AIHow AI Works

Tag: GPU

Nemotron 3.5 Lightning: One GPU, Yes. A Laptop, No.

NVIDIA's 30B Nemotron 3.5 Lightning runs on a single desktop GPU, but its Q4_K_M GGUF is 25.3 GB — above every 24 GB mobile card.

NVIDIA’s $12.9B Hugging Face Deal Is a Bet on Open Models

NVIDIA will acquire Hugging Face for $12.9 billion. Why the deal's incentives favor open models — and where the neutrality risk actually sits.

Self-Hosted vs API: When Running Your Own LLM Pays Off

The raw GPU price is a trap. A 2026 breakdown of the real cost of self-hosting an LLM and where the break-even against API pricing actually sits.

From Token to Transistor: What Happens When You Send a Prompt

Follow one prompt through the machine — tokens, matrix multiplications, the memory that everything waits on, and the chips built around a single operation.

Training vs Inference: Two Economies, Two Opposite Bottlenecks

Training is network-bound and paid once. Serving is memory-bound and paid forever. The crossover point, and why the 2024 chip-demand call was backwards.

What One AI Query Costs: Tokens, Bytes, Joules, Cents

One question, one model, one machine — costed end to end. Then four dials that move the same answer by more than 10,000x.

Why NVIDIA’s Moat Is Software, and What Would Break It

Rivals have shipped comparable silicon for years without shifting the market. The moat is the software above the chips and the rack around them.

TPUs vs GPUs vs ASICs: AI Hardware Guide 2025

Complete guide to TPUs, GPUs, and ASICs for AI workloads. Compare architectures, performance, efficiency, and market trends as of December 2025.

China Claims 14nm Chips Rival NVIDIA’s 4nm

China's new 14nm processor with 18nm DRAM achieves 120 TFLOPS, outperforming NVIDIA A100 GPUs and addressing memory bandwidth challenges.

Google TPUv7 Ironwood: Challenging Nvidia

Google's TPUv7 Ironwood commercializes AI chips externally. Anthropic's 1M TPU order signals potential end to Nvidia's CUDA dominance.