Dick Watson
LLNL
The purpose of the two-hour breakout session at the workshop was to obtain feedback from DOE program managers, DOE laboratory managers and scientists, industry, and other attendees.
The breakout session was organized into four subgroups selected at random. Each group had about 25 people in attendance. The groups were asked to consider four main topics:
The discussions in the groups can be briefly summarized under the following headings:
1. There was concensus that the DOE2000 Initiative is valuable.
The number of formal and informal collaborations among DOE laboratories, universities, other agencies, and industry is growing, as is the use of computational modeling. These paradigm shifts in how the DOE community meets DOE missions could be accelerated by the DOE2000 initiative. While ACTS-type technology is a DOE core competence, the participants agreed that missions currently using such technology need significant improvements in capability (e.g., stockpile stewardship, environment, fusion energy, material science), and there are many areas such as environmental remediation for which there is significant untapped potential. It was also recognized that the Collaboratory and ACTS parts of the initiative are tightly interconnected in helping programs reduce cost and improve effectiveness. The goal of DOE2000 is to both use growing commercial off-the-shelf (COTS) products and extend capabilities in areas where such products have limitations such as security, scalability, performance, and functionality.
2. The groups were able to generate many scenarios for DOE2000 capabilities.
The scenarios spanned the DOE scientific mission space in areas such as stockpile stewardship, fusion energy, environment remediation materials, high energy physics, improved automobile engines, combustion research, and materials sciences. In addition, the value of this technology and changed sociology to DOE management information and decision support and technology transfer projects such as AMTEX were also widely recognized. Besides the sharing of large facilities, it was pointed out that the DOE complex also contains many smaller facilities that were created for specific programs that could be useful to a wider community if the type of capabilities represented by DOE2000 were generally available. Further, the ACTS-type tools are commonly developed for specific applications and are not widely available or easily used by multiple applications, and even when groups build such tools for portability there are redundant such efforts. Thus, setting up an ACTS collaboratory would be valuable. There are many cultural, political, and sociological issues needing improvement; and, thus, the DOE2000 vision cannot be achieved just by improvements in technology. because we are talking about interactions among people, people and machines/information, and machines.
3. The DOE2000 functionality must be available across a wide range of cost/performance platforms/networks and be easily used by both people with little interest in and experience with systems as well as those who are experts.
Many university environments are likely to have lower-performance and cost-networks and workstations relative to those available to the DOE laboratories. Further, callaborations with many industrial partners are likely to involve lower performance capabilities. Therefore, the DOE2000 toolbox needs to recognize this reality and need for interoperability. Many if not most of the people who would use these capabilities are not interested in computation technology and will only use DOE2000 functionality if it is easy to use and reliable. The initiative also needs to recognize the growing international collaborations and the implications this has on the above issues.
4. There is a need for high-level DOE and laboratory management support and a high-level champion(s).
This important point was raised in all the groups in many forms. Using the new approaches in a DOE2000 pilot involves risks to Program Managers and the scientists, Managers and scientists willing to take such risks for potential significant gains need to know that such risk taking is both encouraged and will be rewarded . Related to this need was an expressed feeling that congressional approval was also desirable as participating in pilot programs will require some reprograming of resources. Not only was the need for high-level management support felt necessary, but management support is needed at all levels. For example, in the ACTS area it takes more effort to build general-purpose, reusable supported tools than to more narrowly meet a specific programmatic need. Taking these extra steps needs to be recognized as valuable and be rewarded.
5. There is a need for a clearer understanding of the DP/ER DOE2000 technology and infrastructure plans and capabilities as well as what is commercially available.
In order for program managers, principal investigators, and scientists to understand better what pilots might make the most sense, the groups requested a clearer picture of DP/ER plans for core capabilities. The programs also expressed a desire to identify champions who could participate, not only in establishing pilots but in working with DP/ER on requirements and DOE2000 planning of the core technology roadmap.
6. There is a need to clarify the DOE2000 management and funding processes.
Programs expressed the need to be assured that if they committed to participating in the Initiative, the needed capabilities for success would be there, and they would have a voice in the management of the Initiative. Concerns were also expressed about the long-term continuity of funding and functionality. For example, if a program invests in participating, it wants assurance that once it begins to rely on a capability, that capability will be there and be properly supported over the long term.
7. Specific capabilities needed for the scenarios.
We won't make a list of capabilities needed because the ones listed in the groups were generally those anticipated in earlier planning. This conclusion was useful. One new twist was the notion of an "electronic watercooler," namely, telepresence mechanisms to allow the serendipity of running into colleagues in the coffee room, at the water cooler, in the hallway. It was felt that many useful exchanges among team members and others occur in this fashion.