Changelog

New Features

• 3261c74 feat: generate the dashboard images in the README
• 78d490d feat: switch the container images to a shell-less distroless base

Bug Fixes

• 2526dd6 fix(dashboards): label every honest absence instead of showing no data
• e3af660 fix(dashboards): stop states sharing colours and panels clipping
• 080c257 fix(dashboards): tell a missing fan apart from missing data
• 8f0b43c fix: derive clean metric names for fields with unknown unit suffixes

Documentation

• 318defe docs: highlight the NVML backend and tone down the capture request
• f29aa86 docs: refresh the dashboard images

Dependency Updates

• 6a533b6 chore(deps): update dependency goreleaser/goreleaser to v2.17.1
• 897dbef chore(deps): update dependency sigstore/cosign to v3.1.3
• 3ee3dec chore(deps): update docker images
• 17f9f1d chore(deps): update docker images (#480)
• ca255df chore(deps): update docker/dockerfile docker tag to v1.25
• bd7ec4f chore(deps): update docker/dockerfile docker tag to v1.26
• d6f4049 chore(deps): update docker/login-action action to v4.5.2
• 6b29e8e chore(deps): update docker/login-action action to v4.6.0
• 1eb3e07 chore(deps): update goreleaser/goreleaser docker tag to v2.17.1
• a1b4514 chore(deps): update grafana/grafana docker tag to v13.1.3
• cebf51e chore(deps): update grafana/grafana-image-renderer docker tag to v5.12.0

Build, CI & Chores

• 6f4a616 test: fuzz the parsers that read driver and user input

Other Changes

• 5dcd6ec capture: NVIDIA GeForce RTX 2080 SUPER (610.57.04, linux-x86_64)

v1.14.0 Latest

Latest

Compare

Choose a tag to compare

Changelog

Bug Fixes

• 5270339 fix: keep the cgo build flavor out of the scoop manifest
• bc280bf fix: pull in the network target the systemd service waits for
• f05ab81 fix: stop a missing driver entry point from killing the exporter

Documentation

• 050dc92 docs: add an agent guide and refresh the documentation

Dependency Updates

• 99370a8 chore(deps): update actions/checkout action to v7.0.1
• d071ed1 chore(deps): update docker/login-action action to v4.5.0
• a2156b1 chore(deps): update docker/login-action action to v4.5.1
• 3779b66 chore(deps): update grafana/grafana docker tag to v13.1.1
• a07e2f4 chore(deps): update module github.com/prometheus/client_golang to v1.24.0
• a5a4206 chore(deps): update module github.com/prometheus/client_golang to v1.24.1
• 7d296d5 chore(deps): update module github.com/prometheus/common to v0.70.1
• 5569300 chore(deps): update prom/alertmanager docker tag to v0.33.1

Build, CI & Chores

• 9f7bded chore: run the toolchain through docker instead of host binaries

v1.13.1

Compare

Choose a tag to compare

Changelog

New Features

• 6fd9d07 feat: ship GPU health alert rules with the helm chart

Bug Fixes

• b11a7f5 fix: stop a running Windows service before uninstalling it

Documentation

• 4aad7a8 docs: simplify the readme

Dependency Updates

• 5788abf chore(deps): update dependency sigstore/cosign to v3.1.2

v1.13.0

Compare

Choose a tag to compare

Changelog

New Features

• 2907833 feat(nvml): add per-MIG-instance metrics
• d2cb648 feat(nvml): count XID errors and serve the inactive remapped rows
• d55ba77 feat: add NVML metric panels to both dashboards
• 4d87490 feat: add NVML-only metrics and expose the CUDA version
• 3b45f8e feat: add a demo backend serving synthetic data without a GPU
• f9f09ae feat: add a demo flavor to the local dev compose stack
• 3107d66 feat: complete the helm chart rollout, security, and scheduling knobs
• 25e14d1 feat: serve each simulated machine through both dev stack flavors

Dependency Updates

• eb8c566 chore(deps): update actions/setup-go action to v7

Build, CI & Chores

• 85d72fe chore(ci): run the build on pull request base changes
• 15ed0fd chore: name the dev stack data sources after their flavor

v1.12.0

Compare

Choose a tag to compare

Changelog

New Features

• 2f27d6c feat: add a multi-GPU overview dashboard
• f43b5b0 feat: add an experimental NVML collection backend
• 6b078b2 feat: add value ranges and a config file to the fake nvidia-smi
• cbef58d feat: expose GPU recovery action and fabric state as metrics
• 6bea5d0 feat: fluctuate values and simulate multiple GPUs in the fake nvidia-smi
• a354542 feat: harden the http serving and scrape handling
• c251821 feat: let the fake nvidia-smi override field values via --set
• 5482e48 feat: overhaul the Grafana dashboard
• 5640f54 feat: warn when the driver stops providing an NVML function

Bug Fixes

• 594e4f0 fix: record the symbol extraction tool and make the fatal error nil-safe

Documentation

• d6f659c docs: describe the use cases in the readme
• 911c455 docs: update the Grafana dashboard screenshots

