Add a new notebook for SNMG benchmark runs

[naimnv]

This PR adds new notebook for SNMG benchmark runs. It

• simplifies bookkeeping and data presentation.
• update each cugraph function wrappers to handle both SG and MG calls depending on a flag. The object is to have a common set of function wrappers for SG, SNMG and MNMG and put them in a common module to avoid duplication.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[acostadon]

Looks great to me.
just need to add your changes to the notebook credits and add 2024 to the copyrights at the bottom,
I ran it on my workstation up to scale 24 on a 2-GPU workstation and it ran perfectly.
Running on scale 25 now.

[BradReesWork]

You need to create a new Notebook and not just add MG to this notebook. The issue is that the notebook compare against NetworkX and Nx does not support DASK or any distributed data. You would need to update the code to always collect that data back to a single host.

This will not work in MG
gdf = generate_data(v)
pdf = gdf.to_pandas()

dask_cudf does not have a "to_pandas" function

[BradReesWork]

Needs a new Notebook for MG

[naimnv]

Sure. We decided not to merge it last week.

[naimnv]

This will not work in MG
gdf = generate_data(v)
pdf = gdf.to_pandas()

dask_cudf does not have a "to_pandas" function

@BradReesWork Just to let you know, the updated notebook's generate_data has mg=Flase by default.
def generate_data(scale, edgefactor=16, mg=False):

From correctness point of view, this code will always work.

But I agree that we need to collect the benchmark graphs back to single host and run Nx algorithms on them.

[naimnv]

Early next week, I can work on this again to make a Notebook for MG.

[naimnv]

@BradReesWork @acostadon
This PR has been updated to include a new notebook for SNMG benchmark runs. I will go over it again, but early review is always valuable.

[BradReesWork]

/ok to test

[naimnv]

/ok to test

[naimnv]

/ok to test

[naimnv]

/ok to test

[BradReesWork]

/merge