of engineering Brennen Smith joins a Data Insights episode to unpack GPU-dense clouds, hidden storage bottlenecks, and a “universal orchestrator” for long-running AI agents at scale.
Walmart and Cornelis Expose the Costs of Network Inefficiency
Billions of customer interactions during peak seasons expose critical network bottlenecks, which is why critical infrastructure decisions must happen before you write a single line of code.
Giga Computing on Rack-Scale AI: From Training to Inference
Recorded at #OCPSummit25, Allyson Klein and Jeniece Wnorowski sit down with Giga Computing’s Chen Lee to unpack GIGAPOD and GPM, DLC/immersion cooling, regional assembly, and the pivot to inference.
Data Insights at OCP: DataraAI on Edge Robotics
Durgesh Srivastava unpacks a data-loop approach that powers reliable edge inference, captures anomalies, and encodes technician know-how so robots weld, inspect, and recover like seasoned operators.