Refactors & Internal Changes

• fcf2112 refactor: rework the all-in-one Windows install script

Dependency Updates

• 670f569 chore(deps): update dependency golang/go to v1.26.5
• 72dcd8f chore(deps): update dependency markdownlint-cli to v0.49.1
• f12304c chore(deps): update docker images
• a13e13a chore(deps): update github actions
• 7fa2645 chore(deps): update github actions
• 2918e0c chore(deps): update go packages
• f938c53 chore(deps): update module github.com/prometheus/common to v0.70.0

Build, CI & Chores

• 2feac44 chore: extend the compose dev stack for dashboard development
• 19bdbdc chore: lint the Grafana dashboard with the grafana dashboard-linter
• b51a2ca chore: resume winget publishing
• da0601d chore: satisfy the linters for the cgo build
• f9341ec chore: simulate two nodes in the compose dev stack
• f6977bb test: check driver library symbols against the recorded inventories
• b7e15a2 test: compare the two backends field by field on a real GPU
• 5eef198 test: detect driver drift against the recorded captures
• 47e983a test: tighten the drift checks after review

Other Changes

• de6404e capture: NVIDIA H100 80GB HBM3, plain and MIG (590.48.01, linux-x86_64)

v1.11.0

Compare

Choose a tag to compare

Changelog

New Features

• 2fee27f feat: sign the release checksums, images and charts

Documentation

• bc56dd8 docs: document how to verify release signatures

Dependency Updates

• b6a5aeb chore(deps): update arduino/setup-task action to v3
• 0d67f4b chore(deps): update dependency goreleaser/goreleaser to v2.17.0
• d797b8f chore(deps): update docker images

v1.10.0

Compare

Choose a tag to compare

Changelog

Bug Fixes

• 69e63c0 fix(chart): retry the connection test until the exporter responds
• 9c25603 fix: convert millisecond fields to seconds in metric names
• 23425a0 fix: repair the raw content links consumed at runtime
• 05a2fed fix: support quoted paths in the nvidia-smi command

Documentation

• e574975 docs: clarify test main exit and cell trimming comments
• 96b741f docs: explain how the captures drive the integration tests
• 49daf49 docs: move secondary documentation and community files off the root

Refactors & Internal Changes

• 6897635 refactor: move the exporter entry into an internal package

Build, CI & Chores

• 9eaea5f chore: count cross-package coverage in the CI test run
• 6b67149 chore: gather the host install assets under one directory
• cf0e175 chore: make the lint task reliable and run all of it in CI
• 51787a8 chore: move the dev compose environment and drop the obsolete samples
• ec33534 chore: run CI on every change and validate packaging in it
• 61a9f79 ci: drop the cross-platform test job
• 027de4b test: add a fake nvidia-smi that replays captures
• cc3e608 test: add a parser for the GPU capture files
• 74a7c04 test: add capture-driven integration tests with golden files
• ea8cef9 test: call the expected outputs what they are
• 0eee0d3 test: embed the capture corpus as a Go package
• c47bcb5 test: harden the capture replay for portability and maintenance
• 039437a test: move the integration suite under the internal tree
• 7a48be7 test: support failing selected invocations in the fake nvidia-smi

v1.9.1

Compare

Choose a tag to compare

Changelog

New Features

• 43dedb0 feat: add per-process GPU metrics

Dependency Updates

• 05aef9a chore(deps): update helm/kind-action action to v1.14.0 (#439)

Build, CI & Chores

• 5755d57 chore: add the Artifact Hub repository ID to the chart repo metadata
• d5ea334 test: add nvidia-smi captures from an H200, with and without MIG

v1.9.0

Compare

Choose a tag to compare

Changelog

New Features

• 47f166f feat: add the Helm chart, reworked for runtime-injected GPU access
• dbd7242 feat: publish the helm chart as part of the release
• 810c84b feat: use the nvidia container runtime for GPU access in containers

Documentation

• b5691bc docs: make the helm chart the primary Kubernetes install path

Dependency Updates

• 180cabd chore(deps): update docker/login-action action to v4.4.0 (#437)
• ad700af chore(deps): update github actions (#435)

Build, CI & Chores

• 31c7ffc chore: link the stuck winget pull requests in the pause note
• ab9b249 chore: lint and install-test the helm chart in CI
• 7da5877 chore: pause winget publishing until the package clears review
• 1275d67 test: add nvidia-smi capture from an L40S

v1.8.0

Compare

Choose a tag to compare

Changelog

New Features

• ae70328 feat: decouple metric collection from nvidia-smi execution (#432)

Bug Fixes

• 9796b4e fix: ship the matching architecture binary in docker images

Dependency Updates

• a338ec9 chore(deps): update docker/setup-qemu-action action to v4.2.0 (#434)
• 5e4e4e7 chore(deps): update github actions (#420)
• 75f4a81 chore(deps): update module github.com/prometheus/exporter-toolkit to v0.17.0 (#431)
• a3fe687 chore(deps): update module github.com/prometheus/exporter-toolkit to v0.17.1 (#433)

v1.7.0

Compare

Choose a tag to compare