Replaces the stale 40.5 placeholder (a llama3.2:1b CPU-era number) with 13.2,
measured on the Reference Build (NOMAD6, 780M) under the v2 AI harness with
llama3.1:8b and iGPU acceleration (OLLAMA_IGPU_ENABLE — the provisioning fix in
PR #1074). This is the AI channel's reference for the uncapped v2 score.
Must ship together with #1074 in v1.34.0: the reference assumes iGPU-accelerated
AMD boxes, so shipping it without the iGPU fix would score AMD installs on CPU
numbers against a GPU bar. The exact value gets a final confirm against the
shipping Ollama config before GA.
Must stay byte-identical to the leaderboard's score_service.ts REFERENCE_SCORES_V2
(the server recomputes the score on submit); the matching leaderboard change is a
sibling PR. Changing the leaderboard reference has no effect on existing rows —
the v1->v2 backfill neutralizes the AI channel (ratio 1 regardless of the
reference), and there are no real v2 submissions yet.
Sibling to the Score v2 app client (#1094); stacked on it (same file/constant).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Phase 4 of the NOMAD Score v2 effort (stacked on #1089). Captures the raw
channel values the leaderboard scores from, computes the uncapped v2 score
in-app (byte-matching the leaderboard's server-side recompute), submits the
v2 payload, and surfaces it in the benchmark UI.
- Capture raws: per-channel sysbench values + O_DIRECT disk (W4), single/multi
thread CPU + memory, W6 consistency companions, and run-environment metadata
(run_environment, storage_path_type, gpu_compute_detected, #1016).
- AI hardening: reference model llama3.1:8b, num_predict=256, VRAM eviction
before the run + unload own model after, pre-flight disk check (~6.5GB, only
when the model is uncached), median-of-3 run (W7).
- Score: _calculateNomadScoreV2 (uncapped, reference->1000) byte-matches the
leaderboard recompute. nomad_score_v2 stays null unless a full run with AI.
- Payload: submitToRepository sends the v2 raws + score alongside legacy v1.
- UI: v2 headline score, legacy v1 as secondary, raws/env in details.
- Migration: 13 nullable columns; double (not knex float(8,2)) so millions-scale
raws keep full precision and don't break the leaderboard byte-match.
Reference constant REFERENCE_SCORES_V2.ai_tokens_per_second is still the stale
40.5 placeholder (TODO in source); locking it to the measured 13.2 is a
follow-up gated on the AMD iGPU fix (#1074). Live prod v2 submit is separately
blocked by the leaderboard TTFT floor (tracked in the leaderboard follow-up).
Verified on NOMAD3 (RTX 5060) dev env: migration applies, raws recorded,
v2 score computed, backend + benchmark.tsx typecheck clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Score v2 Phase 1. Safe under v1 — no scoring, weight, reference, or
submission-payload changes; only the benchmark's failure behavior and
forensic metadata.
- Fail-on-parse-miss (W3): the four SCORED sysbench metrics (CPU
events/sec, memory ops/sec, disk read/write MiB/s) now THROW when the
regex misses or the value is <= 0, instead of silently returning 0.
A parse failure (e.g. an upstream image output-format change) now
fails the run with a clear error via the existing _runBenchmark
try/catch, rather than submitting a phantom zero sub-score. Secondary/
informational fields keep their existing defaults.
- Pin sysbench by digest (W3): severalnines/sysbench@sha256:64cd003b...
(was :latest), so a latest-tag format change can't break the parsers
fleet-wide. Digest validated on the NOMAD6 reference build.
- Record provenance (W7): sysbench_digest + ollama_version (from Ollama
/api/version, null-tolerant) stored on each result. New nullable
columns + additive migration.
Verified: pinned digest pulls/runs/parses on NOMAD3; migration applies
(columns present); typecheck clean.
Part of the NOMAD Score v2 effort.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Every Docker pull went through `followProgress(pullStream, resolve)`, passing
the Promise's resolve as dockerode's onFinished(err, output) callback — so the
error argument was ignored and a failed pull (dropped/metered connection, bad
manifest, registry error, disk full mid-pull) resolved as if it had succeeded.
The code then tried to create/start a container from a missing or partial
image, surfacing a confusing downstream error rather than the real cause. (#790)
Add a DockerService.pullImage() helper that rejects when followProgress reports
an error, and route all five pull sites through it:
- service install
- AMD ROCm image pull
- service update
- force-reinstall / recreate (forcePull)
- sysbench benchmark image pull
Closes#790
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Closes#810.
## Bug A: HSA_OVERRIDE_GFX_VERSION=11.0.0 was unconditional
PR #804 set HSA_OVERRIDE_GFX_VERSION=11.0.0 for any AMD GPU. The inline comment
claimed this was harmless on supported discrete cards (gfx1030 RX 6800, etc.) — empirically false. With the override, Ollama crashes during GPU discovery on
gfx1030 and falls back to CPU silently. Affects every NOMAD user with an
RX 6800 or other RDNA 2 discrete card.
