NB688000-21-02120 CD-492_Redacted.pdf

DatePhone No.

ting - Not exceeding $ 0,000:

Title

APPROVAL REQUIREMENTS

Date

Advocate - Over $ ,000 but not exceeding $ million:

Title gnature Phone No.

Date

- Over $ million:

Title Signature

NOTE: Use the total cost estimate specified on the front of this document in determining the necessary approvals. Each review must be preceded by lower level approval(s), e.g., over $ million all approvals are required.

DatePhone No.

SEE ATTACHED.

MEELONDRELL

LUMPKIN

Digitally signed by MEELONDRELL

LUMPKIN

Date: 2021.07.31 09:31:50 -06'00'

This contract requires the implementation of technically demanding, detailed modifications to an existing piece of complex custom software. This ARTIQ experimental control framework was developed in previous contracts with close collaboration between NIST and M-Labs. The contract also stipulates maintenance of aspects of the ARTIQ system such that compiled ARTIQ binaries will be provided for NIST hardware variants (now no longer the standard ARTIQ hardware variants used for testing) to include additional feature developments and protect against regressions. The implementation of the modifications requires a very detailed understanding of ARTIQ. This includes an understanding of the architecture of ARTIQ and the various scientific and technical reasons why certain design choices were made; a demonstrated track record of successful significant contributions to the ARTIQ system; expertise with the many software tools used in ARTIQ, including Python, Rust, Migen, MiSoC/mor1kx, LLVM, llvmlite, and Qt, and pyQt; expertise in FPGA and embedded systems design, including but not limited to high-performance open-source TDC/DTC cores, the mor1kx processor, Zynq processors, the wishbone bus architecture, and MiSoC; and proven ability to work with scientific staff to debug complex, interrelated software and hardware problems arising in the laboratory environment.

The ARTIQ experimental framework was developed as a custom software/hardware package for NIST by M-Labs, Ltd. Because they are the developers of ARTIQ, M-Labs is uniquely suited to develop extensions and modifications to ARTIQ. M-Labs alone possesses the full set of required criteria given above, which are absolutely essential to a time-efficient, cost-efficient implementation of the deliverables in this contract.

In the bids for the original ARTIQ contract (SB1341-14-SU-0271 - May 2014) and responses to a sources sought notification (December 2013), no other potential bidder demonstrated familiarity with the full set of technologies employed in ARTIQ (as listed above). Furthermore, for previous ARTIQ modifications contracts (SB1341-15-SE-0363 - September 2015, SB1341-16-SE-0309 - September 2016, 1333ND19PNB680914 - September 2019), there were no responses from other vendors to the notification of intent to sole source to M-Labs. Furthermore, nobody other than M-Labs has made contributions at the required level of complexity to the existing ARTIQ system, demonstrating understanding of the architecture and design philosophy and mastering the necessary technologies to implement features successfully within the ARTIQ framework.

It is, in theory, possible that other programmers could be contracted to modify and provide necessary maintenance on ARTIQ since the source code is not proprietary to M-Labs. However, because no other potential programmers have the requisite experience either with ARTIQ or with all of the technologies used in ARTIQ, working with such programmers would require them to conduct extensive research to learn about and become familiar with ARTIQ its included technologies. The amount of time required would depend on the programmer's existing skills. Still, we estimate that a minimum of 4 weeks of full-time effort would be required to reach a functional level of expertise required to begin work on the deliverables in this contract.

Furthermore, such programmers would also require extensive discussions with the NIST group, duplicating effort already performed with M-Labs throughout ARTIQ development, regarding the scientific rationale for different design choices, our detailed preferences and requirements for software functionality, and an understanding of the laboratory and connected hardware ecosystem which ARTIQ operates. Based on the amount of time already committed to this process by NIST and M-Labs in previous ARTIQ contracts, we estimate that this would involve at least two weeks of full-time work by both NIST and the potential alternate contractor (in addition to the minimum 4 weeks mentioned above for learning new technologies), involving multiple personnel on both sides. This represents an unacceptably significant time commitment of NIST staff to a task that exactly duplicates discussions previously carried out with M-Labs, given the many other work responsibilities of NIST staff.

For these reasons, M-Labs, Ltd. can reasonably produce the deliverables in this contract in a reliable and timely manner. The use of alternative contractors would create an unacceptably high risk of failure (due to lack of requisite expertise) and cause an unacceptably large time burden on NIST staff.