ECAS Exploring Clouds for Acceleration of Science

[Jamie Sunderland <>]

Thanks Benedikt,

 

I did raise that exact issue with NSF.   I actually mentioned that “(for example) the IceCube project might want to use CloudBank”, and there could be an inconsistency (for example again) if SDSC was one of the awardees and they did not recover indirects on cloud spend, while the other awardee was recovering indirects because of their institutional policies. Would this SDSC /UCSD simply miss out on up to $250k, or do they have an extra $250k to spend on staff or do they have $750k to spend on cloud?  Are we even allowed to change the policy with regards to recovering indirect on cloud half-way through E-CAS? Is there a difference because we are under a cooperative Agreement with NSF rather that a standard award?

 

..as you can see it is a complicated precedent laden question, which is why NSF need a bit of time to consider.

 

Jamie.

 

--

Jamie Sunderland

 

From: Benedikt Riedel <>
Date: Wednesday, April 8, 2020 at 10:03 AM
To: Jamie Sunderland <>
Cc: "Aliaga, Daniel G" <>, "" <>
Subject: Re: [ecas] Meeting

 

Hi Jamie,

 

Another question would be if we can make a sub-award out of the cloud funds, i.e. move the cloud $ to for example UCSD to reduce the overhead cost.

 

Thanks,

 

Benedikt

 

On Wed, Apr 8, 2020 at 9:47 AM Jamie Sunderland <> wrote:

I have raised the staffing issue with NSF.

• I don’t think we can take money from the cloud-spend for staffing, because the cloud provider contributed resources for phase 1 on the understanding that there would be a certain level of actual cash-spend during phase 2.  It may however be OK to include a small amount  for cloud management software like CloudCheckr if you feel you want/need it.
• I have asked NSF whether we can exclude indirect costs from the cloud spend.  This would have the effect of feeing up money that would have otherwise been taken up by university administration to be used to fund further RA resources.  I believe most universities charge indirects on cloud spend – but there are a few exceptions such as UCSD, U.Washington and the CloudBank Project.  This is a complicated matter and NSF have asked for some time to consider it.

Therefore, in preparing your budgets, I would recommend you work out estimated use of cloud resources and the staff resources you would like to have for phase 2 – but we cannot confirm more than 1 RA at this stage.

 

Regards, Jamie.

 

--

Jamie Sunderland

 

From: "Aliaga, Daniel G" <>
Date: Wednesday, April 8, 2020 at 9:06 AM
To: Jamie Sunderland <>
Cc: "''" <>
Subject: RE: Meeting

 

Wonderful! Any update on whether some of the $500k towards cloud credit can be used for RA/staff? (a small portion, say 10% or so)

 

D.

 

 

From: Jamie Sunderland [mailto:]
Sent: Wednesday, April 8, 2020 9:59 AM
To: Aliaga, Daniel G <>
Cc:
Subject: Re: Meeting

 

(Copying this to the group)

 

Hi Daniel,

 

Yes, I was planning a brief meeting today:  If you can’t make it – just please send me a brief report for March.

Templates and submission guide attached.

 

They can also be found in the Google drive. https://drive.google.com/open?id=1w1xfWfwKe2wwNX0k2HCwqjHSDFnHc7yq

 

Report is due 30^th of April.

 

Thanks., Jamie.

 

--

Jamie Sunderland

 

From: "Aliaga, Daniel G" <>
Date: Wednesday, April 8, 2020 at 8:53 AM
To: Jamie Sunderland <>
Subject: Meeting

 

 

Jamie,

 

Hope all is well and thanks for leading the great talks last week, it was great to hear about the diversity of projects!

 

1. Is there a 2pm meeting today?
2. Can you send the template for the report?
3. When exactly is the report due?

 

Stay healthy!

 

-Daniel.

 

 

---------------------------------------

Daniel G. Aliaga

Associate Professor

Department of Computer Science

www.cs.purdue.edu/~aliaga

www.cs.purdue.edu/cgvlab

 

--

Benedikt Riedel

Global Computing Coordinator IceCube Neutrino Observatory

Computing Manager Wisconsin IceCube Particle Astrophysics Center

University of Wisconsin-Madison