Update hardware-failure.md

Changed the gpu-burn example to use tc instead of d to create more of a workload when testing.

content/hardware-failure.md

@@ -156,7 +156,7 @@ make
 And now we can run it like so (this example will run it for 60 minutes/1 hour):
 
 ```bash
-./gpu_burn -d 3600
+./gpu_burn -tc 3600
 ```
 
 ## Test CPU thermals