[GraphBolt][Regression] MultiGPU benchmark on papers100M

[mfbalin]

🔨Work Item

IMPORTANT:

• This template is only for dev team to track project progress. For feature request or bug report, please use the corresponding issue templates.
• DO NOT create a new work item if the purpose is to fix an existing issue or feature request. We will directly use the issue in the project tracker.

Project tracker: https://github.com/orgs/dmlc/projects/2

Description

Let's run the multiGPU graphbolt benchmark with the following additional configurations for differing numbers of GPUs:

1. --dataset=ogbn-papers100M --gpu-cache-size=5000000
2. --dataset=ogbn-papers100M --gpu-cache-size=0

[mfbalin]

@Rhett-Ying