Explainers

Evergreen explainers on AI models, labs, chips, data centers, benchmarks, robotics and events.

Gemini 4 Argon research review: Benchmarks & access
AI model explainers ·

Gemini 4 Argon review: strong evaluations, restricted access

Argon has strong published evaluations, but restricted access leaves limited evidence from everyday workflows.

Claude Opus 5.5 research review: Review coverage
AI model explainers ·

Claude Opus 5.5 review: long-workflow gains and missed bugs

Opus catches additional bugs in CodeRabbit’s review pipeline, while missing some issues the baseline caught.

Claude Sonnet 5.5 research review: Effort & token use
AI model explainers ·

Claude Sonnet 5.5 review: strong results, heavy max-effort runs

Sonnet’s review speed and token use vary with the workload and effort setting.

GPT-6.1 Sol research review: Coding & latency
AI model explainers ·

GPT-6.1 Sol review: capable coding, uneven interactive speed

Sol’s published coding gains come with higher token use and mixed evidence on interactive speed.

AI data center explainers ·

AI data center vs traditional data center: what's different

An AI data center runs dense, liquid-cooled accelerator racks networked into one cluster, backed by large power deals.

AI chip explainers ·

Nvidia B200 vs H100: what Blackwell changes

B200 vs H100 from Nvidia's own spec sheets: two dies instead of one, FP4, fifth-generation NVLink and more memory.

AI chip explainers ·

H100 rental price: what clouds charge per GPU-hour

The H100 rental price on the providers' own pricing pages runs from $1.78 to $11.06 per GPU-hour.

Dark navy grid cover reading "Nvidia H100 specs: memory, FP8 and power", with eight cyan bars, each split into a solid dense half and a dashed sparse half
AI chip explainers ·

Nvidia H100: specs, memory and what it is used for

Nvidia H100 specs: 80 GB HBM3 (SXM) or 94 GB (NVL), FP8 dense and sparse, 700 W per GPU, 10.2 kW per DGX H100.

The Model Press

What are you looking for?

Search by headline, topic or keyword.