Acknowledging ACCRE in Publications
When possible, we would greatly appreciate an acknowledgment of ACCRE in presentations and publications. Here is our suggested text:
This work leveraged the resources provided by the Vanderbilt Advanced Computing Center for Research and Education (ACCRE), a collaboratory operated by and for Vanderbilt faculty. ACCRE is comprised of over 3,000 researchers from more than 40 campus departments and six schools have utilized ACCRE for their research and education programs. The ACCRE computing cluster currently boasts over 16,000 processor cores and continues to expand.
We also maintain an up-to-date list of publications resulting from research conducted using ACCRE resources, including journal articles and conference proceedings, on our website.
Grant Assistance and Letters of Support
We are delighted to assist you in developing proposals that incorporate ACCRE resources. Additionally, we offer support letters for the portion of your research and educational endeavors that rely on our facility. Please don't hesitate to reach out to ACCRE Administration for any assistance you require, including budgetary considerations.
Summary Describing the ACCRE Facility
For your reference, here is a general description of the ACCRE facilities. Please feel free to use any or all of the included text. For more detailed technical information, consult the technical details page and FAQ.
Advanced Computing Center for Research and Education (ACCRE) at Vanderbilt University
The Vanderbilt Advanced Computing Center for Research and Education (ACCRE) is a researcher-driven collaboratory, operated by and for Vanderbilt faculty. It was initially funded by a seed grant of $8.3 million from Vanderbilt, with the vision of becoming self-supporting through contributions from Vanderbilt researchers. Over 3,000 researchers from more than 40 campus departments and five schools have harnessed ACCRE for their research and education programs. ACCRE's computing cluster currently comprises over 16,000 processor cores and is continuously expanding. Funding for all hardware comes from external grants or startup funds contributed by collaboratory faculty. ACCRE is supported by a team of 17 personnel responsible for cluster maintenance, operation, user education, and outreach.
The ACCRE high-performance computing cluster features over 720 nodes each equipped with 128 or 256 GB of memory. All compute nodes run a 64-bit Linux OS, featuring 250 GB – 1 TB hard drives and dual copper gigabit Ethernet ports. Sixty compute nodes are equipped with 4X Nvidia Titan X or GeForce RTX 2080 Ti GPU cards and are interconnected via a low-latency 25 or 40/56 Gb/s RoCE network. Recent additions to the cluster include 20 compute nodes containing 4X NVIDIA A6000 GPU cards, with an additional 18 A6000 cards being installed in existing servers. Nagios continuously monitors all compute nodes. Resource management, job scheduling, and usage tracking are managed by the integrated scheduling system, SLURM. These utilities include an "advance reservation" system for reserving node blocks for specified periods, such as educational or research purposes.
Daily Operation and Maintenance
ACCRE is staffed by a dedicated team of 19 support personnel, including nine system administrators, programmers, and researchers, with a combined experience of over 80 years+ in computing. Additionally, a six-person team handles facilitation and software engineering tasks. Support for system services is available 24/7/365 for urgent issues, with on-call support during nights and weekends. Cluster uptime has exceeded 95% over the past five years. An online support ticket system is used to track and resolve problems and user questions. ACCRE staff are responsible for maintaining core system hardware, core system software, networking, user support, tape backup services, disk storage, logistical storage development work, education, and management/finance support.
Research Security at ACCRE: Key Measures
ACCRE places a paramount focus on the security and privacy of user data. Here are the key security measures and practices in place:
- Data Center Location: Sensitive research data at ACCRE is exclusively stored on file, database, or computational servers located within Vanderbilt's highly secure data centers in Davidson County, TN. These data centers are specifically approved for handling a range of sensitive data types, including FERPA, Export Controlled, PII, HIPAA, and Sensitive Human Subject Research data.
- Science DMZ Security Model: ACCRE follows the Science DMZ security model, designed to provide a secure and high-performance environment for scientific data sharing and analysis, ensuring data protection while enabling efficient research collaboration.
- Continuous Monitoring: All servers within ACCRE's environment are continuously monitored, 24/7, for network and physical intrusion attempts. This vigilant monitoring helps identify and respond to any potential security threats promptly.
- Patch Management: ACCRE's dedicated staff handles server patching based on risk assessments and availability requirements, ensuring servers are protected against known vulnerabilities, reducing the risk of security breaches.
- Group-Based Access Controls: Access to sensitive data is tightly controlled through group-based access permissions, adhering to the principle of least privilege, minimizing unauthorized access risks.
- Isolated Server Instances: ACCRE employs separate server instances within protected data centers, facilitating secure data analysis without removing data from the facility, ensuring data security.
- Remote Access Security: Stringent security measures govern remote access to the ACCRE environment, including the use of strong encryption methods and active accounts in two distinct directories. These measures ensure that only authorized personnel can access the environment remotely.
- Data Disclosure Requirement: Principal Investigators (PIs) must complete a data disclosure before creating a research group and allocating resources.