Dr. Arman Khalatyan — Bloghttps://arm2arm.github.io/blog/2026-11-10T00:00:00ZDr. Arman KhalatyanEight GPUs, Two Generations: Benchmarking Open LLMs for an Astronomy Institutehttps://arm2arm.github.io/blog/benchmarking-open-llms-on-a-mixed-gpu-server-10-11-2026.html2026-11-10T00:00:00ZWe measured open-weight language models on a mixed H200 / RTX PRO 6000 server and built AstroBench-AIP, a benchmark of 216 real astronomy computing tasks. The serving engine matters more than GPU count, and a 27B model on two GPUs serves 64 users at about 1 J per token.Redefining Research: When AI Does the Science, Who Becomes the Scientist?https://arm2arm.github.io/blog/redefining-research-ai-and-scientific-expertise-07-11-2026.html2026-11-08T00:00:00ZAI can accelerate discovery. But if students delegate the work through which expertise is learned, who will recognize when machines are wrong?