| Summary: | In the context of the gradual obsolescence of Moore's Law, chip manufacturing
technology has entered the post-Moore era. Traditional semiconductor technology
has encountered physical and economic bottlenecks, making it impossible to
continue improving performance by reducing the process size. Quantum computing,
as a cutting-edge technology, leverages the fundamental principles of quantum
mechanics, such as quantum superposition, quantum entanglement, and quantum
interference, showing great potential. Quantum computers, by operating on qubits,
can achieve processing capabilities far beyond classical computers, particularly in
solving complex optimization problems, machine learning, cryptography, and
financial analysis.
In quantum systems, controlling and reading qubits require precise microwave pulses.
These pulse signals must have fast frequency switching characteristics, low phase
noise and high frequency resolution. This thesis provides a detailed introduction to
the design and optimization of waveform generators based on Direct Digital
Synthesizers (DDS), exploring the performance of different waveform generation
methods in terms of chip area, power consumption, and maximum operating
frequency. In this thesis, a new DDS architecture that achieves performance
optimization through techniques such as memory folding, multi-level storage
structure, and rolling window readout in lookup table design is proposed.
Through simulation and synthesis result analysis, the work analyzes and compares
the performance of different DDS architectures in terms of signal quality, system
latency, and power consumption. Several improvement methods are ultimately
proposed, including the introduction of Digital Phase-Locked Loops (DPLL) to
reduce phase noise, the use of Gaussian envelope-modulated microwave pulse
signals to enhance the accuracy of qubit operations, and the optimization of output
signal quality through digital-to-analog converters and low-pass filters.
|
|---|