The IT-Division will directly improve the strength of science at LBNL, from experimental design through results verification and reproducibility, through an integrated, efficient Science Services organization. The main Science-IT program is called Science Accelerator.
Science IT offers Lab Science Staff, Students and Affiliates support via five main service offerings. Wherever possible, support is provided free or at the lowest possible cost, leveraging Laboratory buying power, UC collaborations, etc.
FREE Consultation, contact us!
Science Service Objectives:
- Create, Grow, Deploy, and Sustain a high-value Science IT consulting capability, consisting of highly qualified staff, effective engagement program management, cutting edge portal capabilities, and innovative methods to find and fulfill the needs of Laboratory scientists.
- Create, Grow, Deploy and Sustain scientific data management, storage, and orchestration capabilities supporting both the needs of scientific programs across the science activity lifecycle, and the long-term interests of the Laboratory as a central hub for global science data and research results.
- Grow, Deploy and Sustain access to flexible, affordable, useful compute capabilities tailored to the needs of scientists who may in future, but currently do not, or cannot, use national compute resources. These compute capabilities may be provided through a mixture of onsite resources, or expertise easing use of offsite commercial resources, as needed by the customer.
Science Offerings
- Science Lifecycle User Support: This is the most important component of Science Accelerator. We provide a “high-touch” support model that attempts to identify science user project lifecycle needs, and to assist science efforts by ensuring users have access to whatever IT resources they need, regardless of whether they are provided organically by LBL or not. Depending on the customer, we use either an embedded support or a consulting model to provide research IT expertise to scientific programs. Customer needs, engagement history, and services being provided are tracked and managed by the Science Accelerator team. Depending on services needed, Science Services may coordinate responses involving multiple parts of the IT Division, for example when a customer needs faster LabNet connectivity to support instrument data collection. Science Accelerator engagements with customers are managed by a Science Services engagement lead who works to develop, maintain, and keep a long term (hopefully multiproject) relationship with science staff.
- High Performance Compute (HPC): Support for large scale compute resources is provided by the Science Services HPC team. We maintain, grow, and support the Lawrencium & Savio clusters, and we also support and facilitate access to commercial HPC as needed from Amazon Web Services, Google Cloud Platform, and other sources as needed.
- Mid-Tier Scientific Computing: Through our Scientific Virtual Machine (SVM) service, we provide access to highly reliable, customizable compute resources on-demand that are suitable for simulation, data analytics, and other scientific applications. Our SVM service supports automatic synchronization of libraries and applications used on Lawrencium, allowing users to easily develop codes that will perform at HPC scale as needed.
Storage as a Service: We support a number of storage options depending on latency, access, and data workflow needs, and provide scientific data management design and implementation support. Existing, or in process of deployment, resources include:
A scratch file-system supporting HPC
Access to the full set of AWS, GCP and GSuite hot and cold storage offerings, including free Globus-to-GDrive connectivity
High speed, highly reliable project storage
High speed data transfer and cache between on- and off- premises storage via a dedicated flash memory component (AvereFXT)
Unlimited Google Drive storage
Unlimited endpoint data back-up
Access to the Lawrencium Storage Condo to meet dedicated project needs
And many more options for transfer & archiving, connectivity to external resources (NERSC, SDSC, UCB, etc) via our Data DMZ and Data Transfer Nodes
- Science Accelerator Portal (in development): The portal provides a central knowledge base, information on services, and how science users can work with the Science Services team to get these resources. The purpose of the portal is not to simply list what is available, rather we want to enables the development of scientific campaigns, through the tailored provision of integrated support and research IT infrastructure.
If you want further information regarding Science Accelerator and our Services, enter a help ticket by sending an email to help@lbl.gov requesting a consultation meeting, it's FREE!