The correct value depends on the gfx version:
- gfx1030, gfx1100, gfx1101, gfx1102: officially supported by ROCm — no override
- gfx1031..gfx1036 (RDNA 2 variants + iGPUs like Rembrandt 680M): 10.3.0
- gfx1103, gfx1150, gfx1151 (Phoenix 780M, Strix 890M, Strix Halo): 11.0.0
### Resolution chain in `_resolveAmdHsaOverride()`
1. KV `ai.amdHsaOverride` — manual override; accepts 'none' to disable, or a
semver-style value to force.
2. Marker file `/app/storage/.nomad-amd-gfx` — written by install_nomad.sh
based on lspci codename. Mapped to override via `_mapGfxToHsaOverride()`.
3. Default: `11.0.0` — preserves prior behavior so existing iGPU users
(780M / 890M, the dominant AMD population today) don't regress on upgrade.
Discrete RDNA 2 users on existing installs can opt out via
`ai.amdHsaOverride='none'` and force-reinstall AI Assistant, OR re-run
install_nomad.sh to refresh the marker file.
The helper is used in both `createContainer` (initial install) and
`updateContainer` (image update) paths, replacing the unconditional push.
## Bug B: BenchmarkService had no AMD discrete detection path
`BenchmarkService.getHardwareInfo()` had three GPU detection fallbacks:
1. `si.graphics()` — empty inside Docker for AMD
2. nvidia-smi — NVIDIA only
3. AMD APU regex from CPU model — integrated only
Result: AMD discrete cards (RX 6800, RX 7900 XTX, etc.) showed up as
"GPU: Not detected" on the leaderboard despite ROCm working. Corrupts
leaderboard data quality for that population.
Fix: after the existing fallbacks, call `SystemService.getSystemInfo()` and
read `graphics.controllers[0].model`. That path already handles AMD via the
marker file + Ollama log probe added in PR #804, so we're reusing existing
plumbing rather than duplicating detection logic.
## install_nomad.sh changes
The existing AMD detection block already runs lspci. Added a codename parse
step that maps Navi 21/22/23/24, Rembrandt, Phoenix1/Phoenix2, Strix/Strix
Point/Strix Halo, and Navi 31/32/33 to gfx versions, then writes
`/opt/project-nomad/storage/.nomad-amd-gfx`. Unknown codenames write nothing
(admin handles missing-marker case via the backward-compat default).
## Validation
Both bugs were originally surfaced and validated empirically on RX 6800 /
gfx1030 / Ubuntu 24.04 + kernel 6.17 + ollama/ollama:rocm during the #810
filing. Validation grid from that report:
| Run | NOMAD Score | tok/s | GPU detected |
|-----------------------------------------------|-------------|-------|-------------------------|
| Pre-fix (Bug A active) | n/a | 0 | yes, but library=cpu |
| HSA_OVERRIDE removed, Bug B unfixed | 73.8 | 221.6 | "Not detected" |
| Both fixes hot-patched (this PR's behavior) | 73.7 | 216.0 | AMD Radeon RX 6800 |
Local checks: `npm run typecheck` clean, `npm run build` clean.
The log2 normalization formula `50 * (1 + log2(ratio))` produces negative
values (clamped to 0) whenever the measured value is less than half the
reference. For example, a CPU scoring 1993 events/sec against a 5000
reference gives ratio=0.4, log2(0.4)=-1.32, score=-16 -> 0%.
Fix by dividing log2 by 3 to widen the usable range. This preserves the
50% score at the reference value while allowing below-average hardware
to receive proportional non-zero scores (e.g., 28% for the CPU above).
Also adds debug logging for CPU sysbench output parsing to aid future
diagnosis of parsing issues.
Fixes#415
Rotate the HMAC secret used for signing benchmark submissions to the
community leaderboard. The previous secret was compromised (hardcoded
in open-source code and used to submit a fake leaderboard entry).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The previous lspci-based GPU detection fails inside Docker containers
because lspci isn't available, causing Ollama to always run CPU-only
even when a GPU + NVIDIA Container Toolkit are present on the host.
Replace with Docker API runtime check (docker.info() -> Runtimes) as
primary detection method. This works from inside any container via the
mounted Docker socket and confirms both GPU presence and toolkit
installation. Keep lspci as fallback for host-based installs and AMD.
Also add Docker-based GPU detection to benchmark hardware info — exec
nvidia-smi inside the Ollama container to get the actual GPU model name
instead of showing "Not detected".
