Skip to main content

Backend.AI Accelerator Plugin for Intel® Gaudi® HPUs

Project description

Backend.AI Accelerator Plugin for Habana

Just install this along with Backend.AI agents, using the same virtual environment. This will allow the agents to detect Habana devices on their hosts and make them available to Backend.AI kernel sessions.

$ pip install -e <Path to this repository>

This open-source edition of Habana plugins support allocation of one or more Habana devices to a container, slot-by-slot.

Compatibility Matrix

TBD

Generating Docker macvlan interfaces for inter-container network

#! /bin/bash
set -e
for pci_bus_id in $(hl-smi --format=csv,noheader -Q bus_id); do
    for net_dev in $(grep PCI_SLOT_NAME /sys/class/net/*/device/uevent | grep "${pci_bus_id}" | awk -F: '{print $1}' | awk -F/ '{print $5}'); do
        docker network create --attachable -o parent="${net_dev}" -d macvlan "gaudi_${net_dev}" >/dev/null
        echo "Docker network gaudi_${net_dev} created"
    done
done

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

backend_ai_accelerator_habana-26.3.0.tar.gz (8.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

backend_ai_accelerator_habana-26.3.0-py3-none-any.whl (10.8 kB view details)

Uploaded Python 3

File details

Details for the file backend_ai_accelerator_habana-26.3.0.tar.gz.

File metadata

File hashes

Hashes for backend_ai_accelerator_habana-26.3.0.tar.gz
Algorithm Hash digest
SHA256 bd416460034be54444632a96c31fa26b7a659654e7a30446e0f0a73b2faff184
MD5 ffa7429f5f28ad83a29d399e805a6d7c
BLAKE2b-256 9497a5b9e7fbb93af028e373332ab5c0f80657216879b8630c591448838484e4

See more details on using hashes here.

File details

Details for the file backend_ai_accelerator_habana-26.3.0-py3-none-any.whl.

File metadata

File hashes

Hashes for backend_ai_accelerator_habana-26.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 81db6596b4b29283626310041250e042a0e8e656c1d12424d616c94a902e31b1
MD5 11f509866498c3904f5552e472c90fcd
BLAKE2b-256 f7542b808beb59fa0dbed0953393f5b42c7cf704741677da262bcedeaeca382d

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page