They summarize our coverage. We write it.
Newsletters like this one rebroadcast our headlines - often without the full review, the source reading, or the analysis underneath. Our weekly briefing sends the work they paraphrase, straight from the desk, before they get to it.
Free, weekly, no spam. One email every Tuesday. Unsubscribe anytime.
LG AI Research released EXAONE 4.5, a 33B open-weight vision-language model that posts higher STEM scores than GPT-5-mini and Claude 4.5 Sonnet - but a non-commercial license caps its real-world reach.
Intel's Arc Pro B70 launched on March 25 with 32GB GDDR6 and 367 TOPS for $949, undercutting NVIDIA's RTX Pro 4000 by $850. The hardware case is strong. The software story is not.
ByteDance's DeerFlow 2.0 is a powerful open-source agent harness that executes long-horizon tasks inside Docker sandboxes - impressive engineering, but not a turnkey solution.
NVIDIA's new Nemotron-Cascade-2-30B-A3B activates just 3B parameters per token, runs on a single RTX 4090, and outscores NVIDIA's own 120B model on coding and math benchmarks.
Apple's cheapest Mac ever packs the A18 Pro iPhone chip with a 16-core Neural Engine - but its 60 GB/s memory bandwidth puts a hard ceiling on what local models you can actually run.
Apple launches M5 Pro and M5 Max MacBook Pros with Neural Accelerators in every GPU core, 128GB unified memory, and 614GB/s bandwidth - enough to run Llama 70B on a laptop.
A developer cracked Apple's undocumented ANE private APIs, measured its real throughput at 19 TFLOPS FP16 (not the marketed 38 TOPS), and trained a 109M-parameter transformer on hardware Apple designed exclusively for inference.
macOS RDMA over Thunderbolt 5 has turned four Mac Studios into a 1.5TB unified memory cluster that runs Kimi K2 at 25 tokens per second - a setup that would cost $780K with NVIDIA H100s.
Full specs and benchmarks for the Apple M4 Max SoC - up to 128GB unified memory at 546 GB/s, 3nm process, and why it has become the quiet favorite for running 70B+ models locally.
Full specs and benchmarks for the NVIDIA GeForce RTX 3090 - 24GB GDDR6X at 936 GB/s, Ampere architecture, and why used 3090s remain the best value option for local AI inference in 2026.
Full specs and benchmarks for the NVIDIA GeForce RTX 4090 - 24GB GDDR6X, 1,008 GB/s bandwidth, Ada Lovelace architecture, and why it remains the default home lab GPU for local AI inference.
Full specs and benchmarks for the NVIDIA GeForce RTX 5090 - 32GB GDDR7, 1,792 GB/s bandwidth, Blackwell architecture, and what it means for local AI inference.
