v1.31.0.9

What's Changed

• group partitions based on domain and location_id fields by @biluriuday in #162
• exit dp with non-zero exit code if the driver is not loaded by @biluriuday in #170
• [Fix] fix TensorFlow GPU example to sync with latest image by @yansun1996 in #171
• use libdrm as fallback for fetching product name by @biluriuday in #172

Full Changelog: v1.31.0.8...v1.31.0.9

amd-gpu-helm-0.21.0

A Helm chart for deploying Kubernetes AMD GPU device plugin

v1.31.0.8

What's Changed

• Platform support: RedHat OpenShift 4.19
• Update Helm Chart to 0.20.0 to point to release 1.31.0.7 by @sriram-30 in #138
• modify gpu allocator logic to avoid fragmentation of unused GPUs by @biluriuday in #141
• Bump rocm-docs-core from 1.18.2 to 1.21.1 in /docs/sphinx by @dependabot[bot] in #140

Full Changelog: v1.31.0.7...v1.31.0.8

amd-gpu-helm-0.20.0

A Helm chart for deploying Kubernetes AMD GPU device plugin

v1.31.0.7

What's Changed

• Node labeller vram label support for partitions by @sriram-30 in #108
• Bump rocm-docs-core from 1.17.0 to 1.17.1 in /docs/sphinx by @dependabot in #109
• Bump rocm-docs-core from 1.17.1 to 1.18.1 in /docs/sphinx by @dependabot in #113
• Sphinx config and updated docs for device plugin by @AMD-melliott in #114
• Change SimpleHealthCheck to use /sys/class/kfd by @fluidnumerics-joe in #116
• Bump rocm-docs-core from 1.18.1 to 1.18.2 in /docs/sphinx by @dependabot in #118
• Update rhubi-based image labels for OpenShift certification by @yansun1996 in #120
• Support updateStrategy for helm chart daemonsets by @jaeyung1001 in #121
• Fix k8s-node-labeller cleanup on node labels by @yansun1996 in #123
• Device Plugin and Node Labeller support for gpu partitions by @sriram-30 in #117
• update example device-plugin yaml file by @biluriuday in #124
• add fallback mechanism in case allocator init fails by @biluriuday in #125
• Add documentation for new gpu-partition related node labeller arguments by @sriram-30 in #126

New Contributors

• @sriram-30 made their first contribution in #108
• @AMD-melliott made their first contribution in #114
• @fluidnumerics-joe made their first contribution in #116
• @jaeyung1001 made their first contribution in #121
• @biluriuday made their first contribution in #124

Full Changelog: v1.31.0.6...v1.31.0.7

amd-gpu-helm-0.19.0

A Helm chart for deploying Kubernetes AMD GPU device plugin

v1.31.0.6

What's Changed

• Fixed labeller typo for cu-count

Full Changelog: v1.31.0.5...v1.31.0.6

v1.31.0.5

What's Changed

• Bump rocm-docs-core from 1.15.0 to 1.17.0 in /docs/sphinx by @dependabot in #106
• exporter endpoint svc to check for gpu health by @spraveenio in #100
• remove pulse from base images, add timeout grpc req by @spraveenio in #107

Full Changelog: v1.31.0.4...v1.31.0.5

v1.31.0.4

What's Changed

• node labeller "-product-name" failing for some platforms on k8s by @spraveenio in #104
• Update mounts of labeller in helm chart by @amdlin in #105

Full Changelog: v1.31.0.3...v1.31.0.4

amd-gpu-helm-0.18.0

A Helm chart for deploying Kubernetes AMD GPU device plugin