Goalque has just released automate-eGPU.sh v1.0.0. This is an incredibly useful tool to provide a functioning eGPU in macOS on “nearly every Mac model, thanks to feedback from eGPU community”.
The tool’s features are summarized below. For full instructions, visit Goalque’s comprehensive post in our Mac Setup forum.
1) Download the script on your desktop: curl -o ~/Desktop/ https://raw.githubusercontent.com/goalque/automate-eGPU/master/automate-eGPU.sh 2) chmod +x ~/Desktop/ 3) cd ~/Desktop 4) sudo ./
What’s new in 1.0.0
A new method for rebuilding cache files
AMD Polaris and Fiji support (limited CUs)
-skipdriver renamed to -skip-web-driver
Removed secondary data scraping methods
New licensing: https://github.com/goalque/automate-eGPU/blob/master/SCRIPT-LICENSE.txt
This script automates Nvidia and AMD eGPU setup on OS X with:
Native AMD support
Detects your OS X product version and build version
Automatic Nvidia web driver download and installation
Automatic IOPCITunnelCompatible mods + Nvidia web driver mod
Detects Thunderbolt connection
Detects your Mac board-id and enables eGPU screen output
Automatic backups with rsync and uninstalling
Detects GPU name by scraping device id from http://pci-ids.ucw.cz
April 2017 Update
In the first week of April, 2017, both Apple and Nvidia made big announcement in regards to the future of the Mac platform. Apple recognized its mistake with the Late 2013 Mac Pro. It vowed to completely redesign its next Mac Pro to make up with the creative professionals. Within two days, Nvidia announced Pascal drivers for macOS. This introduced a whole host of GTX 10 series graphics cards for Mac computers.
The good news is this installation script still works great with Nvidia Pascal cards. We’ve tried GTX 1050 Ti, GTX 1060, GTX 1070, and GTX 1080 Ti with great success. One bad news with AMD cards however. Apple made changes to 10.12.4 which somehow broke AMD eGPU through Thunderbolt 3 connection. This means the Late 2016 MacBook Pro is currently not able to use any AMD GPU through Thunderbolt 3 external enclosures. We are figuring out a fix.
Example implementations using automate-eGPU.sh for installation
Below are benchmark scores on a Late 2016 13″ MacBook Pro comparing eGPU Radeon RX 480 Nitro+ vs. iGPU Intel Iris 550 courtesy of ikir.
Here’s a video showing the successful use of this tool to get a working eGPU under macOS courtesy of theitsage:
Inquire about your eGPU build or share your successful implementation with automate-eGPU.sh v1.0.0 in the Mac Setup forum.