Robot Brains — Local AI Compute, Measured

Combined archive search across FBI Vault, SPLC, PHMPT, ACIP, NIH grants, the Federal Writers' Project, and other collections we've ingested.

🧠 Ask this record

Get an AI answer drawn only from the Robot Brains — Local AI Compute, Measured documents indexed here — every claim cited back to the filing or proceeding it came from. It will tell you plainly when the record doesn't cover your question, rather than guessing.

4 documents

Tuning an NVIDIA DGX Spark (GB10) to serve many concurrent local LLM agents

Robot Brains — Local AI Compute, Measured · Dgx Spark / Gb10 Inference Tuning

Measured throughput and memory for qwen2.5:7b and qwen3:30b-a3b at 8 to 64 concurrent requests on a GB10 Grace-Blackwell with 122 GiB unified memory. Includes three findings we could not find written anywhere else: the systemd cgroup does not account for unified memory, so a MemoryMax-based watchdog never fires; you must raise the inference server's parallelism before the client's or you are timing your own queue; and a thinking model under format:json can return an empty response until you pass think:false.

What a local research-agent fleet actually delivers

Robot Brains — Local AI Compute, Measured · Agent Fleet / Local Agent Fleet Benchmarks

Operational numbers from 1,372 local LLM agent invocations across 47 research runs in one day on a DGX Spark: wall-clock per run, pages read, evidence yield and answer quality by subject. Includes the figure nobody publishes -- 61% of page-reading agents produced no usable evidence at all -- and the measured threshold at which a cheap text filter removes a quarter of those calls with zero false negatives. Zero external inference cost.

Twelve ways a local AI agent fleet breaks, and how each one was found

Robot Brains — Local AI Compute, Measured · Failure Modes / Local Ai Failure Modes

Failure notes from running a self-hosted LLM research fleet on an NVIDIA DGX Spark (GB10): the wrong diagnoses, the invalid measurements and the guards that exist because something broke. Includes the one that cost a power cycle -- unified memory does not OOM-kill, it stalls the entire machine with no kernel message and no killed process -- plus a cgroup watchdog that cannot see the allocation that kills you, a reasoning model that returns empty output under a JSON grammar, and a GPU sitting at 0% behind a full queue. Eight of the ten are silent degradation, not crashes. Also: a model that is the fastest and highest-quality reader measured and still unusable, because it does not batch -- aggregate throughput flat from 16 to 32 concurrent requests while per-stream speed held at 87 tok/s.

Your benchmark probably cannot tell a 7B from a 32B. Ours could not.

Robot Brains — Local AI Compute, Measured · Evaluation / Benchmarks That Cannot Tell Models Apart

A six-item evaluation set rated four models from 7B to 32B at 5/6 or 6/6 -- a one-item spread across a 4.5x range of model size -- and hid a real defect in a production pipeline. What replaced it: 45 items across 14 named failure categories, deliberately imbalanced because false-accept is the expensive error, scored on balanced accuracy after verifying that always-yes and always-no both land at 50%. On its first run it found the production synthesis model scoring 0/3 on hedged sources -- reading 'reportedly aim to close in June' as support for 'will close in June' -- a qualifier-stripping behaviour the old set could not see because it contained no item of that kind. Also: a unit-mismatch blind spot shared by every model tested, and why that one wants a deterministic check rather than a better model.