NPPS Leadership Team

[Torre Wenaus <wenaus AT gmail.com>]

---------- Forwarded message ---------
From: Gao, Haiyan <hgao AT bnl.gov>
Date: Wed, Nov 29, 2023 at 1:33 AM
Subject: NPP Ad hoc committee on Computing and Data Strategic Plan
To: Begel, Michael <begel AT bnl.gov>, Dunlop, James C <dunlop AT bnl.gov>, Mukherjee, Swagato <swagato AT bnl.gov>, Purschke, Martin <purschke AT bnl.gov>, Boyle, Peter <pboyle AT bnl.gov>, Jaffe, David <djaffe AT bnl.gov>, Kirby, Michael <mkirby AT bnl.gov>, Wenaus,Torre J <wenaus AT bnl.gov>, Brown, Kevin A <brownk AT bnl.gov>, Kim, Dohyun <dohkim AT bnl.gov>
Cc: Capasso, Frances <capasso AT bnl.gov>

Dear colleagues:

 

The lab is starting a Strategic Planning process for Computing and Data that expects to take about a year, consisting of three phases. As part of the input (phase I) to this planning process, we are asked to respond to the request below from DDST: a-e.

 

I am writing to appoint a NPP ad hoc committee with Michael and Jamie as two co-chairs. The response deadline is Feb 2, and I would like your committee to complete your responses by Jan 26, 2024.

 

Let me thank you in advance for this great service you are doing for our Directorate and I would also like to thank Jamie and Michael for agreeing to serve as the two co-chairs.

 

Best, Haiyan

 

*****************************************************************************

Haiyan Gao (she/her/hers)

Associate Laboratory Director, Nuclear and Particle Physics

Brookhaven National Laboratory

Building 510, 1-34

Email: hgao AT bnl.gov

Phone: 631-344-2431 (office), 631-461-2867 (cell)

 

*****************************************************************************

 

From: Hill, John <hill AT bnl.gov>
Date: Saturday, November 18, 2023 at 1:46 PM
To: Gao, Haiyan <hgao AT bnl.gov>, Misewich, James <misewich AT bnl.gov>, Schoonen, Martin <mschoonen AT bnl.gov>, Asner, David <dasner AT bnl.gov>, Kleese Van Dam, Kerstin <kleese AT bnl.gov>
Cc: Hewett, JoAnne <jhewett AT bnl.gov>, Emrick, Ann <emrick AT bnl.gov>, Ohlsen, Maria <mohlsen AT bnl.gov>
Subject: Computing and Data Strategic Plan

Haiyan, Jim, Martin, David and Kerstin,

 

As you know I am running a process to look at the way we do data and computer science here at BNL in order to make sure that we are positioned for success.

 

The first step is to perform a self-assessment (Phase I). Phase II will consist of us answering some guiding questions for ourselves, so that we may then write a coherent strategic plan (phase III). I expect the whole process to take something like a year – this is  a lot of work, but it vital that we get this right. Appended below is an outline of Phases I,II and III. I also expect to bring in outside experts to give us advice along the way.

I am writing today to ask to you each to answer the following question for your organization (i.e phase I):

 

1. What do you need from a data science perspective from BNL?

                                                               i.      Compute

                                                             ii.      Storage

                                                           iii.      Infrastructure (software, hardware, services, support, OSS policy,…)

                                                           iv.      AI/ML

                                                             v.      Workforce development in data science

1. Training
2. Recruitment
3. Retention

2. Were there opportunities  you lost out on because we did not have computing capabilities (expertise or access to hardware)
3. What are the key areas where data science will be needed by you in the next 5-10 years - what is needed to be competitive?
4. Please carry out a SWOT analysis for computer science within your organization
5. In the area of scientific data and computing, for each of ITD, CSI and SDCC document:

                                                               i.      What is working

                                                             ii.      What are the barriers to working more effectively with that organization

                                                           iii.      What would be most helpful to you in enabling your science

 

Please answer these questions in a maximum of ten pages. For those of you with large scale facilities within your organizations, or large scale science (e.g. ATLAS), you will want to provide two sets of answers to the above questions. One for the small scale science and one for the large facility-driven science. This could be two separate documents, or delineated within one with a higher page count..

 

Please get these to me by Friday February 2^nd, 2024. If you can get them to me earlier, that would be very helpful, but I understand all that you have going on.

 

If you have any questions now, or as you get to writing, please reach out. Happy to discuss at any time.

 

Thank you very much,

J.

 

 

Phase I: Self-assessment

1. Charge each of ITD, SDCC and CSI with documenting:

1. What is your mission as you currently see it
2. SWOT for your organization in meeting that mission
3. Benchmark your organization against equivalents at other national labs (size, mission, capabilities, impact)
4. Barriers in achieving your mission (external and internal to BNL)

2. Charge Science Directorates: Facilities and Core Depts Separately

1. What do you need from a data science perspective from BNL?

1. Compute
2. Storage
3. Infrastructure (software, hardware, services, support, OSS policy,…)
4. AI/ML
5. Workforce development in data science

1. Training
2. Recruitment
3. Retention

2. Were there opportunities  you lost out on because we did not have computing capabilities (expertise or access to hardware)
3. What are the key areas where data science will be needed by you in the next 5-10 years - what is needed to be competitive?
4. Please carry out a SWOT analysis for computer science within your organization
5. In the area of scientific data and computing, for each of ITD, CSI and SDCC document:

1. What is working
2. What are the barriers to working more effectively with that organization
3. What would be most helpful to you in enabling your science

Phase II: Determine guiding parameters for strategic plan

1. BDISC Brookhaven Data Infrastructure Steering Committee (reconstitute with DDST as Chair)

Charge it with answering the following questions as input for the strategy

1. What is our vision for data at BNL to enable our mission and BNL vision
2. Building on analysis of in 1) and 2), determine areas of computer science R&D where BNL is leading, or poised to take a lead.
3. What are ASCR's priorities in the next 5 years? What about other Directorates within SC? And elsewhere in the federal government
4. Where can BNL have the biggest impact?
5. What are the possible growth areas in Computer Science (if different from ASCR priorities) and how can we prepare for them?
6. Identify role of BNL in

1. Developing AI/ML
2. Utilizing AI/ML

7. Identify unexploited synergies across existing lab efforts in computing.
8. What security posture is required for different use cases?
9. Determine areas of CS R&D required to support BNL mission-driven computing (development of new technologies or frameworks, as opposed to development required to adopt/adapt existing ones)
10. Identify unexploited synergies/engagements with Stonybrook CS (Institute for Advanced Computational Science) or other partners
11. How should BNL interact with the ASCR facilities including HPDF and LCFs
12. Analysis of on-prem vs off-prem for various use-cases
13. Identify areas of  ITD, CSI and SDCC which need improvement in order to support the lab mission successfully

Phase III: Write Strategic plan

 

1. With the input from phases I and II, develop strategy doc to achieve vision. Document should include

1. Vision for data science at BNL
2. Strategy to support to support small experiments and theory
3. Strategy to support large scale facilities
4. Strategy for computer science at BNL
5. Cross-cut strategies

1. Role for HPDF and other LCFs
2. Role of cloud and on-prem compute and storage
3. Cyber security
4. AI/ML strategy
5. Workforce development
6. Infrastructure development

1. Compute
2. Storage
3. Network
4. Policy
5. ..

6. Identify the organizational structure that would best position the lab for achieving its data strategy.
7. Timeline with milestones to execute strategy

 

 

--

-- Torre Wenaus, BNL NPPS Group, ATLAS Experiment

-- BNL 510A 1-222 | 631-681-7892