B08_SOL_Attachment_3_-_Brand_Name_Justification_for_an_Exception_to_Fair_Opportunity_Redacted.pdf

United States Department of the Interior

INTERIOR BUSINESS CENTER

Washington, DC 20240

SINGLE SOURCE DETERMINATION

UNDER THE SIMPLIFIED ACQUISITION THRESHOLD (SAT)

FOR COMMERCIAL PRODUCT OR SERVICE

AUTHORITY: RFO 12.102(a)

Solicitation No. 140D0426Q0576

1. Contracting activity/agency.

Department of the Interior (DOI), Interior Business Center (IBC), Acquisition Services Directorate (AQD), U.S. Department of Health and Human Services, National Institutes of Health (NIH), National Cancer Institute (NCI), Center for Biomedical Informatics and Information Technology (CBIIT), Infrastructure Branch (IB).

2. Description of supplies or services required to meet agency needs.

NCI requires University of Chicago’s Globus High Assurance Subscription and Globus for Amazon S3 High Assurance Level support for Government-owned equipment. The NCI produces and receives many large data sets that need to be transferred via network/internet connections. Globus is the industry standard tool used for these transfers, allowing NCI to efficiently move these large data sets.

Software updates and revisions shall be provided, to include: all minor updates and major upgrades that are released and/or recommended during the term of the support plan. Defective software shall be replaced at no additional cost to the Government.

The items covered include the following:

Description Period of Performance Globus High Assurance Subscription Renewal 8/18/2026 – 8/17/2027 Globus for Amazon S3 (High Assurance Level) 8/18/2026 – 8/17/2027 the period of Performance 08/18/26 – 08/17/27.

3. Identification of the single source to be solicited.

The purpose of this acquisition is to allow the NCI access to Globus - the high-performance file transfer tool supported by many of the large super-computing sites and collaborators for the purpose of performing large scale parallel data transfers between NCI, NIH, and non-NCI locations. The University of Chicago has noted that Carahsoft is a reseller. Therefore, this acquisition will be solicited on a brand name basis utilizing the procedures in Revolutionary Federal Acquisition Regulation (FAR) Overhaul (RFO).

4. Basis for Single Source Determination.

The Contracting Officer has determined that only one source is reasonably available.

☒ Brand-name requirement – The particular brand name is essential to the requirement and market research indicates similar products do not meet or cannot be modified to meet agency needs.

5. Detailed Rationale.

Globus is the only secure, Federal Information Security Modernization Act (FISMA)-aligned, high-assurance large-data-transfer platform in wide use across national supercomputing centers and biomedical research institutions. The service is critical to NCI research operations, supporting secure data transfers tied directly to federally funded cancer research efforts. The NCI uses Globus as its standardized, validated mechanism for transferring protected and high-volume scientific datasets. A lapse would disrupt active scientific studies and undermine existing data pipelines.

The agency requires:

• High Assurance capability with comprehensive audit logging, session/device isolation, and additional authentication assurance

• Integration with NCI’s existing managed endpoints

• Compliance with NIH security requirements for sensitive biomedical data

• Continued compatibility with established workflows and multi-institution collaborations

No alternative platform offers the combined features of:

• High-performance parallel data transfers

• FIPS-aligned data security controls

• Integrated identity assurance

• Enterprise-grade logging and auditing

• Direct compatibility with existing NCI Globus endpoints

Disrupting these workflows would significantly delay data movement, require revalidation, and incur major operational cost.

Only Globus can fulfill NCI’s high-assurance technical, security, and interoperability requirements.

6. Market Research.

A technical market review of alternative high-speed data transfer utilities in use across supercomputing and biomedical centers was conducted via web searches by the program office. The following alternatives were evaluated and determined to be not viable for the specific operational and infrastructure requirements:

1. IBM Aspera (Faspex / High-Speed Transfer Server)

• Reason for Rejection (Cost, Architectural Siloing, and IAM Isolation): While IBM

Aspera supports a high-speed web GUI and connects to cloud storage, it is technically non-viable for our multi-tier environment due to its rigid licensing architecture. Aspera utilizes a proprietary, siloed licensing model where the software is bound to specific storage nodes. To bridge NCI’s entire storage ecosystem ([Lustre/GPFS, local servers, and S3 clouds]), the institution would be forced to purchase, deploy, configure, and maintain separate enterprise licensing systems for each individual storage backend. The total cost of ownership (TCO) and administrative overhead is severely prohibitive compared to Globus. Furthermore, Aspera lacks native integration with academic federated identities forcing internal IT staff to manually manage separate access credentials for every external researcher.

2. Open-Source Command-Line Utilities High-Performance (HPN)-SSH, BBCP, Fast Data Transfer (FDT)

• Reason for Rejection (Lack of Web GUI and High Operational/Scripting Overhead):

Open-source point-to-point transfer tools such as HPN-SSH, FDT, and BBCP are highly efficient for raw command-line data piping, but fail to meet the mandatory requirements of NCI’s facility:

o No Web GUI: These solutions operate strictly via a Command-Line Interface (CLI). They require specialized technical proficiency, rendering them completely unusable for clinical staff and medical investigators who require a drag-and-drop web portal.

o Siloed Identity & Local Account Overhead: These tools cannot perform virtual, temporary "Guest Sharing." To transfer or share data, external collaborators must be manually provisioned with local Linux operating system accounts and SSH keys across NCI’s internal firewalls, creating a massive administrative burden and increasing the security attack surface.

o No Native Automation or Resilience: These utilities lack robust, automated "fire-and-forget" error recovery. If a network drop or timeout occurs during a multi-terabyte transfer, the job fails or hangs. To achieve parity with Globus, NCI’s team would be forced to write, maintain, and audit complex custom wrapper scripts to verify checksums and resume interrupted transfers.

3. MinIO Enterprise (Object Storage Gateway Layer)

• Reason for Rejection (Incompatible Storage Paradigm): MinIO provides an excellent web GUI and bridges local arrays, but it functions exclusively as an object storage platform. Adopting MinIO would require transforming all our underlying parallel file systems into an S3-compliant object store. This breaks a substantial portion of NCI’s downstream bioinformatics and data analytics pipelines, which strictly rely on direct POSIX file paths (e.g., traditional tools expecting standard Linux directory paths like /mnt/cluster/data/). Additionally, MinIO cannot orchestrate native, high-speed point-to-point transfers to external, non-MinIO systems at other universities.

4. Nextcloud Enterprise

• Reason for Rejection (Severe Performance Bottlenecks): Nextcloud offers a highly functional, consumer-friendly web interface suitable for administrative workflows and small documents. However, its software architecture is written in PHP and is fundamentally incapable of multi-stream, high-throughput parallel data movement over Wide Area Networks (WANs). Attempting to utilize Nextcloud for massive datasets (such as multi-terabyte genomic sequencing data or raw structural biology images) results in persistent server timeouts, web server buffer overflows, and high rates of packet drop/file corruption.