Snapdragon hw acceleration#8
Open
eoinjordan wants to merge 7 commits intomasterfrom
Open
Conversation
Chocolatey ImageMagick on Windows runners lacks PNG codec delegates, causing prepare-branding.ps1 to fail with: magick.exe: no decode delegate for this image format Fix: generate header.bmp (150x57) and welcome.bmp (164x314) on macOS using Homebrew ImageMagick from img/linux-collection.png, commit them to the repo, and remove the Install ImageMagick and Generate installer branding assets CI steps entirely. BMPs will be regenerated locally when the source image changes.
…ect graph image The previous header.bmp and welcome.bmp were generated from img/linux-collection.png which is a raw sensor data graph screenshot, not the Edge Impulse brand logo. Regenerate from studio/public/themes/default/png/logo.png (same source as the serial-daemon installer). Also update prepare-branding.ps1 to look for img/edge-impulse-logo.png first, falling back to the monorepo logo path.
- Add detect-snapdragon.ps1 script to identify Qualcomm devices and QNN runtime - Integrate Snapdragon detection with installer staging pipeline - Bundle detection helper in installer for post-install verification - Add comprehensive Snapdragon section to README with setup/verification steps - Support automatic CPU fallback on unsupported devices This enables Windows ARM64 users on Snapdragon processors to leverage: * Neural network inference via Snapdragon NPU * DSP acceleration for signal preprocessing * Model compilation optimizations * Graceful fallback to CPU mode if QNN unavailable Supported devices: Windows 11 ARM64 on Snapdragon, RB3 Gen 2, IQ-9
eoinjordan
force-pushed
the
snapdragon-hw-acceleration
branch
from
0366f92 to
96f3df5
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request introduces a comprehensive Windows offline installer for the Edge Impulse Linux CLI, enabling installation on Windows devices without requiring
npm, build tools, or internet access to npm. It adds a full GitHub Actions build workflow, Windows-specific installer scripts, and documentation, with special support for ARM64 Snapdragon hardware acceleration.The most important changes are:
Windows Installer Infrastructure:
.github/workflows/build-windows-installer.ymlto automate building Windows installers for x64 and ARM64 via GitHub Actions, producing artifacts and optionally uploading releases.windows-installer/installer.nsi(NSIS script) to create a signed, branded Windows installer that bundles Node.js, prebuilt modules, and CLI shims, and manages PATH/uninstall cleanly.stage.ps1), branding (prepare-branding.ps1), and hardware detection (detect-snapdragon.ps1) to automate packaging and ARM64 optimization detection. [1] [2] [3].gitignoreto exclude installer build output and generated branding assets.Documentation:
README.mdwith a new section on using the Windows offline installer, outlining included components, requirements, and instructions for downloading artifacts.Hardware Acceleration Support:
detect-snapdragon.ps1to check for ARM64 Snapdragon devices and QNN runtime presence, enabling hardware acceleration for neural network inference on supported platforms.These changes make the Edge Impulse Linux CLI easily installable on Windows, especially in restricted environments, and provide enhanced performance on ARM64 Snapdragon devices.