WebGPU on EC2 Windows - requestAdapter Failed to create WebGPU Context Provider

[wenhan chong]

Hi All,

While trying to run Chrome with WebGPU on EC2 Windows instances unattended, I noticed inconsistent behaviors for WebGPU. Each browser session is a coin flip between having and not having WebGPU. The code fails to get adapter from navigator.gpu.requestAdapter. Errors show "Failed to create WebGPU Context Provider". This only happens when I run the tests unattended. After connecting to the EC2 instances through remote desktop, the same set of automated tests are able to render WebGPU consistently without issues. I also noticed several error messages including GPU process crash in chrome://gpu when running unattended. I have attached dxdiag output and chrome://gpu reports for both unattended sessions and remote desktop connected sessions. 
Trying to understand what is causing the inconsistency and if there are additional options required to run WebGPU on unattended EC2 windows instance.

Regards,
Wen Han

[wenhan chong]

Attaching additional output from chrome://gpu showing errors from failure in unattended browser sessions. Some of the browser sessions are also not able to detect nvidia T4s, falling back instead to ANGLE swiftshader device.