Tested on nomad3 (Intel Core Ultra 9 285HX + RTX 5060): AI performance
went from 12.7 tok/s (CPU) to 281.4 tok/s (GPU) — a 22x improvement.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When running in Docker, the systeminformation library cannot see the
host's GPU hardware. This adds a fallback to detect Intel integrated
graphics from the CPU model name, similar to how we handle AMD APUs
with Radeon graphics.
Intel Core Ultra processors (Meteor Lake, Arrow Lake) include Intel
Arc Graphics integrated. This change detects "Core Ultra" in the CPU
brand and reports "Intel Arc Graphics (Integrated)" as the GPU model.
Note: This is for display purposes only - Ollama does not support
Intel integrated graphics for AI acceleration.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* feat(benchmark): Require full benchmark with AI for community sharing
Only allow users to share benchmark results with the community leaderboard
when they have completed a full benchmark that includes AI performance data.
Frontend changes:
- Add AI Assistant installation check via service API query
- Show pre-flight warning when clicking Full Benchmark without AI installed
- Disable AI Only button when AI Assistant not installed
- Show "Partial Benchmark" info alert for non-shareable results
- Only display "Share with Community" for full benchmarks with AI data
- Add note about AI installation requirement with link to Apps page
Backend changes:
- Validate benchmark_type is 'full' before allowing submission
- Require ai_tokens_per_second > 0 for community submission
- Return clear error messages explaining requirements
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix(benchmark): UI improvements and GPU detection fix
- Fix GPU detection to properly identify AMD discrete GPUs
- Fix gauge colors (high scores now green, low scores red)
- Fix gauge centering (SVG size matches container)
- Add info tooltips for Tokens/sec and Time to First Token
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix(benchmark): Extract iGPU from AMD APU CPU name as fallback
When systeminformation doesn't detect graphics controllers (common on
headless Linux), extract the integrated GPU name from AMD APU CPU model
strings like "AMD Ryzen AI 9 HX 370 w/ Radeon 890M".
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* feat(benchmark): Add Builder Tag system for community leaderboard
- Add builder_tag column to benchmark_results table
- Create BuilderTagSelector component with word dropdowns + randomize
- Add 50 adjectives and 50 nouns for NOMAD-themed tags (e.g., Tactical-Llama-1234)
- Add anonymous sharing option checkbox
- Add builder tag display in Benchmark Details section
- Add Benchmark History section showing all past benchmarks
- Update submission API to accept anonymous flag
- Add /api/benchmark/builder-tag endpoint to update tags
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* feat(benchmark): Add HMAC signing for leaderboard submissions
Sign benchmark submissions with HMAC-SHA256 to prevent casual API abuse.
Includes X-NOMAD-Timestamp and X-NOMAD-Signature headers.
Note: Since NOMAD is open source, a determined attacker could extract
the secret. This provides protection against casual abuse only.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
- Add OLLAMA_API_URL environment variable for Docker networking
- Use host.docker.internal to reach Ollama from NOMAD container
- Add extra_hosts config in compose for Linux compatibility
- Add downloading_ai_model status with clear progress indicator
- Show model download progress on first AI benchmark run
- Fail AI-only benchmarks with clear error if AI unavailable
- Display benchmark errors to users via Alert component
- Improve error messages with error codes for debugging
Fixes issue where AI benchmark silently failed due to NOMAD container
being unable to reach Ollama at localhost:11434.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add Benchmark to Settings sidebar navigation
- Fix Luxon DateTime bug when saving submission timestamp
- Add privacy explanation text before Share button
- Add error handling and display for failed submissions
- Show "Submitting..." state and success confirmation
- Add link to view leaderboard after successful submission
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add comprehensive benchmarking capability to measure server performance:
Backend:
- BenchmarkService with CPU, memory, disk, and AI benchmarks using sysbench
- Database migrations for benchmark_results and benchmark_settings tables
- REST API endpoints for running benchmarks and retrieving results
- CLI commands: benchmark:run, benchmark:results, benchmark:submit
- BullMQ job for async benchmark execution with SSE progress updates
- Synchronous mode option (?sync=true) for simpler local dev setup
Frontend:
- Benchmark settings page with circular gauges for scores
- NOMAD Score display with weighted composite calculation
- System Performance section (CPU, Memory, Disk Read/Write)
- AI Performance section (tokens/sec, time to first token)
- Hardware Information display
- Expandable Benchmark Details section
- Progress simulation during sync benchmark execution
Easy Setup Integration:
- Added System Benchmark to Additional Tools section
- Built-in capability pattern for non-Docker features
- Click-to-navigate behavior for built-in tools
Fixes:
- Docker log multiplexing issue (Tty: true) for proper output parsing
- Consolidated disk benchmarks into single container execution
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Chris Sherwood
Bortlesboat
Jake Turner
chriscrosstalk