Microsoft is currently conducting the largest infrastructure buildout that humanity has ever seen.

CXL Is Dead in the AI Era

Groq Inference Tokenomics: Speed, but at What Cost?

TSMC CFET and 3D Stacking, AMD 3D Device, Hynix HBM 4, Micron 3D DRAM, FeRAM

AMD MI300 performance – Faster than H100, but how much?

Microsoft Swallows OpenAI's Core Team – GPU Capacity, Incentives, IP

AMD MI300 Ramp, GPT-4 Performance, Asp and Volumes

Deciphering Latest Restrictions on AI and Semiconductor Manufacturing

TPUv5e: New Benchmark in Cost-Efficient Inference, Training for <200B Parameter

Google Gemini Eats the World

GPT-4 Architecture, Infrastructure, Training Dataset, Costs, Vision, Moe

AMD AI Software Solved – MI300X Pricing, Perf, PyTorch, FlashAttention, Triton

AI Server Cost Analysis – Memory Is the Biggest Loser

Intel Emerald Rapids Backtracks on Chiplets – Design, Performance and Cost

Meta’s Bizarre AI Infrastructure Choice Costs Them $100s Millions

Google “We have no moat, and neither does OpenAI”

The Future of the Transistor

Amazon’s Cloud Crisis: How AWS Will Lose the Future of Computing

Google Apollo: The >$3B Game-Changer in Datacenter Networking

Embracing Chaos: The Imperfect Art of Semiconductor Manufacturing, Lithography

Intel Cuts Employee Pay to Maintain Quarterly Dividend

The Inference Cost of Search Disruption – Large Language Model Cost Analysis

How Nvidia’s CUDA Monopoly in Machine Learning Is Breaking

Lithography Intensity and Long-Term Wafer Demand

Arm Changes Business Model – OEM Partners Must Directly License From Arm

The Semiconductor Heist Of The Century: ARM-China Has Gone Rogue, Operating As An Independent Company With Inhouse IP/R&D

Intel Betting the Farm

The iPhone 13 Does Not Have Satellite Internet, Band N53 and Globalstar

Apple CPU Gains Grind to a Halt and the Future Looks Dim

The Tesla Dojo Chip Is Impressive, but There Are Some Major Technical Issues

More →