Modelling and Design of Multi-Parameter Optimization Techniques for Superconducting Comparators and ADCs
Abstract
The original university lab demonstrations of circuits that can divide a data stream by two in the 1990s above 750 GHz raised the expectation that superconducting electronics (SCE) based ADCs and digital processing circuits ought to be clockable above 100 GHz without any time interleaving/parallelization. However, so far, the complexity of fully optimizing the current biasing schemes (which compensate for the fabrication non-uniformities of the underlying 2 terminal Josephson Junction devices) has prevented complex circuits from running that fast. In particular, an 8-bit ADC can have as many as 48 independent biases to be optimized. A new design iteration can require a week of manual labor to have its performance optimized. Thus, the research objective of this work is to develop closed-loop, low-latency and scalable hardware optimization methods to allow each circuit to be optimized. In addition to automating the setup of individual die, one major benefit of this work will be to allow prediction from fabrication parameter gradients of the likely ADC performance. This feedback to the fab as to which parameters are most critical to better control can only help the SCE community. It is also believed that the cost function mechanism of determining optimal biasing can be generalized and applied to other ADC and, indeed, to DSP circuits. The proposed research brings together novel concepts and tools from signal processing to SCE hardware and mixed-signal circuit design, to advance basic science and engineering of future exascale TB/s communication links and serial processors. The investigation of the fundamental sub-system constraints and higher level integrated system modeling and optimization is expected to unravel cross-domain design techniques for developing highly efficient and scalable heterogeneous temperature based multi-platform systems. The research will benefit several emerging millimeter-wave cases, such as backhaul, fiber-to-home, and mobile access networks. It will also impact other more DoD specific areas, including radar, imaging, sensing, localization, and vehicular communication networks.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 09, 2020
- Source ID
- N000142012109
Entities
People
- Subhanshu Gupta
Organizations
- Office of Naval Research
- United States Navy
- Washington State University